RECOMMENDED FOR PUBLICATION
                               Pursuant to Sixth Circuit I.O.P. 32.1(b)
                                      File Name: 25a0059p.06

                   UNITED STATES COURT OF APPEALS
                                 FOR THE SIXTH CIRCUIT



                                                            ┐
 KARU GENE WHITE,
                                                            │
                               Petitioner - Appellant,      │
                                                             >        No. 21-5958
                                                            │
       v.                                                   │
                                                            │
 LAURA PLAPPERT, Warden,                                    │
                              Respondent - Appellee.        │
                                                            ┘

Appeal from the United States District Court for the Eastern District of Kentucky at Lexington.
                  No. 5:02-cv-00492—Karen K. Caldwell, District Judge.

                                    Argued: June 27, 2024

                             Decided and Filed: March 14, 2025

            Before: BATCHELDER, STRANCH, and THAPAR, Circuit Judges.

                                     _________________

                                           COUNSEL

ARGUED: Margaret O’Donnell, ATTORNEY AT LAW, Frankfort, Kentucky, for Appellant.
Elizabeth Hedges, OFFICE OF THE KENTUCKY ATTORNEY GENERAL, Frankfort,
Kentucky, for Appellee. ON BRIEF: Margaret O’Donnell, ATTORNEY AT LAW, David M.
Barron, KENTUCKY DEPARTMENT OF PUBLIC ADVOCACY, Frankfort, Kentucky, for
Appellant. Elizabeth Hedges, Bryan D. Morrow, OFFICE OF THE KENTUCKY ATTORNEY
GENERAL, Frankfort, Kentucky, for Appellee.

       THAPAR, J. (pp. 2–47), (app. 48–53), delivered the opinion of the court in which
BATCHELDER, J., concurred. STRANCH, J. (pp. 54–86), delivered a separate dissenting
opinion.
 No. 21-5958                             White v. Plappert                                 Page 2


                                        _________________

OPINION

                                        _________________

       THAPAR, Circuit Judge. In 1979, Karu Gene White used a crowbar to beat a blind
seventy-five-year-old and two other seniors to death while robbing them. In the forty-five years
since, White has offered various reasons why he shouldn’t be executed.

       Here, he argues that Kentucky shouldn’t have sentenced him to death because his trial
counsel failed to investigate and present mitigating evidence. Neither argument holds up. Thus,
we affirm the district court’s denial of his habeas petition.

                                                  I.

                                     A. Factual Background

       After months of planning, White and two accomplices perpetrated a horrific home
invasion. His target? A small, coal-heated home tucked away in the heart of Appalachia. On
the first floor of that house, Charlie Gross, a seventy-five-year-old minister, ran a grocery store
with his seventy-four-year-old wife Lula and her seventy-nine-year-old brother Sam, a local
farmer. The Grosses ran the store to pay the healthcare expenses of their son, Marvin, who
suffered from a mental health disorder that relegated him to a nursing home.

       White, a twenty-year-old, was a lifelong friend of the Grosses. He was looking to pick
up some extra cash and figured that the store was a perfect target. After all, he knew the Grosses
were old and disabled, making it “easy to knock them out” or kill, if need be. R. 125-1, Pg. ID
5906. White also knew that they were saving money for their “dumb boy,” as he put it. Id. at
Pg. ID 5907. And he knew that the Grosses—Great Depression survivors who didn’t trust
banks—stored their money in bulk cash around the house.
 No. 21-5958                                     White v. Plappert                                             Page 3


        As White planned the operation, he recruited two teens to assist: his seventeen-year-old
half-brother, Tommy, and Tommy’s fifteen-year-old friend, Charlie Fisher.1 He gave each
specific roles and instructions for the heist.

        On the night of the crime, the three hitchhiked to the store and waited until it was dark.
White and his accomplices then approached the store, armed, masked, and wearing gloves and
socks on their hands to avoid leaving fingerprints.                     White wielded a crowbar, with his
accomplices carrying a lug wrench and a tree limb. With White “giving the orders,” the three
sprang into action. R. 125-3, Pg. ID 6051. At White’s direction, Fisher, the youngest of the
group, walked into the store and gave the signal: He picked out a bottle of soda and set it down
on the counter loudly in front of Charlie Gross. White and Tommy then bolted through a side
door of the store and began brutally bludgeoning the three elderly owners to a gory pulp. After
the first couple blows from White and his accomplices, Sam, lying on the ground, began to beg
for his life. He pled “Oh, my God, don’t kill me.” Id. at Pg. ID 6019. Meanwhile, Charlie
attempted to stand up, raising his arms to defend against additional blows. In the face of their
feeble resistance, White’s violence only intensified: He proceeded to plunge his crowbar into the
skulls of Sam and Charlie—repeatedly—until their mutilated corpses lay silently in pools of
blood. The three murderers—their clothes covered with the Grosses’ blood—then ransacked the
house in search of cash.

        The aftermath at the store evinced White’s relentlessness. See Exs. A–J, Commonwealth
v. White, 79-CR-024 (Powell Cir. Ct. 1979) (crime-scene photographs). The coroner, state
trooper, and detective all testified that it was the most brutal murder they’d ever worked. The
coroner likened the scene to a slaughterhouse. Officers couldn’t set foot inside without wading
through a bloody morass. One of Charlie’s eyes had ruptured and fallen out of his skull. His
entire head had caved in. Lacerations covered the side of his face—so many that medical
professionals couldn’t keep an accurate count. On the floor, police also found part of Sam’s
brain seeping out from the back of his head. Lula’s body was coated with blood. Her head had
been bashed five or six times, leaving multiple seven-inch gashes on it. Muscles were detached
from the victims’ limbs, and fingernails were stripped from their fingers. And when it came time
        1
            To avoid ambiguity between Charlie Gross and Charlie Fisher, we refer to the latter as “Fisher.”
 No. 21-5958                            White v. Plappert                                 Page 4


to prepare the bodies for burial, the coroner had to use large vinyl “disaster pouches” to contain
the pile of each victim’s remains. R. 125-1, Pg. ID 5932.

       As for the murderers? They escaped with $7,000 and a handgun. Afterward, White
urged his accomplices to dispose of evidence and obstruct the murder investigation.

                                             B. Trial

       Police eventually tracked down the murderers. Kentucky charged them with burglary,
three counts of robbery, and three counts of capital murder.           White v. Commonwealth,

 (Ky. 1983) [hereinafter “White I”]. The state indicated it would seek the
death penalty against White.

       The suspects’ families retained two attorneys, who ended up representing the defendants
jointly. White and his codefendants interacted with their attorneys dozens of times before trial.
Each time, they maintained their innocence. According to one attorney, they “were[n’t] inclined
to even discuss the possibility of a guilty plea.” R. 124-25, Pg. ID 4944.

       The defendants also had an alibi. They told their attorneys that they were at a nearby
college dance the night of the murders. An eyewitness corroborated that story. And White’s
family members believed as much. Thus, counsel never had “any hint” that the defendants had
committed the crimes. R. 124-8, Pg. ID 3621.

       For its part, the state’s case against the defendants was weak. It had only circumstantial
evidence and no testimony or direct evidence placing them in the store at the time of the
murders. Desperate for a stronger case, the state offered immunity agreements to both Tommy
and Fisher in exchange for testimony against White. Both defendants rejected the offers. Why?
“[I]t was not possible” for them to testify, they claimed, because they “had no knowledge of any
guilty conduct.” R. 124-7, Pg. ID 3506.

       Armed with this information from the defendants, their families, and the state, White’s
counsel prepared an alibi defense.      They devoted most of their efforts to shoring up the
defendants’ story, finding ways to poke holes in the state’s circumstantial case, and showing how
it would have been “utterly out of character” for the defendants to commit murders like these.
 No. 21-5958                           White v. Plappert                                Page 

-25, Pg. ID 4921. And, if the jury voted to convict, the attorneys planned to “put on
family, friends, and neighbors” to testify to the defendants’ tame backgrounds, limited criminal
records, and young ages. 

 at Pg. ID 4942.

       As part of this preparation, White’s counsel devoted at least a dozen hours to
interviewing White’s family, friends, and others who knew him.           These included “long
discussions” with White’s relatives, during which counsel asked about his behavioral history and
the “kind of kid he had been.” 

 at Pg. ID 4922–23. White’s family informed counsel that he
was “a good little boy.” 

 They gave “no indication” “at any time” about any “erratic” details
in his childhood, such as child abuse or exposure to traumatic events. 

 at Pg. ID 4921.
Counsel, for their part, didn’t ask any questions on these topics. Given the promise of an alibi
approach, counsel believed a deep dive into White’s childhood wouldn’t have been an efficient
use of time.

       The case took an abrupt turn on the fourth day of voir dire. That day, Fisher changed his
tune. He now agreed to testify against White in exchange for total immunity. White’s attorneys
withdrew from their representation of Fisher, and a new attorney was appointed to represent him.
With an eyewitness prepared to take the stand, an alibi defense was no longer viable; White’s
counsel pursued an insanity defense instead.

       To buy time to develop this new defense, counsel moved for an indefinite continuance
and to redo voir dire. The court denied these motions but appointed a psychologist and ordered a
competency hearing for White. That psychologist found White competent to stand trial.

       Over the next two weeks, counsel feverishly prepared White’s new insanity defense.
They hit an initial roadblock: White’s retained psychologist concluded that he wasn’t insane.
So, counsel attempted to prove insanity through lay testimony. See Jewell v. Commonwealth,

 (Ky. 1977) (holding defendants may establish insanity through lay
testimony), overruled in part on other grounds by Payne v. Commonwealth, 

(Ky. 1981).
 No. 21-5958                                  White v. Plappert                                          Page 6


        As part of this effort, counsel performed a second round of interviews with many of
White’s family members, including his grandmother, mother, stepfather, four of his siblings, and
one of his aunts.

        With the alibi defense gone, White’s family started to sing a different tune. No longer
did White have a tame childhood. Instead, the family revealed a number of “old skeletons in the
closet” about his childhood. R. 124-7, Pg. ID 3512; R. 124-11, Pg. ID 3919. These interviews
uncovered harrowing tales about White’s traumatic upbringing.                      A few examples suffice:
White’s mother attempted to suffocate him when he was an infant. As a five-year-old, White
witnessed firsthand his uncle’s gruesome murder of his father. And his mother, who tried and
failed to put him up for adoption on the black market, hated White—and White knew it. Counsel
presented much of this testimony to the jury during White’s trial. See Table 1, infra (left
column).

        At trial, counsel argued that White’s horrific upbringing left him insane—unable to
conform his conduct to law. Counsel pointed to White’s stealing, sexual advances towards his
sisters,2 and his physical and sexual abuse of animals as evidence of White’s trauma-induced
insanity.

        Counsel’s insanity defense came up short. The jury found White guilty on all counts.
The penalty phase immediately followed. There, both parties agreed not to introduce additional
evidence. White’s counsel recapped the mitigating evidence from the guilt phase, including
White’s extensive abuse as a child, drug problems, and exposure to violence and murder at a
young age. “Hundreds of days” of abuse and violence, counsel argued, had stripped away
White’s capacity to choose right over wrong. R. 124-18, Pg. ID 4537. According to counsel,
that incapacity born of trauma mitigated his culpability for the crimes.

        The jury found the presence of a statutory aggravator: White was engaged in a first-
degree robbery or first-degree burglary when he committed his murders. And it unanimously
concluded the aggravating factors outweighed any mitigating evidence.                           The jury thus

        2
           White argues some of the testimony about these sexual advances was partially fabricated. We discuss this
in greater detail in Part IV.C.2, infra.
 No. 21-5958                            White v. Plappert                                  Page 7


recommended death sentences for each of the three murders. The trial judge adopted the
recommendation, noting “if there ever was a case where the death penalty is justified, this is it.”
R. 125-5, Pg ID. 6107.

                                  C. Subsequent Proceedings

       White appealed his conviction and sentence on numerous grounds.             The Kentucky
Supreme court affirmed, and the U.S. Supreme Court denied certiorari. White I, 

; White v. Kentucky, 

 (1984).

       Afterward, White sought post-conviction relief in state court. Among other claims, he
argued his trial counsel rendered ineffective assistance by failing to sufficiently investigate and
present mitigating evidence. White argued this alleged failure prejudiced him in the penalty
phase of his trial. Had counsel performed a reasonable investigation and introduced additional
mitigation evidence at trial, he argued, the jury might not have voted for death. After filing for
relief, White took no action on his claims for eight years, despite carrying the burden of
prosecuting them under Kentucky law. Mem. Op. 2–3, White v. Commonwealth, 94-SC-326-MR
(Ky. Feb. 22, 1996) [“White II”], withdrawn, Order Granting Pet. Reh’g, id. (Ky. Aug. 29, 1996).
The trial court then dismissed those claims, finding that none merited relief under Kentucky’s
post-conviction rules. Order 1–4, Commonwealth v. White, 79-CR-24 (Powell Cir. Ct. Apr. 18,
1994). The Kentucky Supreme Court eventually vacated that order and directed the trial court to
hold an evidentiary hearing. Order Granting Pet. Reh’g, White v. Commonwealth, 94-SC-326-
MR (Ky. Aug. 29, 1996).

       The hearing occurred in 1999, some two decades after White’s original trial. Forty-three
witnesses testified, including White’s trial counsel, trial experts, family members, school
principal, and neighbors. Mem. Op. 2–3, White v. Kentucky, No. 94-SC-326-MR, at 5 (Ky. May
16, 2002) [hereinafter “White III”]. During these depositions, additional details about White’s
child abuse and violent household emerged. So did testimony about White’s history of head
injuries, seizures, and mental-health issues.    Relevant testimony from these proceedings is
presented in the right-hand column of Table 1, infra.
 No. 21-5958                            White v. Plappert                                  Page 8


       The trial court denied relief on all grounds, and the Kentucky Supreme Court affirmed.
See generally White III. In affirming, the state supreme court held that counsel’s pretrial
investigations were reasonable given the circumstances. Id. at 3–6. And it found that counsel
uncovered and presented sufficient mitigating evidence to overcome a claim of deficient
performance. Id. Lastly, the court concluded that “[e]ven if all the testimony White suggests
had been introduced during the penalty phase,” it would not have “changed the sentence of the
jury.” Id. at 6. The U.S. Supreme Court again denied certiorari. White v. Kentucky, 

 (2003).

                                D. Federal Habeas Proceedings

       White then filed for federal habeas relief on various grounds, including ineffective
assistance during the penalty phase. At White’s request, the district court stayed proceedings so
he could pursue an intellectual-disability claim in state court. See Atkins v. Virginia, 

 (2002). White then “refused” to present the merits of that claim in state court for the next
nine years. R. 58, Pg. ID 1087. Instead, he argued the state had to first pay for an independent
expert to perform a mental evaluation of him. The Kentucky Supreme Court disagreed. See
Commonwealth v. Paisley, 

 (Ky. 2006). Two years later and pursuant to the
Kentucky Supreme Court’s decision in Paisley, the lower court ordered White to undergo a
psychiatric evaluation at a state-run facility. White “flatly refused to cooperate,” R. 58, Pg. ID
1087–88, and he petitioned the Kentucky Supreme Court to prohibit the lower court from
ordering him to submit to the evaluation. See White v. Payne, 

 (Ky. 2010), as
modified on denial of reh’g (Mar. 24, 2011). The Kentucky Supreme Court denied White’s
petition. 

. At this point, the presiding state-court judge retired, leaving White’s case in
limbo for over a year. Thus, even though the district court had stayed federal proceedings so
White could exhaust a mental-disability claim, “the better part of a decade” had passed without
any adjudication of the merits of that claim. R. 58, Pg. ID 1089.

       White then attempted to use these delays—to which he’d contributed—as grounds for
excusing his failure to exhaust that mental-disability claim so that he could include it in his
federal habeas petition. The district court denied that motion but ordered White’s federal habeas
action to proceed on his exhausted claims. The court then denied relief on all grounds. White v.
 No. 21-5958                            White v. Plappert                                  Page 9


White, No. CV 5:02-492 (KKC), 

 (E.D. Ky. Sept. 16, 2021). With respect to
White’s penalty-phase ineffective-assistance claim, the court concluded (1) counsel’s mitigation
investigation was reasonable given his ostensibly strong alibi defense and his family’s initial lack
of candor, (2) the mitigating evidence would not have made a difference to the jury, and (3) the
Kentucky Supreme Court reasonably applied Supreme Court precedent in reaching these
conclusions. 

 at *75–81. The district court then denied White a certificate of appealability.
But this court granted him a certificate of appealability on penalty-phase ineffective assistance of
counsel. We now affirm.

                                                II.

       To be eligible for federal habeas relief, White must show that “law and justice require
relief.” Brown v. Davenport, 

 (2022) (citing 

).

                                             A. Law

       Start with the law.       Because the Kentucky Supreme Court adjudicated White’s
ineffective-assistance claim on the merits, the Antiterrorism and Effective Death Penalty Act of
1996 (AEDPA) applies. See generally 

. “AEDPA recognizes a foundational
principle of our federal system: State courts are adequate forums for the vindication of federal
rights”—including constitutional ones. Burt v. Titlow, 

 (2013). In doing so,
AEDPA “erects a formidable barrier to federal habeas relief for prisoners” like White “whose
claims have been adjudicated in state court.” 

 White must make two showings to surmount
this barrier and satisfy habeas’s “law” prong. See James v. Corrigan, 

, 394–95 (6th
Cir. 2023). First, he needs to overcome AEDPA’s relitigation bar: White must prove that the
Kentucky Supreme Court either contravened or unreasonably applied clearly established
Supreme Court precedent. 

(d). Second, he must show that his sentence violates
the Constitution. 

 § 2254(a). Here, that means proving he received ineffective assistance of
counsel under Strickland v. Washington, 

 (1984).
 No. 21-5958                              White v. Plappert                             Page 10


                                      1. AEDPA Relitigation Bar

       Because “phrasing mirrors thought,” we must precisely articulate “the true issue before a
federal court” in this context. Brown v. Allen, 

 (1953) (opinion of Frankfurter,
J.). The issue before us is whether there has been an “extreme malfunction[] in the state criminal
justice system[].” Harrington v. Richter, 

, 102–03 (2011) (quotation omitted). We
are not concerned with ordinary errors. 

 “[E]ven serious errors aren’t enough to warrant
relief.” James, 85 F.4th at 393. To overcome AEDPA deference, a habeas petitioner like White
must show the state court’s last reasoned decision is either “contrary to” or an “unreasonable
application of” clearly established Supreme Court precedent.

       A decision can be “contrary to” clearly established precedent in one of two ways: it
arrives at a different outcome than the Supreme Court did when faced with materially
indistinguishable facts, or it applies a different legal rule than one that the Supreme Court set
forth. Williams v. Taylor, 

 (2000). White argues the latter: that the Kentucky
Supreme Court applied the wrong legal standard.

       When determining whether a state court contravened clearly established Supreme Court
precedent, federal courts’ deference to the state court’s decision is “near its apex.” Sexton v.
Beaudreaux, 

 (2018) (per curiam).            What is the precise extent of that
deference? We first presume that state courts “know and follow the law.” Woodford v. Visciotti,

 (2002) (per curiam). AEDPA accordingly “demands that state-court decisions
be given the benefit of the doubt.” 

 Federal courts may issue habeas relief only “in cases
where there is no possibility fairminded jurists could disagree that the state court’s decision
conflicts with [the Supreme Court’s] precedents.” Harrington, 

. So, when a
federal court is at all unsure whether the state court’s decision may have run contrary to clearly
established Supreme Court precedent or unreasonably applied that precedent, the federal court
cannot issue federal habeas relief.

       When we apply this deferential standard, the object of our deference is the state court’s
“decision” as a whole. 

(d)(1); see also Harrington, 

; Rogers v.
Mays, 

 (6th Cir. 2023) (en banc), cert. denied sub nom. Rogers v. Pounds, 144
 No. 21-5958                            White v. Plappert                                 Page 

 (2024). So, the fact that a state court opinion may use imprecise shorthand at times
provides no grounds for issuing relief. Rogers, 

. The question is whether the
state court’s bottom-line judgment and its opinion as a whole—not particular sentences or “stray
thought[s]”—adequately track the applicable Supreme Court standard. 

 at 391–92; see also
Davis v. Carpenter, 

 (6th Cir. 2015) (noting that federal courts in the AEDPA
context “review the state court’s ‘decision,’ not the court’s intermediate reasoning” (citation
omitted)). The Constitution doesn’t set mandatory opinion-writing standards on state courts, so
neither can we. Johnson v. Williams, 

 (2013). And if we wrongly “flyspeck
state-court opinions,” Rogers, 

, we risk incentivizing state courts to not issue
opinions in the first place to avoid the wrath of unduly harsh federal-court opinion graders.

       A state court “unreasonabl[y] appl[ies]” Supreme Court precedent when it uses the right
legal rule, but it reaches a bottom-line judgment that’s “so obviously wrong” that no fair-minded
jurist could possibly agree. Shinn v. Kayer, 

 (2020). In determining whether a
state court behaved unreasonably, federal courts may consult only the specific, binding holdings
of contemporaneous Supreme Court cases—not dicta, not subsequent holdings, and not general
statements of law that aren’t themselves holdings. Brown, 

. If the state court
lays out its reasoning, as the Kentucky Supreme Court did here, we apply our holistic, deferential
standard to the stated reasoning, see Wilson v. Sellers, 

 (2018), as well as
alternative reasoning that “could have supported” the court’s decision, Harrington, 

. And in evaluating a state court’s decision, we may not focus on particular snippets from the
court’s opinion. Cf. Rompilla v. Beard, 

 (2005) (determining whether court
reached an “objectively unreasonable conclusion,” not whether its analysis was objectively
unreasonable) (emphasis added); Porter v. McCollum, 

 (2009) (assessing whether
the state court’s “decision” was unreasonable). Rather, we start by presuming the state court’s
conclusion was reasonable. Woodford, 

. Then, we look at the record before the
state court and the Supreme Court’s then-controlling precedent and ask whether “any reasonable
argument” supports the state court’s decision. Davis, 

 (citation omitted). If so,
we must deny habeas. Harrington, 

. This approach is far more deferential than
many other permissive standards of review that populate the United States Reports, such as clear
error. See Virginia v. LeBlanc, 

 (2017) (per curiam). That is unsurprising: in
 No. 21-5958                            White v. Plappert                                Page 12


enacting AEDPA, Congress stopped just short of imposing “a complete bar” on relitigating
claims rejected by state courts. Harrington, 

. “If this standard is difficult to
meet, that is because it was meant to be.” 

        Thus, under either AEDPA’s “contrary to” or “unreasonable application of” tests, White
faces a steep climb for relief.

                                     2. Ineffective Assistance

        Because White raises an ineffective-assistance claim, his climb gets even steeper. The
Sixth Amendment protects a criminal defendant’s right “to have the Assistance of Counsel for
his defense.” U.S. Const. amend. VI. And a defendant suffers a Sixth Amendment violation
when a court concludes (1) defense counsel’s representation “fell below an objective standard of
reasonableness” and (2) but for counsel’s poor performance, there’s a reasonable probability that
defendant’s case would have come out differently. Strickland, 466 U.S. at 687–88, 694.

        These two requirements are commonly called Strickland’s “deficient-performance” and
“prejudice” prongs. Both prongs are “highly demanding” and among the “most deferential”
standards in American law. Shinn, 

 (citation omitted); Harrington, 

. White must show that his counsel were so incompetent that they did not function as counsel
at all. Strickland, 

. And since AEDPA applies, White must overcome two layers
of deference: deference to the defense attorney and deference to the state court. Cullen v.
Pinholster, 

, 190 (2011); Titlow, 

; Haight v. Jordan, 

,
831–32 (6th Cir. 2023) (per curiam), cert. denied, 

 (2024). That’s “a nearly
insurmountable hurdle.” James, 85 F.4th at 395.

        Start with the deficient-performance prong. At the outset, courts “must apply a ‘strong
presumption’” that counsel behaved reasonably.          Harrington, 

 (quoting
Strickland, 

). But White must do more than overcome this presumption. It’s not
enough for him to convince us that his counsel “took an approach that no competent lawyer
would have chosen.” Dunn v. Reeves, 

 (2021). White must also prove that
“every fairminded jurist would agree that every reasonable lawyer would have made a different
 No. 21-5958                            White v. Plappert                                 Page 13


decision.” 

 (internal quotation marks omitted). Thus, our “scrutiny of counsel’s
performance” is “highly deferential.” Strickland, 

.

       And if White makes that showing, he then must also establish prejudice. To establish
prejudice in the capital-sentencing context, White must first show a “reasonable probability” that
a competent attorney “would have introduced” the extra mitigating evidence.               Wong v.
Belmontes, 

 (2009) (per curiam) (citation omitted). Then, he must show a
“reasonable probability that the jury would have rejected a capital sentence after it weighed the
entire body of mitigating evidence . . . against the entire body of aggravating evidence.” 

 “A
reasonable probability means a ‘substantial,’ not just ‘conceivable,’ likelihood of a different
result.” Shinn, 

 (citations omitted). The gap between a substantial likelihood and
a preponderance of the evidence is “slight and matters ‘only in the rarest case.’” Harrington,

 (citation omitted). And when AEDPA applies, a petitioner like White must also
prove that every fair-minded judge would agree that it’s substantially likely that the weight of the
additional mitigating evidence would have changed the jury’s calculus. Shinn, 

.

       Atop all this deference, when the applicable Supreme Court precedent sets out vague
standards like “reasonably effective assistance” and “a substantial likelihood of a different
result,” AEDPA deference is even stronger than usual. 

. State courts thus have broad
discretion in the context of Strickland claims: “[B]ecause the Strickland standard is a general
standard, a state court has even more latitude to reasonably determine that a defendant has not
satisfied that standard.” Knowles v. Mirzayance, 

 (2009). And we are even
more wary of upsetting the state court’s decision on the basis of evidence that comes to light
years, much less decades, after trial. Cf. Herrera v. Collins, 

, 417–18 (1993);
Burgess v. Booker, 

 (6th Cir. 2013); Taylor v. Illinois, 

(1988). So, when a petitioner like White relies on mitigating evidence that first appears two
decades post-trial, we have yet another reason to pause before granting relief. That hesitance is
especially warranted when the petitioner is responsible for much of the delay.
 No. 21-5958                               White v. Plappert                             Page 14


                                              B. Justice

       Even if a petitioner can prove that habeas relief is warranted as a matter of law, courts
may nevertheless deny habeas through the exercise of their equitable discretion. Brown, 

. After all, habeas is a discretionary remedy. 

. Specifically, if a petitioner
like White overcomes AEDPA and Strickland’s doubly deferential gauntlet and thereby
convinces the court that “law” requires relief, he then must convince the court that “justice” also
warrants relief. 

.

                                                   *

       Altogether, White must navigate a dense forest of presumptions and deference to obtain
relief. He must overcome the presumption that his counsel performed adequately and that the
Kentucky Supreme Court knew and correctly applied the law. He must show that all reasonable
judges would agree that all reasonable attorneys would have performed differently than his. He
must prove that all fair-minded judges would have concluded that, but for his attorneys’
unreasonable performance, there’s a substantial likelihood that the jury would have changed its
mind. And he must prove that justice requires resentencing.

       White can’t make these showings. First, White hasn’t shown that the Kentucky Supreme
Court contradicted Strickland. Second, he hasn’t demonstrated that the court unreasonably
applied Strickland: reasonable jurists could conclude that White’s counsel’s performance was
adequate and didn’t result in prejudice.

                                                  III.

       The Kentucky Supreme Court’s deficient-performance and prejudice analyses didn’t
contravene clearly established Supreme Court precedent. The court consistently quoted and
paraphrased the applicable standards from Strickland. White’s arguments assailing its decision
zero in on single words and sentences in the opinion, which are not our concern. See Rogers, 69
F.4th at 391–92.
 No. 21-5958                            White v. Plappert                                Page 15


                                   A. Deficient Performance

       When a state court quotes and cites the applicable Supreme Court precedent, its decision
is not “contrary to” that precedent. See Woodford, 537 U.S. at 22–24. Throughout its opinion,
the Kentucky Supreme correctly cited and quoted Strickland when discussing deficient
performance.    See White III, at 3–6.     Under Strickland, the “proper measure of attorney
performance” is “reasonableness under prevailing professional norms.” 

; see
Bobby v. Van Hook, 

, 11–12 (2009) (per curiam) (“professionally reasonable
judgments”); Rompilla, 

 (“an objective standard of reasonableness under
prevailing professional norms” (cleaned up)). When evaluating counsel’s performance, the
Kentucky Supreme Court assessed whether the attorneys’ decisions were “professionally
reasonable.” White III, at 5 (characterizing counsel’s performance as “not clearly unreasonable
or unprofessional”). Indeed, when applying Strickland’s deficient-performance test to the issue
of mitigation investigations, the Kentucky Supreme Court paraphrased the U.S. Supreme Court.
Just as the Supreme Court assesses counsel’s investigatory decisions “for reasonableness in all
the circumstances,” Strickland, 

, the Kentucky Supreme Court stated that
counsel’s investigatory decisions must be “reasonable under all the circumstances,” White III, at
4.

       White’s counterarguments are unavailing. He argues that the Kentucky Supreme Court
contravened clearly established precedent when it distinguished Williams v. Taylor on the
grounds that Williams “involve[d] a total failure to inquire into [the petitioner’s] life or family
background.” White III, at 5; see 

. White contends that counsel can still perform
deficiently even if they perform some degree of investigation.

       While that’s true, it doesn’t help White. The Kentucky Supreme Court didn’t claim that
counsel are deficient only if they fail to perform any investigation. Rather, the court correctly
noted that Williams didn’t control here because White’s counsel did present mitigating evidence
while Williams’ counsel didn’t. In Williams, counsel “failed to conduct an investigation that
would have uncovered” reams of mitigating evidence helpful to Williams. 

.
Counsel’s “failure to introduce the . . . voluminous amount of evidence that did speak in
Williams’ favor” fell short of even Strickland’s low bar. 

. Indeed, Williams had
 No. 21-5958                            White v. Plappert                                  Page 16


argued that the jury would’ve reached a different outcome had they heard “any of this evidence.”

 (citation omitted). By contrast, White’s counsel uncovered and presented three days’
worth of mitigating testimony. Thus, the Kentucky Supreme Court didn’t misread Williams or
misapply Strickland.

       The Kentucky Supreme Court also did not contravene clearly established Supreme Court
precedent by observing that “the defendant’s own statements or actions” impact its assessment of
the reasonableness of counsel’s conduct. White III, at 4–5. That’s what Strickland said: “The
reasonableness of counsel’s actions may be determined or substantially influenced by the
defendant’s own statements or actions.” 

; see also Titlow, 

(holding that defendant’s professed innocence is a relevant factor in assessing counsel’s
reasonableness). Thus, the court didn’t contradict clearly established law (or common sense, for
that matter) when it acknowledged that a defendant’s words and actions factor into the deficient-
performance analysis.

       In short, none of these individual statements from the state court’s opinion contradict
clearly established Supreme Court precedent.

       By contrast, the court might have articulated a legal standard that contradicted Supreme
Court precedent when it stated that “[b]ecause the defendants originally claimed they were not
guilty, there was no reason to investigate White’s background or his physical or mental health.”
White III, at 5. Compare Williams, 

 (noting counsel’s “obligation to conduct a
thorough investigation of the defendant’s background”). Even if this single potential misstep did
run afoul of Supreme Court precedent, it doesn’t provide grounds for granting White habeas
relief. Recall that our task is to assess whether the state court’s decision as a whole runs contrary
to clearly established Supreme Court precedent. Harrington, 

. A single sentence
in an opinion that the state court was not even obligated to write in the first place is not our
concern; it’s whether a fair-minded judge could reasonably conclude that the state court
“proper[ly] fram[ed]” the overarching question. Woodford, 

. It did: given its
repeated citing, quoting, and paraphrasing of Strickland and overarching focus on the
reasonableness of counsel’s actions under the actual circumstances, the state court’s bottom-line
 No. 21-5958                            White v. Plappert                                 Page 17


judgment and decision as a whole did not contravene clearly established Supreme Court
precedent.

       At any rate, it’s less than clear that this piece of the Kentucky Supreme Court’s opinion
contradicted clearly established Supreme Court precedent to begin with. Recall that Strickland
itself said that “a particular decision not to investigate must be directly assessed for
reasonableness in all the circumstances, applying a heavy measure of deference to counsel’s
judgments,” and that “[t]he reasonableness of counsel’s actions may be determined or
substantially influenced by the defendant’s own statements or actions.” 

. The
Court added that “when a defendant has given counsel reason to believe that pursuing certain
investigations would be fruitless or even harmful, counsel’s failure to pursue those investigations
may not later be challenged as unreasonable.” 

 It’s true that the Williams Court spoke of
counsel’s “obligation to conduct a thorough investigation of the defendant’s background.”
529 U.S. at 395–96. But Williams didn’t confront a situation in which the defendant at first
credibly claimed innocence. That matters, since the Court has cautioned us to not frame its
holdings at too high a level of generality. Nevada v. Jackson, 

 (2013) (per
curiam). Further, though Williams cited ABA guidelines calling for investigations into the
defendant’s background, those ABA guidelines “are only guides,” not hard and fast rules or
precedents. Strickland, 

.

                                           B. Prejudice

       Likewise, the Kentucky Supreme Court didn’t contradict clearly established law
regarding prejudice. Quoting Strickland, the Kentucky Supreme Court observed that the “critical
question” is whether counsel’s performance “so undermined the proper functioning of the
adversarial process that the trial cannot be relied on as having produced a just result.” White III,
at 3 (quoting Strickland, 

).       This quote mirrors Strickland’s definition of
prejudice, which asks whether counsel’s error “undermine[s] confidence in the outcome.”
Strickland, 

. Thus, the Kentucky Supreme Court articulated the correct standard.

       The Kentucky Supreme Court was also correct to state that overwhelming aggravating
circumstances make it more difficult for a petitioner to show prejudice from counsel’s failure to
 No. 21-5958                                 White v. Plappert                                       Page 18


introduce additional mitigating evidence. White objects to the court’s statement that, “‘where
the aggravating circumstances are overwhelming, it is particularly difficult to show prejudice at
sentencing due to the alleged failure to present mitigating evidence.’” White III, at 3–4 (quoting
Bonin v. Calderon, 

 (9th Cir. 1995)). But this is a correct statement of
ineffective-assistance law. The prejudice inquiry requires courts to reweigh the aggravating and
mitigating evidence in light of the new, unpresented mitigating evidence.3 Pinholster, 563 U.S.
at 198.     Thus, when there are “overwhelming aggravating factors,” there’s usually “no
reasonable probability that the omitted evidence would have changed the conclusion that the
aggravating circumstances outweighed the mitigating circumstances and, hence, the sentence
imposed.” Strickland, 

.

        The Supreme Court’s decision in Wong v. Belmontes illustrates this commonsense
insight. There, the habeas petitioner had used a metal bar to mangle the victim’s skull. 558 U.S.
at 27. The petitioner had also committed a prior murder. Id. at 28. It was thus “hard” for the
Supreme Court “to imagine” what mitigating evidence could possibly outweigh this aggravating
evidence of heinous conduct and prior violence. Id. at 27–28. So too here: the Kentucky
Supreme Court concluded that White—whose murderous acts were comparably gruesome—
faced “particular[] difficult[y]” in showing prejudice. White III, at 3.

        White also claims that the Kentucky Supreme Court imposed a novel “nexus”
requirement on White’s mitigating evidence. Appellant Br. 80; see Eddings v. Oklahoma, 

, 113–15 (1982) (holding that state courts can’t limit mitigating evidence to only those
acts that excuse the defendant’s criminal acts). He points to the Kentucky Supreme Court’s
statement that “[e]ven if all the testimony White suggests had been introduced during the penalty
phase[,] it would not have been sufficient to counter or explain the brutalities of the three
murders.” White III, at 6. But this language merely states that White’s mitigating evidence




        3
          Kentucky’s death-penalty jury instructions indicate that the jury need not recommend death even if the
aggravating factors outweigh the mitigating factors. Based on this, White argues that reweighing the aggravating
and mitigating factors isn’t appropriate for assessing prejudice. But our caselaw makes clear that a reweighing
approach nonetheless governs the prejudice prong for Kentucky death sentences. E.g., Haight v. Jordan, 

 (6th Cir. 2023) (per curiam), cert. denied, 

 (2024).
 No. 21-5958                            White v. Plappert                                 Page 19


wasn’t enough to outweigh—or “counter”—the aggravating evidence. Again, that’s the correct
way to think about prejudice. See Thornell v. Jones, 

, 1311 (2024).

       White is also wrong to claim that the Kentucky Supreme Court erred by discounting the
persuasive value of White’s mitigating evidence because it lacked a connection to his murders.
Courts may “attach[] diminished persuasive value” to evidence that lacks a nexus with the crime
of conviction. Id. at 1310. That is what the Kentucky Supreme Court did when it noted that
White’s mitigating evidence doesn’t “explain” the brutal murders that he committed. White III,
at 6. That observation was correct as a factual matter: White doesn’t tie his unpresented
mitigating evidence to the night of the murders. It was also correct as a legal matter: the
Kentucky Supreme Court never said it would ignore any evidence that wasn’t tied to the
murders; it simply discounted the weight of that evidence because it lacked a connection with the
murders. Therefore, the court did not err. See Thornell, 144 S. Ct. at 1310.

       White’s attack on a single adverb in the Kentucky Supreme Court’s opinion also comes
up short. In the concluding sentence of the penalty phase section of the court’s opinion, it stated
that there was no “possible or reasonable chance that the omitted testimony would have had such
an unquestionably favorable impact that it would have changed the sentence of the jury.” White
III, at 6. According to White, this articulation contradicts clearly established precedent regarding
prejudice.

       Had the court omitted the word “unquestionably,” the challenged sentence would be an
indisputably correct paraphrase of Supreme Court precedent. Cf. Williams, 

(requiring “a reasonable probability that the result of the sentencing proceeding would have been
different” (citation omitted)). White’s argument thus hinges on the word “unquestionably.”
Specifically, he argues that the court’s inclusion of this single word increased his burden under
Strickland: the court would grant relief only if there’s a chance the mitigating evidence would be
“unquestionably favorable.” And because the U.S. Supreme Court doesn’t require petitioners to
show that their mitigating evidence is “unquestionably” positive, White argues that the court
contradicted clearly established precedent.
 No. 21-5958                                    White v. Plappert                          Page 20


       But squabbles over a few words—much less one adverb—aren’t enough to prove that a
state-court decision is “contrary” to Supreme Court precedent. Woodford, 537 U.S. at 23–24;
Rogers, 69 F.4th at 391–92. A state court’s usage of an “imprecise” adverb does not provide
grounds for upending its decision. Woodford, 537 U.S. at 23–24. And the Supreme Court
instructs us to “presum[e] that state courts know and follow the law.” Id. at 24. Here, one can
fairly read the Kentucky Supreme Court’s “unquestionably favorable” line as saying that White’s
unpresented mitigating evidence had no “possible or reasonable chance” of “chang[ing] the
sentence of the jury” because that evidence wasn’t “unquestionably favorable.” White III, at 6.
That interpretation of the court’s opinion is reasonable, avoids any conflict with Strickland, and
aligns with our presumption that the state court knew and followed the law. So, that’s the
interpretation we adopt.

       Relatedly, when a state court has already properly recited the Strickland standard, if at all
possible, we should not read language later in its opinion to “supplant[] Strickland” and thus
“needlessly create internal inconsistency in the opinion.” Holland v. Jackson, 

(2004) (per curiam). Here, the Kentucky Supreme Court began its ineffective assistance analysis
by properly reciting the Strickland standard. See White III, at 3. That initial, correct articulation
of the standard provides further reason for us to not strain to read the court’s later
“unquestionably favorable” language as problematic.

       When we analyze the “opinion as a whole” and read its “words in context,” it becomes
even more apparent that the Kentucky Supreme Court’s “unquestionably favorable” phrasing did
not put its decision at odds with Strickland. See Rogers, 

. In the paragraphs
preceding the “unquestionably favorable” sentence, the Kentucky Supreme Court assessed the
prejudice of counsel’s failure to introduce a psychological report at trial.4 That report contained
numerous unfavorable details about White, such as his antisocial tendencies and his lack of
empathy. Thus, the court reasoned that “[a]lthough White maintains [the report] is mitigating
evidence, it can just as easily be considered as detrimental to his case.” White III, at 5. Then,
when discussing extra mitigating testimony about White’s background, the court noted that the
additional evidence would have been “cumulative and perhaps counterproductive.” Id. at 6. The
       4
           White doesn’t challenge the non-introduction of this report here.
 No. 21-5958                            White v. Plappert                                 Page 21


court further concluded that White’s counsel behaved reasonably in deciding “not to present
evidence which conflicted with the mitigation theory.” Id. Given this context, one could
reasonably interpret the court’s subsequent “unquestionably favorable” phrase as characterizing
White’s mitigating evidence, rather than articulating the legal standard for prejudice. Finally, the
“unquestionably favorable” line is the last sentence in the penalty-phase section of the opinion.
Like most conclusions, this sentence merely ties together the factual and legal determinations of
the foregoing analysis; it summarizes why White’s evidence doesn’t show prejudice. Nitpicking
a concluding sentence that’s summarizing otherwise unobjectionable analysis on the basis of a
single adverb would betray a “readiness to attribute error [that] is inconsistent with the
presumption that state courts know and follow the law.” Woodford, 

. So, we must
decline the invitation to nitpick.

        Altogether, these considerations support the conclusion that the Kentucky Supreme
Court’s use of the word “unquestionably” didn’t contradict the Supreme Court’s prejudice
standard. Cf. Sutton v. Bell, 

, 758 n.3 (6th Cir. 2011) (giving effect to an alternative
reading of a state-court opinion when doing so would avoid the conclusion that the court applied
the wrong standard). After all, when denying ineffective-assistance claims, the Supreme Court
frequently emphasizes that the unpresented mitigation evidence was “by no means clearly
mitigating” and “by no means uniformly helpful to the petitioner.” Pinholster, 563 U.S. at 201;
see also Burger v. Kemp, 

 (1987); Penry v. Lynaugh, 

 (1989)
(holding that evidence of a history of abuse is “a two-edged sword”), overruled on other grounds
by Atkins, 

. The Kentucky Supreme Court did the same, and thus didn’t contradict
clearly established Supreme Court precedent.

                                                 *

        In sum, the Kentucky Supreme Court’s decision wasn’t contrary to clearly established
federal law, as determined by the Supreme Court in Strickland.

                                                IV.

        We now turn to whether the Kentucky Supreme Court “unreasonabl[y] appli[ed]” clearly
established Supreme Court precedent to White’s claims. 

(d)(1). White offers
 No. 21-5958                            White v. Plappert                                Page 22


three primary theories of ineffective assistance in the penalty phase. He challenges (A) the depth
and timing of counsel’s mitigation investigation, (B) counsel’s failure to introduce certain
mitigating evidence after learning about it, and (C) counsel’s decision to introduce certain
unfavorable evidence. But a fair-minded judge could conclude that none of these arguments
satisfies Strickland.

                             A. Counsel’s Mitigation Investigation

        For White’s first theory of ineffective assistance, he claims that his counsel’s mitigation
investigation was so inadequate that it fell outside the realm of reasonableness. And he argues
that no fair-minded judge could conclude otherwise.         But White’s attorneys performed a
thorough investigation under the circumstances and then introduced ample mitigating evidence.
Thus, a fair-minded judge could find counsel’s investigation adequate. Moreover, a judge could
reasonably conclude that, even if counsel had presented extra evidence, it would not have
changed the jury’s mind.

                                    1. Deficient Performance

        Under Strickland, counsel has “a duty to make reasonable investigations or to make a
reasonable decision that makes particular investigations unnecessary.” Strickland, 

. A federal court reviewing an ineffective-assistance claim in the AEDPA context should
begin by assessing whether there is on-point Supreme Court precedent governing the attorney
conduct at issue. See, e.g., Davis, 798 F.3d at 473–74. Because the Strickland duty “spawns few
hard-edged rules,” rarely will such precedent exist. Rompilla, 

. Still, the
Supreme Court has set forth various specific guideposts delineating what sort of conduct renders
counsel’s assistance ineffective in the sentencing phase of the capital context. For example,
counsel might be deficient beyond fair-minded dispute if he fails to perform any investigations
whatsoever. Williams, 

. Likewise, counsel acts unreasonably when he fails to
investigate “potentially powerful mitigating evidence” that “stared [him] in the face.” Van Hook,

 (citing Wiggins v. Smith, 

 (2003)). So too if counsel never
attempts to obtain “material that [he] knows the prosecution will probably rely on as evidence of
aggravation at the sentencing phase of trial.” Rompilla, 

. If counsel’s conduct
 No. 21-5958                            White v. Plappert                                Page 23


fails to abide by these specific guideposts laid down by the Supreme Court, then no fair-minded
judge could disagree that counsel has failed to clear Strickland’s low bar, and habeas relief will
be appropriate. See, e.g., Davis, 

.

       But if counsel’s conduct doesn’t breach any of these specific constraints, we next assess
whether the state court unreasonably applied Strickland’s general mandate that counsel’s
behavior must fall within “the wide range of reasonable professional assistance.” 

. Here, AEDPA deference flexes its muscles. Whether a state court’s application of a legal
rule is unreasonable under AEDPA hinges in large part on how permissive the underlying rule is.
Yarborough v. Alvarado, 

 (2004). The more general the rule, the more leeway
the state court enjoys—because the more standard-like the rule is, the more room there is for
distinct yet “reasonable” applications of it. 

 The Strickland standard “is a general standard.”
Knowles, 

; see also Wiggins, 

 (noting that the Supreme Court has
“declined to articulate specific guidelines for appropriate attorney conduct” under Strickland);
Strickland, 466 U.S. at 688–89 (“No particular set of detailed rules for counsel’s conduct can
satisfactorily take account of the variety of circumstances faced by defense counsel or the range
of legitimate decisions regarding how best to represent a criminal defendant.”). So, the state
court enjoys immense discretion here. We will upend its rejection of a petitioner’s ineffective
assistance claim only if there is zero “reasonable argument” available that counsel in fact
satisfied Strickland’s vague standard. Harrington, 

; Davis, 

.

       When we apply this two-step framework here, first we find that the Kentucky Supreme
Court’s decision did not run afoul of any clearly established Supreme Court precedents regarding
deficient performance under Strickland.       Second, we conclude that there are numerous
reasonable arguments that counsel’s performance satisfied Strickland’s general reasonableness
standard.

                                                (a)

       Counsel’s behavior did not fall short of any attorney conduct standards clearly set forth in
applicable Supreme Court precedents. Each precedent is readily distinguishable.
 No. 21-5958                           White v. Plappert                                Page 24


       Consider Williams v. Taylor. There, counsel failed to introduce “voluminous amount[s]”
of mitigating evidence on the defendant-turned-petitioner’s behalf. 

. As a result,
the jury did not hear “any of this evidence.” 

 (citation omitted). But unlike counsel in
Williams, White’s attorneys uncovered and presented myriad details about his “mistreatment,
abuse, and neglect during his early childhood.” 

.

       Nor is Wiggins v. Smith on point. In Wiggins, the defendant was convicted of murder and
sentenced to death. 539 U.S. at 515–16. Counsel made a strategic choice to “introduce[] no
evidence of [the defendant’s] life history”—even though it contained mitigating evidence of
“severe physical and sexual abuse [the defendant] suffered at the hands of” family members. 

The Court concluded that counsel’s decision to not pursue leads after the defendant’s troubled
background came to light and then to not introduce such mitigating evidence was unreasonable.

 at 523–25. The Court also concluded that the state court’s holding to the contrary (namely,
its application of Strickland) was unreasonable. Id. at 527, 534. Not so here: Unlike the
attorneys in Wiggins, once White’s counsel learned about his history of abuse and trauma, they
chased down hours of testimony on the matter. Cf. id. at 525. Moreover, White’s attorneys then
presented that mitigating evidence.

       Finally, Rompilla v. Beard is inapposite. There, the Court held that counsel must “make
reasonable efforts to obtain and review material that counsel knows the prosecution will
probably rely on as evidence of aggravation at the sentencing phase of trial.” 

.
Unlike the petitioner in Rompilla, White doesn’t identify any unexplored mitigating evidence
that should have been obvious to counsel in light of the prosecution’s looming aggravation
arguments. 

, 385–86.

                                               (b)

       Without a controlling, specific Supreme Court holding on point, we turn to whether the
Kentucky Supreme Court somehow unreasonably applied Strickland’s more general standard. A
fair-minded judge could conclude that White’s counsel’s conduct was acceptable when assessed
under Strickland’s general reasonableness standard.
 No. 21-5958                                 White v. Plappert                                        Page 25


        Counsel’s decision to not speak with each and every one of White’s siblings and
immediate family members was reasonable. When our court found AEDPA-grade deficient
performance in Cauthern and Mason, it was because “counsel fail[ed] to investigate a
defendant’s nearest relatives at all.” Cauthern v. Colson, 

 (6th Cir. 2013)
(emphasis added); see Mason v. Mitchell, 

 (6th Cir. 2008) (finding counsel
deficient because he “failed to conduct his own independent investigation and interview
members of [the defendant’s] family regarding the circumstances of his childhood and
background”).      Here, by contrast, White’s counsel decided that they had collected enough
relevant information after speaking with White, his mother, grandmother, stepfather, cousin,
half-brother, three half-sisters, and aunt/believed-half-sister. From these relatives, counsel had
obtained extensive evidence—including specific, detailed accounts—of the abuse that White had
suffered as a child, acts of self-harm, exposure to gruesome violence, struggles in school, and
potential head injuries. Thus, counsel determined that they did not need to interview a handful of
other family members.          And there’s no indication that these other family members were
harboring “potentially powerful mitigating evidence” that counsel should have known about.
Van Hook, 

. Given all this, “it was not unreasonable for his counsel not to identify
and interview every other living family member.” Id.5

        A fair-minded judge could also reasonably conclude that counsel conducted the
interviews of White and his family members in a reasonable manner. While we don’t have a
record of every question that counsel asked outside the courtroom, there’s no reason to believe
their questioning was inadequate. After all, counsel’s efforts turned up scores of troubling
details about White’s conception, childhood, abuse, car accidents, exposure to violence,
substance-abuse issues, self-harm, difficulties in school, and toxic family relationships. One
must strain to imagine that counsel didn’t ask any questions about these topics—which we can’t
do. Cf. Woodford, 

. Thus, White’s contentions that counsel should have asked—
but didn’t—about the circumstances of White’s birth, childhood, possible abuse, suicide attempt,

        5
          White cites 1980 ABA guidelines and caselaw suggesting reasonably competent counsel would have also
investigated White’s employment and educational histories. Yet, when identifying unpresented mitigating evidence,
White mentions school only in passing. And he doesn’t mention his employment history at all. Indeed, White’s
brief doesn’t identify any school- or employment-specific evidence that he believes could have changed the jury’s
mind. Thus, we don’t understand him to be challenging counsel’s failure to investigate these matters.
 No. 21-5958                            White v. Plappert                                Page 26


and head injuries are unavailing. At any rate, if counsel didn’t ask questions about these topics,
that decision was reasonable (and nonprejudicial), since White’s family told counsel this
information anyway.

       To be sure, there’s no indication that White or his family told counsel about his purported
suicide attempt. Maybe counsel didn’t ask about it. Even so, a fair-minded jurist could conclude
that counsel acted reasonably. After all, White concedes that counsel asked his family about
whether he experienced any mental health issues. An attorney could reasonably assume that this
question would elicit details about any suicidal thoughts or attempts. Indeed, White told counsel
how he felt that nobody would care if anything happened to him and how “[he] didn’t think [he]
should be here.” R. 124-14, Pg. ID 4224. Likewise, White’s grandmother and mother told
counsel that White had to see a doctor after engaging in acts of self-harm. Based on these
remarks, reasonable counsel could infer that, if White had attempted suicide, someone would’ve
already spoken up about it. Thus, even if White’s counsel didn’t ask specific questions about
suicide, that decision was not unreasonable.

       Nor did counsel fail to “delve into details” about certain “red flags,” such as White’s
abuse, head injuries, and other traumas. Appellant Br. 38, 49. Counsel spent the better part of
three days offering detailed testimony about White’s exposure to violence at a young age, his
physical abuse, his allegedly untreated psychiatric issues, his substance abuse, his educational
difficulties, and his belief that nobody would care if anything happened to him. To be sure,
counsel could have asked more questions and learned more details.           That’s true of every
investigation. But Strickland calls for effective legal representation, not perfect representation.
See Strickland, 

; Greer v. Mitchell, 

 (6th Cir. 2001) (“The
question . . . is not whether counsel was topnotch, but whether he or she functioned at the level
required by the Sixth Amendment[.]”). And at a certain point, an attorney’s “decision not to
seek more mitigating evidence from the defendant’s background than was already in hand” falls
“well within the range of professionally reasonable judgments.” Van Hook, 558 U.S. at 11–12
(citation omitted) (cleaned up). Counsel passed that point here. Over a three-day period, counsel
asked mitigating witnesses well over one thousand questions.
 No. 21-5958                            White v. Plappert                              Page 27


       Counsel’s decision not to further evaluate White for brain injuries was similarly
reasonable. To be sure, an attorney performs deficiently when he receives medical records of a
defendant’s brain damage but fails to investigate further. See Frazier v. Huffman, 

 (6th Cir.), opinion supplemented on denial of reh’g, 

 (6th Cir. 2003). But that’s
not what happened here. Before White’s trial, a retained psychologist examined White and
found “no gross signs of brain damage.” R. 124-22, Pg. ID 4708. Any brain impairments here
were thus speculative.     And failing to spend additional time and money chasing down
speculative injuries is “far from” being clearly unreasonable. Cf. Hall v. Mays, 

, 450
(6th Cir. 2021). Moreover, “[i]t is ‘not unreasonable’ for counsel, ‘untrained in the field of
mental health,’ to rely on the professional opinions of expert[s].” Haight, 

(citation omitted); see also Morris v. Carpenter, 

 (6th Cir. 2015). That is why
this court has concluded that it was not unreasonable for defense counsel to rely on an expert’s
lack of suggestions that defendant had “organic brain damage” or “required any additional
mental health testing.” Clark v. Mitchell, 

 (6th Cir. 2005). So too has this
court concluded that if a “trained psychologist” did not detect any evidence of a particular
mental defect, counsel’s not detecting the same is not “an objectively unreasonable mistake.”
Campbell v. Coyle, 

 (6th Cir. 2001). In sum, we cannot conclude that White’s
counsel’s reliance on the psychologist’s opinion was unreasonable.

       Finally, White’s counsel did not behave in an objectively unreasonable manner when it
comes to the timing and scope of their investigation into White’s background. Therefore, the
state court’s conclusion that counsel conducted a “reasonable investigation” was not an
unreasonable application of Strickland. White III, at 5.

       First, consider timing. Contrary to White’s contentions here, counsel did not begin
investigating White’s background and upbringing too late.         Counsel investigated White’s
background, for both guilt- and penalty-phase purposes, before Fisher agreed to cooperate. At
trial, one of White’s attorneys explained that he had performed “a rather thorough investigation”
before voir dire. R. 124-11, Pg. ID 3916. Counsel recalled spending approximately a dozen
hours interviewing White’s family, friends, and others who knew him in preparation for the case.
 No. 21-5958                             White v. Plappert                                 Page 28


       Testimony from White’s family supports this. During these interviews, counsel asked
about White’s criminal and behavioral histories and “what kind of kid he [was].” R. 124-25, Pg.
ID 4922. This was consistent with counsel’s initial penalty-phase theory: Had a jury convicted
White, counsel would introduce evidence showing that he had an otherwise well-behaved
childhood without any disciplinary issues. See R. 124-7, Pg. ID 3516 (arguing White “doesn’t
walk around here looking for somebody to kill”). Counsel reasoned that these details, combined
with White’s age, would lead the jury to believe White wasn’t someone who deserved death.

       To be sure, counsel didn’t learn about White’s child abuse or exposure to violence until
Fisher agreed to testify—in part because counsel didn’t initially ask about these topics. That
said, it’s unclear whether some witnesses would’ve shared this evidence before that point. For
instance, White’s mother indicated that “nothing” could have “drag[ged] [testimony about
White’s abuse] out of [her].” R. 124-9, Pg. ID 3646. And one of White’s half-sisters indicated
that she testified only because she found it unfair that Fisher was receiving immunity. In any
event, White’s family wasn’t readily volunteering that information to counsel. As counsel put it
at the time, White’s relatives were “protect[ing] the family secrets” until Fisher took the deal.
R. 124-11, Pg. ID 3919. Thus, unlike in Wiggins, counsel didn’t fail to act despite leads that
were “star[ing] them in the face.” Van Hook, 

.

       Next, counsel’s initial decision to limit investigation into White’s background was not
objectively unreasonable. We must view counsel’s choices in context. Strickland, 

. White’s and his co-defendants’ initial, repeated lying to counsel provides critical context
here. See West v. Bell, 

 (6th Cir. 2008) (finding no deficient performance
when “[petitioner’s] lack of cooperation is part of the reason why his counsel did not discover
some potentially mitigating evidence”); Titlow, 

 (“[A] defendant’s proclamation of
innocence . . . may affect the advice counsel gives.”). Until Fisher agreed to testify for the state,
White and his codefendants pervasively lied to counsel.          For seven months straight, they
consistently denied any involvement in the crime whatsoever. They claimed that they were
elsewhere the night of the murder and even proffered eyewitness testimony to support their alibi.
White’s two codefendants were so committed to misleading their counsel that they turned down
immunity on the basis that they knew nothing about the crime. Meanwhile, the state’s case
 No. 21-5958                            White v. Plappert                                Page 29


looked paper-thin. The prosecution had only circumstantial evidence, and no eyewitnesses could
tie the defendants to the crimes. Indeed, the state was so desperate for evidence that it was
willing to give total immunity to two of the three murderers.

       In these circumstances, a reasonable attorney could conclude that White had a strong
chance of acquittal. Accordingly, it made sense for counsel to spend their time investigating
alibi defenses and poking holes in the state’s circumstantial theory—and less time chasing down
details about White’s childhood. Cf. Greer, 

 (holding that a failure to investigate
mitigating evidence might be unreasonable when a guilty verdict is likely).

       White counters that counsel couldn’t rely on his profession of innocence when deciding
not to further investigate his background. In support, he points to the Supreme Court’s decisions
in Titlow, Porter, Sears, and Andrus, as well as our decisions in Jells, Black, Foust, and Johnson.
But none of those cases helps White.

       Start with Titlow. There, the Court noted that “a defendant’s proclamation of innocence
does not relieve counsel of his normal responsibilities under Strickland.” 

. But an
attorney doesn’t have a “normal responsibility[y] under Strickland” to investigate every possible
trial strategy. 

 Strickland noted that there will be times when an attorney may reasonably
conclude that an in-depth investigation is unnecessary. 

. That’s what happened
here. Counsel investigated the strength of White’s alibi defense and spent a dozen hours talking
to his family about his background. Based on that investigation, counsel concluded that a more
exhaustive inquiry into his upbringing was neither necessary nor a good use of time.           Cf.
Rompilla, 

 (“[C]ounsel may draw a line when they have good reason to think
further investigation would be a waste.”). Further, Strickland instructs us to “apply[] a heavy
measure of deference to counsel’s” decisions to limit their investigations. 

.

       As for Porter, Sears, and Andrus? None of those cases applied AEDPA deference to the
deficient-performance prong of Strickland. See Porter, 

 (reviewing deficient
performance de novo); Sears v. Upton, 

 (2010) (per curiam) (noting certiorari
granted on a state post-conviction case, not AEDPA); Andrus v. Texas, 

 (2020)
(per curiam) (same). And they all post-date White III. So they don’t shed any light on whether
 No. 21-5958                                   White v. Plappert                                          Page 30


the Kentucky Supreme Court unreasonably applied clearly established Supreme Court precedent
at the time.6 See Pinholster, 563 U.S. at 202.

         Circuit precedent doesn’t help White, either. Jells involved counsel who didn’t prepare
any mitigation strategy until after the defendant was convicted. Jells v. Mitchell, 

,
492–94 (6th Cir. 2008). Additionally, Jells relied on a line of cases in which attorneys failed to
investigate in “‘circumstances where a finding of guilty [could not] come as a surprise’” to
counsel. 

 at 494 (quoting Greer, 

). Neither situation is applicable here.
White’s counsel uncovered a load of mitigating evidence before White was convicted. Indeed,
counsel had this evidence on-hand before opening statements at trial.                        Moreover, Fisher’s
agreement to testify—and thus, the newfound high likelihood of White’s conviction—did
surprise counsel. Before that point, White’s attorneys reasonably expected an acquittal.

         White fares no better under Black v. Bell, 

 (6th Cir. 2011). In Black, our
court denied a Strickland claim under AEDPA but noted that “counsel cannot rely solely on
information provided by the defendant and his family in determining the extent of a proper
mitigation investigation.” 

. Regardless of whether that point of law was “clearly
established” in 2002, it doesn’t help White here.                  Counsel didn’t rely solely on White’s
profession of innocence and family when they declined to further investigate his background.
Counsel also relied on the ready availability of alibi witnesses and the prosecutor’s demonstrably
weak case at the time. Thus, Black is also inapposite.

         Nor does Foust help White. See Foust v. Houk, 

 (6th Cir. 2011). There, we
held that counsel’s failure to interview family members or seek records constituted ineffective
assistance. 

 at 536–37. But the Foust defendant admitted to many of the crimes, and counsel
presented no evidence during the guilt phase. 

. Thus, mitigating the death penalty was
counsel’s only strategy. And when counsel’s only strategy is mitigation, no fair-minded judge
could conclude that a failure to investigate mitigating evidence is reasonable. Of course, that

         6
          For this reason, our circuit’s decision in Poindexter v. Mitchell is also inapposite. See 

(6th. Cir 2006) (noting AEDPA didn’t apply because defendant filed his habeas petition before AEDPA’s
enactment). The fact that all these post-White III cases did not apply AEDPA deference is relevant: their analyses
cannot be used to assess whether cases that pre-dated them ran afoul of AEDPA. See Wiggins v. Smith, 

 (2003).
 No. 21-5958                            White v. Plappert                                Page 31


holding has no application here: White’s counsel did have a guilt-phase strategy at the time.
And once that strategy was no longer feasible, counsel began gathering mitigation evidence
immediately.

       Our decision in Johnson is even further afield. There, we held that counsel provided
deficient penalty-phase assistance for three reasons. Johnson v. Bagley, 

 (6th
Cir. 2008). First, counsel did not speak with the defendant’s mother about his childhood. 

Second, like counsel in Wiggins, Johnson’s attorneys ignored records in their possession that
“tipped them off” to powerful mitigating evidence. 

 And third, as in Jells, counsel in Johnson
didn’t prepare a mitigation defense until after the state convicted the defendant. 

. None
of that happened here.

       As a final argument, White cites some of his counsel’s post-conviction testimony as proof
that they unreasonably waited to start investigating. Specifically, White points to counsel’s
remarks that they were acting out of “desperation” and “never quite [caught]” the
defense strategy they were pursuing. Appellant Br. 6 (quoting R. 124-24, Pg. ID 4859); Reply
Br. 15 (quoting R. 124-25, Pg. ID 4941).         Such remarks, however, are beside the point.
Whether counsel performed reasonably is an objective test applied to counsel’s behavior under
the circumstances at the time; counsel’s subjective, post-hoc thoughts about their performance
are irrelevant. Fields v. Jordan, 

 (6th Cir. 2023) (en banc), cert. denied sub
nom. Fields v. Plappert, No. 23-6912, 

 (U.S. June 10, 2024).

                                                *

       Perhaps, in hindsight, counsel made mistakes when allocating their time and structuring
their investigations at first. But it’s not our job to play Monday morning quarterback. See Bell v.
Cone, 

 (2002); Harrington, 

; Strickland, 

. We
just need to make sure that the state court’s conclusion that counsel behaved reasonably under
the circumstances was not itself unreasonable. Here, counsel’s client initially had an alibi
backed up by eyewitnesses, while the prosecution’s case looked weak in comparison. For us to
nonetheless hold—decades later and with the benefit of hindsight—that counsel should have
focused more on developing evidence to mitigate a potential conviction rather than avoiding that
 No. 21-5958                                  White v. Plappert                                        Page 32


conviction in the first place would not only be wrong but perverse. In the name of upholding
Strickland’s guarantee of effective legal representation, we would undercut it. We would risk
incentivizing defense attorneys to spend excessive amounts of their limited time, energy, and
resources preparing for any and every eventuality rather than focusing on proving their client’s
innocence in the first instance—even when there’s initially a very strong innocence defense
available. That’s not what clients want; it’s not what the Sixth Amendment requires; and it’s not
something that AEDPA allows us to do.

        In sum, a reasonable judge could conclude that White’s counsel adequately discharged
their duty to investigate and present mitigating evidence. When “counsel investigate[s] and
present[s] evidence regarding [a petitioner’s] mental health, alcohol and drug abuse, and family
background,” the failure to further investigate and present additional evidence “do[es] not show
constitutionally deficient performance.” Haight, 

. Here, counsel uncovered and
presented substantial evidence of White’s childhood abuse, trauma, substance use, self-harm, and
family strife. And White hasn’t shown that counsel’s initial decision to limit their mitigation
investigation is deficient beyond fair-minded dispute.7                Therefore, White’s first theory of
ineffective assistance fails at the deficient-performance step.

                                                 2. Prejudice

        Even if counsel’s failure to perform a deeper or earlier investigation was somehow
deficient, it didn’t prejudice White.          To prevail on this prong, White must show that all
reasonable judges would find that (a) this evidence was different in kind and degree from the
evidence presented at trial, (b) the jury would have found that the evidence diminished White’s
moral culpability, and (c) there’s a “substantial likelihood” a juror would have concluded that the
mitigating evidence outweighed the aggravating evidence. Shinn, 

; Hill v.
Mitchell, 

 (6th Cir. 2005); Williams, 

; Van Hook, 

. White can’t make these showings.


        7
          Even if counsel behaved unreasonably by waiting to investigate until after Fisher turned, White suffered
no prejudice from that decision. As discussed above, despite counsel’s “delay,” they nonetheless uncovered a great
deal of mitigating evidence. And as discussed below, a court could reasonably conclude that whatever extra
evidence counsel would have uncovered by starting earlier wouldn’t have made a substantial difference.
 No. 21-5958                           White v. Plappert                                Page 33


                                               (a)

       Because counsel presented mitigating evidence at trial, White faces a heightened burden
to show prejudice. He must prove that the additional, unpresented evidence “differ[s] in a
substantial way—in strength and subject matter—from the evidence actually presented.” Hill,

. The evidence must paint “an altogether different picture” of the defendant’s
background. Foust, 

. White’s additional evidence can’t bear this heavy burden.

       At White’s trial, the jury heard three days of harrowing testimony about White’s
upbringing. The left column of Table 1 sets forth a sample of that extensive testimony. The
right column presents the extra evidence that White argues counsel should have uncovered.
Each piece of the unpresented mitigation evidence falls into one of seven buckets: (1) attempted
killings of White in utero and as an infant, (2) episodes of physical, sexual, and verbal child
abuse in his house, (3) exposure to traumatic episodes of dishonesty, violence, and death,
(4) testimony that White and his siblings were unloved or unwanted, (5) evidence of mental
health issues, head injuries, and other physical trauma, (6) evidence of substance abuse, and
(7) underperformance in school.

       The first three buckets of unpresented evidence are “merely cumulative” of the evidence
presented at trial. Broom v. Mitchell, 

 (6th Cir. 2006) (citation omitted). For
instance, the jury already learned that White’s mother attempted to suffocate him as an infant,
causing his body to turn blue. And it heard that White knew about the circumstances of his
conception, and how White’s family hated him and lied to him about it. The jury likewise
learned that White’s father raped his children, that White’s family brutally beat him to the point
of leaving marks, and that, on three separate occasions, White’s family members threatened to
kill him with a gun or knife. The jury also heard testimony on how White witnessed his
grandmother’s shooting, his father’s murder, and his brother’s drowning.           Together, this
evidence was more than sufficient to convince jurors that White was the product of “severe”
abuse and an extremely “violent” upbringing. R. 124-22, Pg. ID 4713.

       To be sure, White’s unpresented mitigating evidence adds new, grizzly details about
specific episodes of abuse and violence. These include stories of how White was forced to drink
 No. 21-5958                            White v. Plappert                                Page 34


alcohol at a young age and suck urine out of his bedsheets. But as White’s own psychologist put
it, these episodes are “just more data to substantiate the same opinion”—that violence and abuse
permeated White’s upbringing. 

 at Pg. ID 4712. And it’s not prejudicial when counsel fails to
present additional episodes painting “an even more detailed picture of the [petitioner’s] abuse.”
Caudill v. Conover, 

 (6th Cir. 2018); see Van Hook, 

 (holding no
prejudice when factfinder “was already aware that [petitioner’s parent] had a violent nature, had
attacked [petitioner’s family members], and had beaten [petitioner] at least once”); Pinholster,
563 U.S. at 200 (holding no prejudice when “[t]he ‘new’ evidence largely duplicated the
mitigation evidence at trial”). Accordingly, White can’t show that unpresented evidence in the
first three buckets is both stronger and substantially different from that presented at trial. The
evidence is thus cumulative, and failure to present cumulative evidence can never establish
prejudice. Broom, 

.

       In response, White argues that the jury only heard general accounts of abuse, not specific
details. To be sure, detailed descriptions of trauma can have more mitigating potential than
general accounts. See Jells, 538 F.3d at 500–01. But that doesn’t help White here: the
mitigating evidence the jury heard at trial was already quite detailed. Contra 

 at 498–501.
Consider an example. When describing the murder of White’s father at the hands of White’s
uncle, multiple witnesses recounted the event in haunting detail. They recounted how White’s
uncle first stabbed White’s father with a knife. White—just five years old at the time—and his
siblings then brought their father to his bed. White’s uncle then returned, snapped a broom
handle in half, and drove the splintered end into White’s father’s neck area. As White’s father
struggled for life, he looked White in the eye and told him to take his money after he died.
White’s uncle then pulled out a gun, placed the gun point-blank between the father’s eyes and
repeatedly fired. The assailant then turned to White, placed the gun against his face, said “I will
kill you all,” and pulled the trigger, dry-firing the gun. R. 124-13, Pg. ID 4030.

       This is just one of the many detailed accounts of childhood trauma that the jury heard.
Many more fill the transcripts of White’s trial. Thus, this is not a case in which the jury heard
only high-level, nondescript allegations of child abuse. Rather, it heard multiple days’ worth of
testimony about White’s abuse, exposure to violence, and toxic upbringing. A reasonable judge
 No. 21-5958                                   White v. Plappert                                          Page 35


could therefore conclude that “[a]dditional evidence on these points would have offered an
insignificant benefit, if any at all.” Wong, 

.

                                                        (b)

         To be sure, White identifies some unpresented mitigating evidence that is different in
both kind and degree from that introduced at trial. Specifically, White points to episodes of head
trauma, car accidents, alcohol consumption, and seizures from his childhood. And he adds that
his family attempted to induce a miscarriage while his mother was pregnant with him. Evidence
of this sort wasn’t introduced at trial. Nevertheless, White fails to show how this evidence would
have diminished his “moral culpability” in the eyes of the jury. Cf. Cauthern, 

;
Penry, 492 U.S. at 319–28. Thus, he hasn’t established prejudice.

         Consider White’s alleged head injuries, seizures, car accidents, and childhood alcohol
consumption.        To show prejudice, he must identify how these incidents diminish his
blameworthiness. For instance, he might show how these mitigators “impaired his ability to
appreciate the wrongfulness of his conduct or to conform his conduct to the law at the time of the
murder.” Shinn, 

 (holding no prejudice when petitioner failed to make such
showing). White doesn’t. He offers no medical documentation or testimony that his head
injuries, seizures, car accidents, or alcohol use caused brain damage.8 Without medical evidence
in the record, counsel’s failure to introduce stories about head injuries is not prejudicial. See
Thornell, 144 S. Ct. at 1312 (finding no prejudice when petitioner “alleged a few additional head
injuries from car accidents and fights, but there [was] no medical documentation to corroborate
any of these injuries” (citation omitted)); Jackson v. Bradshaw, 

. 772 (6th Cir.
2012) (no prejudice when post-conviction proceedings don’t turn up new expert findings);
Haight, 59 F.4th at 840–41 (no prejudice when no medical evidence in record showing brain
damage); Foley v. Parker, 

 (6th Cir. 2007) (similar); cf. Sears, 

.
In addition, “witnesses during the sentencing phase testified that [White] experienced a
devastatingly dysfunctional and abusive childhood. . . . The jury also heard evidence regarding

         8
           White faults trial counsel for not ordering a neurological evaluation. Yet, White still hasn’t gotten one
forty-five years later. And what little we have in the record cuts against any claims of brain damage. See R. 124-22,
Pg. ID 4708 (defense expert testifying White exhibited “no gross signs of brain damage”).
 No. 21-5958                            White v. Plappert                               Page 36


[White’s] substance abuse. While additional evidence that [White] suffered a brain injury would
have provided further support for his mitigation case, [White] has not shown this evidence would
be sufficiently compelling to result in a noncapital sentence when weighing all the evidence
presented at sentencing.” See Haight, 

. The same is true for the effects of alcohol
abuse. Cf. 

 (no prejudice when there was no psychopharmacological evidence on the impact
of alcohol abuse in the record).

       White likewise fails to show how his father and grandmother’s unsuccessful efforts to
induce miscarriage—by beating his mother’s stomach and pushing her down a staircase—
diminish his moral culpability. These incidents occurred before White’s birth, and he points to
no medical evidence that they affected his neonatal development or caused later health issues.
Nor does White allege that he knew about these attempted abortion efforts. So it’s unclear how
they could diminish his moral culpability in the jury’s eyes. See R. 124-22, Pg. ID 4691
(White’s psychologist stating that “obviously, what one does not know will not affect one in a
social interactional sense”).

       In sum, acutely disturbing as the additional details of White’s background may be, a fair-
minded jurist could conclude there’s no reasonable probability the jury would find them
mitigating.

                                                (c)

       Lastly, reasonable jurists could conclude that even if all of White’s extra evidence came
in, there’s no substantial likelihood that the jury would have changed its mind.

       When assessing the prejudicial impact of counsel’s failure to introduce mitigating
evidence, we must “keep in mind the State’s evidence on the other side of the scale.” Caudill,

. Here, that evidence was “overwhelming” and the crimes “barbaric.” Dixon v.
Houk, 

 (6th Cir. 2013); White I, 

. The jury learned that
White planned the murders for months and recruited his younger half-brother and a fifteen-year-
old to help. He targeted a blind, seventy-five-year-old lifelong friend, his seventy-four-year-old
wife, who was weakened and had limited mobility, and his seventy-nine-year-old brother-in-law.
White picked them because he knew they’d be easy to subdue. The jury also learned that White
 No. 21-5958                                 White v. Plappert                                        Page 37


called the shots in the operation and used the youngest of the group as a decoy. Cf. Caudill, 

 (jury considered petitioner’s role as mastermind of crime). Then, using a crowbar,
White literally beat the brains out of three victims as they “desperate[ly] struggle[d] for life.”
See Wong, 

 (internal quotation omitted). He then stole their money. See Van
Hook, 

 (robbery as aggravator). Afterward, he encouraged his accomplices to
destroy evidence and instructed them not to cooperate in the murder investigation. And through
it all, White appreciated the wrongfulness of his actions.

        On the other hand, White’s extra mitigating evidence is attenuated from the murders and
thus has “diminished persuasive value.” Thornell, 144 S. Ct. at 1310.9 None of it “provide[s] a
real link” between the alleged mitigating factors—his abuse, head injuries, exposure to sexual
assault and serious violence, or his attempted suicide—and his crimes. Id. at 1311–12. For
instance, White doesn’t argue that his seizures, alcohol consumption as a young child, or car
accidents predisposed him to murder the Grosses in a particularly gruesome way. Thus, White’s
extra mitigating evidence “would have done him little good.” Id. at 1312. So, “it is not
reasonably likely that this evidence would have resulted in a different sentence.” Id. at 1311.

                                                      (d)

        White’s counterarguments don’t move the needle. His invocations of Porter, Williams,
Rompilla, and Wiggins are inapposite.             The Supreme Court has limited those precedents’
prejudice holdings to only those cases in which “defense counsel introduced little, if any,
mitigating evidence.” Id. at 1314. That’s not what happened here.

        White’s citations to our decisions in Cauthern and Goodwin fare no better.10                        Our
decision in Cauthern, like the Supreme Court cases listed above, involved a trial in which the
jury heard no testimony of abuse from the defendant’s family members. 

.
Likewise, Goodwin concerned a defendant whose jury never heard about his low education

        9
         White argues that the Sixth Circuit held otherwise in Hodge v. Jordan, but that decision predates the
Supreme Court’s decision in Thornell and, in any event, has been vacated. See 

 (6th Cir.), reh’g en
banc granted, opinion vacated sub nom. Hodge v. Plappert, 

 (6th Cir. 2024).
        10
           White also cites Harries v. Bell and Sowell v. Anderson, but neither case applied AEDPA. 

 (6th Cir. 2005); 

 (6th Cir. 2011). So neither informs our analysis here.
 No. 21-5958                                   White v. Plappert                                          Page 38


achievement, unstable home life, and physical abuse. Goodwin v. Johnson, 

(6th Cir. 2011). By contrast, White’s jury heard about all these things at trial.

         White next emphasizes that the jury took a total of nine hours to form a unanimous
consensus on the death penalty, and that the jury was initially split. He claims that this proves
the jury was on the fence about the death penalty, and that additional mitigating evidence would
have pushed at least one juror to change his or her mind. But White’s brief cites to no clearly
established Supreme Court precedent holding that a nine-hour deliberation period or an initial
jury spilt proves prejudice. And while our circuit has pointed to jury deadlock in finding
prejudice, that was in a case where the jury heard no evidence about a defendant’s abusive
household. See Mason, 543 F.3d at 770–80. Again, that’s not what happened here.

         Moreover, the Kentucky Supreme Court could have reasonably interpreted the jury’s
deliberation timeline as evidence of the obvious: sentencing another human being to death in
any case is a grave undertaking that might require prolonged consideration.                          Cf. Gregg v.
Georgia, 

 (1976) (White, J., concurring in the judgment) (“Imposition of the
death penalty is surely an awesome responsibility for any system of justice and those who
participate in it.”).      Thus, we can’t say all fair-minded judges would find that the jury’s
deliberation somehow proves prejudice.

         White also claims that his extra evidence isn’t cumulative of the testimony offered at trial
because he had given some of that testimony himself. And the jury was likely skeptical of his
self-serving testimony. Thus, he claims, his mitigating testimony was weaker than it would have
been had a family member offered it. But that contention only speaks to the evidence’s relative
strength, and that’s not enough to show prejudice. Hill, 

 (strength and subject
matter). Moreover, White points to no Supreme Court decision clearly establishing that a
similarly biased third party’s testimony on the same underlying evidence can establish
prejudice.11 So this argument doesn’t get White anywhere, either.




         11
          The dissent’s invocation of Skipper v. South Carolina, 

 (1986), is misplaced. Dissent Op. at
85 n.8. There, the Court found that it was “implausible” to characterize the testimony of “disinterested witnesses” as
“cumulative.” Skipper, 

. White’s family members aren’t disinterested witnesses. Thus, there’s no
 No. 21-5958                                   White v. Plappert                                         Page 39


        Lastly, White argues that he suffered prejudice because the penalty-phase jury
instructions (1) didn’t expressly tell the jury to consider mitigating evidence from the guilt phase,
and (2) didn’t inform the jury that it could consider mitigating evidence beyond the listed
statutory factors. There are myriad issues with this argument. First, White has forfeited these
arguments by failing to raise them in his opening brief. See United States v. Carson, 

 (6th Cir. 2009). Second, White doesn’t argue that counsel was ineffective for failing to
object to the jury instructions. So, to the extent these instructions caused prejudice, it would be a
function of the trial court’s error, not counsel’s. Third, and relatedly, any challenges to the jury
instructions’ correctness aren’t properly before this court.                   Fourth, White’s argument is
unpersuasive on its own terms. The penalty phase of the trial immediately followed the guilt
phase, and the penalty instructions told the jury (1) it may consider all mitigating circumstances
“as have been presented to you in the evidence,” and (2) that the jury wasn’t limited to the
enumerated mitigators. R. 125-4, Pg. ID 6088, 6096; see also White I, 

 (“This
instruction permits the jury to consider every circumstance in mitigation offered by White.”).
Nothing in the instructions suggests the jury couldn’t consider earlier-introduced evidence as
mitigation.

        Finally, to the extent that White argues the jury instructions magnified the prejudicial
effect of counsel’s failure to present additional mitigating evidence, his arguments fail for the
reasons given above. Even if all the extra evidence had made it in, there’s no reasonable chance
that the jury would have changed its sentence.

                                                         *

        To recap: Much of the unpresented evidence was cumulative of evidence that the jury
had already heard. Other portions would have done nothing to diminish White’s culpability.
And none of the evidence has a nexus with White’s murder. Given all this, a reasonable judge
could conclude that the extra evidence did not have a substantial likelihood of changing the
jury’s death-penalty recommendation. Cf. Wong, 558 U.S. at 27–28 (holding “[i]t is hard to
imagine . . . additional facts about [the petitioner’s] difficult childhood outweighing the facts of

precedent establishing that their testimony could establish prejudice: their testimony is different in kind from the
disinterested testimony the Skipper Court concluded was non-cumulative.
 No. 21-5958                             White v. Plappert                                Page 40


[the] murder,” in which the petitioner struck the victim in the head multiple times with a metal
weapon in furtherance of a burglary (emphasis omitted)). Thus, even if White could show that
counsel’s investigation was deficient beyond fair-minded dispute, a reasonable judge could
conclude that there was no prejudice.

                     B. Failure to Introduce Certain Mitigating Evidence

         In addition to arguing that counsel should have uncovered more evidence, White claims
that his attorneys unreasonably failed to introduce some of the evidence that they did uncover.
But a fair-minded judge could conclude that these decisions were reasonable and didn’t prejudice
White.

         First, White highlights counsel’s decision not to introduce testimony about car accidents
during which he allegedly suffered head injuries. Unlike the unpresented mitigation evidence
discussed earlier, counsel had this information on-hand at the time of trial, but nonetheless
decided to leave it out of the guilt phase.

         A fair-minded judge could find this decision reasonable. As noted above, at the time of
trial, White didn’t exhibit any signs or have any medical evidence of brain damage. And an
attorney’s decision to omit lay testimony about potential brain damage is an “inherently
strategic” choice that doesn’t establish deficient performance when the attorney has already
presented mitigating evidence of the petitioner’s physical abuse. Hall, 7 F.4th at 450.

         Next, White faults counsel for not presenting any mitigating evidence during the penalty
phase of the trial. But that decision was reasonable, too. The penalty phase immediately
followed the guilt phase, during which counsel spent three days putting on witnesses who offered
mitigating evidence. Thus, “counsel reasonably could have concluded that the substance of their
testimony was still fresh to the jury.”       Cone, 535 U.S. at 699–700 (finding no deficient
performance when counsel didn’t recall witnesses in sentencing phase, since counsel offered
mitigating testimony during the guilt phase in support of an insanity defense, and sentencing
immediately proceeded after the trial). And under Kentucky law, “evidence properly admitted
on the first phase of the trial can be and is, without being further introduced in evidence,
considered by the jury in arriving at its decision on the sentencing phase.”              Smith v.
 No. 21-5958                            White v. Plappert                                 Page 41


Commonwealth, 

 (Ky. 1980); see Cone, 535 U.S. at 699–700. Trial counsel
is neither ineffective nor prejudicial when it fails to present cumulative evidence—much less
duplicative evidence. See Haight, 

.

       White also takes issue with the Kentucky Supreme Court’s holding that “[t]he decisions
made by defense counsel not to call [a defense expert], not to recall family members and not to
present evidence which conflicted with the mitigation theory were professionally reasonable.”
White III, at 6. According to White, the court here was saying that counsel’s decision not to
introduce the unpresented mitigating evidence was reasonable, because that evidence conflicted
with White’s penalty-phase defense strategy. Under that reading, the court’s conclusion might
be problematic: if trial counsel never discovered that mitigation evidence, how could they have
made an informed decided not to present it? And even if counsel did know about this evidence,
there’s no reason to think it would have conflicted with White’s mitigation theory. Thus, if
White’s reading is correct, the Kentucky Supreme Court’s logic seems unreasonable.

       There are two problems with this line of thinking. First, to reiterate, we review the
reasonableness of a state court’s decision, not the clarity of its prose. See Rogers, 69 F.4th at
391–92. It’s not our job to dissect specific sentences in state-court opinions. Moreover, given
our deferential posture, we may not readily “attribute error” to the state supreme court’s
decision: when confronted with two reasonable interpretations of the state court opinion—one in
line with Supreme Court precedent and the other not—we must opt for the former.                 See
Woodford, 

.

       And White’s argument doesn’t hold up on the merits.                There’s a much more
straightforward way to interpret this part of the court’s opinion. In discussing the “conflict[ing]”
evidence, the court was likely referring to White’s psychological report, which counsel were
aware of at the time. White III, at 6. Just two paragraphs earlier, the opinion discussed that
report, which concluded White had antisocial tendencies, “explosive features,” self-centeredness,
a lack of empathy for others, and thrill-seeking. Id. at 5. In other opinions, judges have
described that report as saying that White had “minimal regard for the rights and feelings of
others” and a tendency to use “power and intimidation” to get whatever he wanted. White, 

, at *71. Moreover, the report concluded that White wasn’t eligible for an insanity
 No. 21-5958                             White v. Plappert                             Page 42


defense. As the Kentucky Supreme Court put it, that evidence could “easily be considered as
detrimental to his case.” White III, at 5.

       Given this context, we understand the Kentucky Supreme Court’s “conflicting evidence”
line to be a call-back to White’s psychological report—not an assessment of White’s unpresented
mitigation testimony. And in this respect, the court’s application of Strickland is eminently
reasonable: a fair-minded jurist could conclude that a report indicating someone isn’t insane
“conflict[s] with [a] mitigation theory” based on insanity. Id. at 6.

       The Kentucky Supreme Court held that counsel’s decision to omit certain mitigating
evidence from White’s trial was reasonable and nonprejudicial. White hasn’t shown that this
holding was unreasonable, so his second theory of ineffective assistance also fails.

                              C. Introduction of White’s Bad Acts

       As a third theory of ineffective assistance, White argues that counsel shouldn’t have
introduced evidence of certain past bad acts. With one possible exception, a fair-minded judge
could conclude that this was a reasonable trial strategy. In addition, the introduction of this
evidence wasn’t prejudicial. White’s bad-acts evidence was weak, redundant against the state’s
bad-acts evidence, and too insignificant to make a difference to the jury.

                                      1. Bad Acts – Generally

       As part of White’s defense, counsel elicited testimony that White got in fights, stole,
abused animals, and made a sexual advance toward one his half-sisters. While this approach
might initially seem unreasonable, context illustrates how it could logically fit into counsel’s
defense strategy. Recall that counsel needed to prove that White couldn’t conform his behavior
to the law. And since experts concluded that White wasn’t insane, counsel opted for another
route. They argued that, because violence, abuse, and trauma so permeated White’s childhood,
he never developed the cognitive ability to pick right from wrong and, thus, couldn’t be held
criminally liable. As evidence of this, counsel brought in testimony of White’s engaging in
highly problematic behaviors.
 No. 21-5958                             White v. Plappert                                  Page 43


       That’s not an unreasonable strategy. See Crawford v. State, 

 (Miss.
2003) (holding counsel is not ineffective under Strickland for presenting evidence of past bad
acts when they are probative of insanity); see also Coates v. Commonwealth, No. 2005-CA-
001780-MR, 

, at *5 (Ky. Ct. App. Mar. 2, 2007) (“the inferences which the
jury could draw from the [prior bad acts] evidence were as favorable to his defenses of insanity
and mental illness as they were against”); Schrock v. Kirkpatrick, No. 16-cv-6364 CJS, 

, at *13 (W.D.N.Y. Mar. 27, 2019) (“trial counsel’s decision . . . to delve into certain
prior bad acts as part of the insanity defense” was a “valid tactical decision[]”); cf. State v.
Hinchey, 

 (Ariz. 1990) (“[O]nce a defendant raises insanity as a defense,
evidence of prior bad acts . . . may assist the trier of fact in determining criminal
responsibility.”); Yarborough v. Gentry, 

 (2003) (per curiam) (finding that counsel’s
presentation of the defendant’s bad character traits to the jury was not ineffective assistance, as it
was a reasonable trial tactic to “buil[d] credibility with the jury and persuade[] it to focus on the
relevant issues in the case”). That’s especially true here, as counsel had few other viable options
for the guilt phase. The alibi defense had collapsed. And of the possible affirmative defenses,
insanity was likely White’s best bet, even if it were a longshot. “Regardless of whether this
testimony damaged or helped [White’s] case in the eyes of the jury, counsel’s questions were
objectively reasonable.” Hodge v. Haeberlin, 

 (6th Cir. 2009).

       To be sure, proving insanity through past bad acts comes with its risks. After all, a jury
could view this same evidence and conclude that White is such a depraved person that the death
penalty is all-the-more appropriate. See Pinholster, 563 U.S. at 201. But that doesn’t mean
counsel was unreasonable for taking this path. Cf. Sears, 

 (noting that reasonable
counsel can attempt to spin otherwise-adverse evidence into effective mitigation); Pinholster,
563 U.S. at 201 (no ineffective assistance when counsel doesn’t introduce mitigating evidence
that has capacity of backfiring). After all, many kinds of mitigating evidence are a “two-edged
sword.” Penry, 

. And when AEDPA and Strickland apply, we may grant relief
only if no fair-minded judge could conclude that the state court’s conclusion that counsel acted
reasonably under the circumstances was itself reasonable.
 No. 21-5958                           White v. Plappert                               Page 44


       But even if counsel’s decision to introduce White’s past bad acts were unreasonable, it
wouldn’t constitute prejudice under Strickland.     Some of the “bad acts” testimony wasn’t
particularly compelling. For instance, after one witness testified that White “scalded [a dog] to
death,” a subsequent witness clarified the alleged abuse:      White had just made the dog’s
bathwater too hot, and he took the dog out as soon as he realized. R. 124-14, Pg. ID 4186.

       And compared to the brutality of White’s murders, his past bad acts were just drops in the
storm. A reasonable jurist could therefore conclude that a few extra bad acts didn’t have a
substantial impact on the likelihood of a death-penalty recommendation.

                                 2. Bad Acts – Alleged Perjury

       The above analysis comes with one caveat. When introducing “bad acts” about White’s
past, counsel allegedly made one decision that, if true, would be objectively unreasonable. In an
effort to show that White was incapable of controlling his behavior, counsel solicited testimony
from his three half-sisters. Each alleged that White had made sexual advances toward them on
different occasions. Louise testified that White came up behind her while she was sleeping one
night and tried “to do something to [her] back end.” R. 124-13, Pg. ID 4111. Mary Ann
testified that she once woke up to an inebriated White lying on top of her, and she didn’t see
whether or not he was wearing clothes. And Mildred claimed that White tugged up her gown
while she was sleeping one night and intended to perform some sexual act.

       At White’s post-conviction hearing two decades later, Louise maintained that the incident
occurred as she described it at trial. But Mary Ann and Mildred each suggested that they had
exaggerated the incidents on counsel’s advice. Specifically, Mary Ann claimed she knew White
was clothed, and only testified otherwise because counsel wanted her to sexualize the story. And
while Mildred stated that her incident did in fact occur, she made clear White never touched her
sexually.   Based on this testimony, White alleges that counsel performed deficiently by
encouraging Mary Ann and Mildred to lie about past bad acts.

       To be sure, suborning perjury would likely be deficient beyond fair-minded dispute.
“Although counsel must take all reasonable lawful means to attain the objectives of the client,
counsel is precluded from taking steps” towards “presenting false evidence or otherwise
 No. 21-5958                            White v. Plappert                                 Page 45


violating the law.” Nix v. Whiteside, 

 (1986). That said, reasonable jurists
could doubt whether counsel actually encouraged White’s half-sisters to lie. For one, these
allegations arose nearly twenty years after White’s trial, casting substantial doubt on their
accuracy. As the Supreme Court remarked with respect to affidavits filed nearly a decade after
petitioner’s trial, “[n]o satisfactory explanation has been given as to why the [witnesses] waited
until the 11th hour.” Herrera, 

. And it is “reasonable to presume that there is
something suspect” about testimony that arises so late in the game. Taylor, 

.
Moreover, the same witnesses who purported earlier to lie under oath made these accusations. If
they were willing to lie under oath to acquit White in 1980, “the potential that they were also
willing to lie under oath . . . to support White’s ineffective assistance claim must be
acknowledged.” White, 

, at *24 n.53.

       In any case, we need not resolve whether White’s counsel actually suborned perjury.
Even if they did, the error would be harmless. Even if Mary Ann and Mildred had not testified
as they did, the jury still would have heard Louise’s truthful account of White’s sexual advance
on her. Thus, counsel’s alleged efforts to suborn perjury didn’t produce any additional prejudice
beyond counsel’s decision to introduce Louise’s testimony. And for reasons we’ve already
discussed, that decision, while risky, was a reasonable way to show White couldn’t conform his
behavior to the law.

                                 D. Miscellaneous Arguments

       White offers two additional theories of deficient performance. Neither carries him to
habeas relief.

       White argues that counsel should’ve investigated further before picking an alibi-based
theory of defense. According to White, counsel should have anticipated that he was guilty and
that Fisher would agree to testify for the state. And had counsel investigated further, White
argues, they wouldn’t have spent the pretrial period working on a futile alibi defense.

       This argument is flawed for three reasons. First, counsel’s decision to pursue an alibi
defense primarily speaks to adequate assistance during the guilt phase. Defendants offer alibis as
 No. 21-5958                             White v. Plappert                                   Page 46


evidence that they didn’t commit the crime—not to show that they deserve a lesser sentence.
That distinction is fatal here, since counsel’s guilt-phase assistance isn’t before this court.

       Second, a reasonable judge could conclude that counsel’s initial decision to pursue an
alibi defense was reasonable. Before Fisher decided to testify for the state, there was strong
reason to believe a jury would acquit White. The only eyewitnesses put White elsewhere the
night of the crime, and the state’s case wasn’t strong. After all, why else would the prosecutor
be willing to face “fifteen kinds of hell” for giving a murderer full immunity in exchange for
eyewitness testimony? R. 124-18, Pg. ID 4537. Likewise, there was reason for White’s counsel
to believe Fisher wouldn’t agree to testify: while represented by that counsel, Fisher had
previously turned down an immunity offer on the grounds he knew nothing about the murders.
Thus, counsel’s initial decision to pursue an alibi defense was reasonable.

       Third, White fails to “affirmatively prove” that counsel’s decision to pick an alibi defense
prejudiced him. Hodge, 

 (citation omitted). At no point does White identify
what guilt-phase defense a reasonable attorney would have picked. Instead, he simply asserts
that counsel should have focused on a “horrendous childhood mitigation presentation” for the
penalty phase. Appellant Br. 29–30. But that doesn’t tell us what guilt-phase theory reasonable
counsel would have used, or how that theory might have produced an acquittal.

       In sum, White’s critique of counsel’s alibi defense is besides the point here and is
baseless. White also criticizes his counsel’s choice of an insanity defense as unreasonable and
unsupported by evidence or experts. This argument suffers from many of the same problems: it
concerns counsel’s guilt-phase assistance; White doesn’t offer an alternative guilt-phase defense;
and, whatever the merits of counsel’s decision, they nevertheless performed a reasonable
mitigation investigation under the circumstances.

                                           *       *       *

       Federal courts may grant habeas only when both “law and justice require.” Brown, 

 (citing 

). After a horrific life that culminated in his committing
horrific murders, Karu Gene White had his day in court decades ago. A jury of his peers found
 No. 21-5958                                 White v. Plappert                                        Page 47


him guilty of murder. He was sentenced to death. Despite years of delay,12 the truth persists:
neither the law nor justice stand in the way of his sentence.

        We accordingly AFFRIM the district court’s denial of habeas relief.




        12
            On at least fifty-four occasions, White has filed motions to extend deadlines, exceed brief-length
maximums, stay mandates, toll limitations periods, hold cases in abeyance, add addenda to those motions,
reconsider the denials of such motions, and object to the denial of such motions: May 31, 1980; July 31, 1980;
September 26, 1980; December 1, 1980; January 27, 1981; February 26, 1981; March 9, 1981; August 10, 1981;
October 9, 1981; December 8, 1981; February 8, 1982; February 22, 1982; March 8, 1982; April 2, 1982; May 7,
1982; June 7, 1982; July 7, 1982; August 6, 1982; September 7, 1982; February 15, 1983; March 17, 1983; March
31, 1983; April 13, 1983; January 11, 1983; January 31, 1984; February 3, 1984; March 9, 1984; July 6, 1984; June
24, 1994; August 17, 1994; November 15, 1994; February 13, 1995; March 24, 1995; April 21, 1995; September 22,
1995; July 11, 2000; December 29, 2000; February 13, 2001; February 15, 2001; April 26, 2001; July 5, 2001; July
17, 2001; September 4, 2001; June 3, 2002; October 21, 2002; October 16, 2003; November 4, 2003; July 8, 2015;
October 13, 2021; June 29, 2022; August 29, 2022; January 19, 2023; April 27, 2023; February 14, 2024. White has
been briefing the same ineffective-assistance claims for twenty-five years. He has had more than enough time and
brief space to develop these issues.
No. 21-5958                            White v. Plappert                             Page 49


              held it up against White’s throat, and said he   •   White’s grandmother
              would be the first one to die. White knew
                                                                      o Would whip White
              that his father intended to kill him and
                                                                        until    he    cried,
              frequently remembered this incident.
                                                                        sometimes drawing
              White’s father went to jail for this behavior
                                                                        blood and leaving
              but was released the next day.
                                                                        welts.
  •   White’s mother                                                  o Whipped one of
         o Would beat White with the buckle-end of a                    White’s         half-
           belt, butcher’s knife, broom handle, and                     brothers so hard he
           vacuum cleaner. This occurred nearly every                   couldn’t
           time White was with his mother, and it left                  comfortably sit in
           welts and “stripes” on White. R. 124-12, Pg.                 school the next day.
           ID 3991, 3993.                                             o Would sometimes
         o Would spank him until he fell over.                          punch holes in the
                                                                        walls when she got
         o Punished him harder than her other kids, not
                                                                        angry.
           because he deserved it, but because she
           hated him.                                                 o Made White suck
                                                                        urine out of his
         o Took her hatred of her stepfather out on
                                                                        bedsheets when he
           White and made him “suffer” for being his
                                                                        wet the bed.
           son. R. 124-12, Pg. ID 3963.
         o Locked White in his bedroom every night as          •   White’s mother
           he screamed and beat fists against the wall.               o Drank alcohol to
         o Would break out in fits and throw dishes                     induce miscarriage
           around the house.                                            of          previous
  •   White believed his mother wanted to “beat [him] to                pregnancy.
      death.” R. 124-14, Pg. ID 4220.                                 o Beat White so badly
  •   White’s family worried his stepfather would kill                  it caused scars.
      him after he got in a fight with his half-sister.        •   White’s stepbrother also
  •   White’s grandmother’s boyfriend kicked White out             sexually assaulted his half-
      of house while he wasn’t wearing pants, put a pistol         sisters.
      to his eye, and told him not to speak or he’d never      •   Family members “shot
      speak again. This man similarly kicked White’s               guns” at White and
      other siblings out of the house.                             threatened to kill him.
  •   As a child, White wore two pairs of pants because
      he would be paddled so badly.
  •   One of White’s half-sisters said she didn’t know if a
      child was ever punished more than White was.
 No. 21-5958                             White v. Plappert                              Page 54


                                        _________________

DISSENT

                                        _________________

       JANE B. STRANCH, Circuit Judge, dissenting. White’s appeal addresses only the
penalty phase of his trial and the resulting sentence. On that issue, the majority’s decision is
wrong. The Kentucky Supreme Court’s decision was both contrary to and an unreasonable
application of clearly established federal law.         White has demonstrated both deficient
performance and prejudice from his trial counsel’s representation at the penalty phase.
Accordingly, I would reverse the district court’s determination, grant White relief, and remand to
the district court for further proceedings.

                                        I. BACKGROUND

       A. Factual background

       The facts of White’s crime are truly disturbing. Prior to the crime, Karu Gene White,
then 19-years-old, learned that Charles Gross, Lula Gross, and Sam Chaney, three co-owners of a
small grocery store in their seventies, were hiding $60,000 to $70,000 in their store, which also
served as their home. See White v. Commonwealth, 

 (Ky. 1983). Over the
coming months, White repeatedly discussed his desire to rob the Grosses and Chaney with his
half-brother, Thomas (“Tommy”) Bowling, their friend, Charles (“Chuck”) Fisher, and others.
In early 1979, the three co-owners were murdered and robbed in their store, which was
burglarized. White, 

. The murder was brutal and bloody. In September 1979,
White, Tommy Bowling, and Chuck Fisher were indicted for the murders. White was tried,
convicted, and sentenced to death. His appeal challenges only his sentencing, not his conviction.
 No. 21-5958                                  White v. Plappert                                         Page 55


        B. Procedural Background

                 1. State Trial and Sentencing

        Kevin Charters1 and James Early jointly represented White, Fisher, and Bowling.
See White, 

. The defendants were tried separately; White’s trial began in
February 1980. Up until the beginning of trial, White’s counsel intended to pursue an alibi
defense, and all of their investigation focused on that defense. This was because White and his
co-defendants all protested their innocence and were able to provide a witness who believed she
had seen them at a college dance around the time of the crime behaving normally. During
postconviction proceedings, when asked about White’s case being a death penalty case, Charters
testified that, “[i]t turned out to be”; when White’s postconviction counsel pressed that the
Commonwealth had announced “they were seeking the death penalty,” Charters responded that
the Commonwealth did so “[i]n a half-hearted way; they couldn’t have even gotten a
conviction.” Charters believed “that the chance of a death penalty in my mind was zero, and the
chance of a conviction was one percent,” and emphasized “the evidence . . . that this type of
conduct was totally, utterly out of character from [the defendants] because of their history.”

        In keeping with this belief, Charters and Early conducted little, if any, investigation into
death penalty mitigation. Questioned about a contingency plan in the event of conviction,
Charters testified that “the contingency plan was to . . . present what I thought was a very strong
defense.” Asked who he questioned regarding White’s background, Charters testified that he
“spoke with Kitty Neace,” White’s grandmother, at length. But when asked what types of
questions he asked her, other than whether White “would have had the capacity to commit this
kind of violence,” Charters stated, “[w]ell, I don’t know.” Charters did not recall any discussion
with White’s family members about them taking the stand at the penalty phase. Early admitted
that he did not investigate White’s background at all at that stage.

        This lack of investigation occurred despite mounting evidence that their client might,
indeed, be guilty. Charters acknowledged that he and Early were aware that White had access to

        1
          Charters was disbarred in 1989 for “[n]eglect to practice.” At the time of the postconviction evidentiary
hearing, Charters no longer practiced law but, instead, ran a liquor store.
 No. 21-5958                            White v. Plappert                                Page 56


“unexplained money after the crime” and had purchased “a used automobile with some cash not
long after the crime.” They may have known that witnesses had observed White “go off into a
remote location and come back with a coffee tin of change like what was missing from the scene
of the crime.”

       Counsel’s assumption that they would be able to proceed with an alibi defense came back
to bite them. On February 15, several days into voir dire, Chuck Fisher, one of White’s co-
defendants, agreed to testify on behalf of the Commonwealth in exchange for immunity. At that
point, Charters and Early requested that the court grant a continuance to address this change in
circumstances, urging that they had proceeded on an innocence theory based on their clients’
representations but that Fisher’s cooperation with the Commonwealth necessitated a change in
strategy to an insanity defense.     Defense counsel argued they had “been snowed by the
defendants,” and though they had suspected all along that the culprits were “crazed,” only after
Fisher agreed to cooperate with the Commonwealth did they suspect that “our clients did it and
they were indeed crazed.”

       The court denied the continuance motion. In relevant part, the court responded:

       We have been here five days and it has cost the state of Kentucky over five
       thousand dollars. And here you boys come up because you found there was a
       witness against you that you didn’t anticipate, maybe, I don’t know why you
       didn’t because practically everybody knew it except you, and I tried to tell you a
       long time ago what you was [sic] into, that you couldn’t represent these people.

       When counsel responded that they “didn’t understand why [the court] said that,” the
judge said that he “knew this boy was going to talk. He told the jailer and several people told
me. I don’t know why you boys didn’t know it.” The judge went on to observe that a county
jailer had informed him that “Mr. Fisher was passing the word out to different people that he
wasn’t going to go to the penitentiary in order to protect you fellows.” Charters stated, “surely
we can’t base our defense on the basis of rumors,” to which the court responded, “I think . . . if
the rumor was out, it was your obligation to check it.”

       Early pressed that defense counsel now intended to rely on an insanity defense for White
and explained that he had not told the court sooner because “until late yesterday and early this
 No. 21-5958                            White v. Plappert                                Page 57


morning defense attorneys were inadvised [sic] and unknowledgeable and unaware of previous
crazy behavior of the defendant.” The court stated that the defense attorneys had, until that
point, “intimated all the way through that the defendant’s not going to testify,” to which Charters
responded that defense counsel did not know that White’s mental health raised “a serious
question” until “last night.” The court stated that everyone had “been on the alert since this case
began, but nobody ever told me that [White] was crazy until this morning.”               After the
Commonwealth “concede[d]” that White was “entitled to a hearing,” however, the court agreed
to grant a hearing on sanity and allow for White’s examination. When White’s counsel asked for
time to prepare, the Commonwealth’s attorney responded, “[y]ou have had better than five
months” to prepare a defense. The court scheduled the hearing for the following Monday
morning.

       Once it became clear that they were going to need to pursue an insanity defense, Charters
and Early conducted a limited investigation into White’s background and family history. The
night counsel found out Fisher would testify for the State, they gathered the family members who
were at the courthouse and had what Charters describes as a “pretty emotional discussion” with
them. Early began asking family members “about prior conduct that was bizarre.” Early had one
conversation of several hours with White and then asked family members to corroborate what
White had said. Early never visited the homes of additional witness or sought out new witnesses
besides the ones he was able to speak to at the courthouse that night. Several of White’s
immediate family members were never contacted including several of White’s half siblings who
would have been willing to testify if asked.        Trial counsel never attempted to subpoena
documentation regarding White’s childhood, education, or mental health or the mental health of
his family.

       For the few family members trial counsel actually spoke to, counsel’s inquiry focused on
evidence of insanity and odd sexual behavior. For example, counsel told White’s mother and
grandmother that they needed to talk about anything that would “prove [White] is insane.” They
did not focus on mitigation evidence or evidence of childhood trauma unrelated to insanity.

       The court appointed Dr. Robert Granacher to evaluate White and assess his competency.
Dr. Granacher evaluated White and found him competent to stand trial but recommended “an
 No. 21-5958                                    White v. Plappert                                          Page 58


evaluation by Forensic Psychiatry in Louisville to rule on his sanity” based on “the sensitivity of
this case.” Defense counsel also retained Dr. Paul Evensen to evaluate White, and Dr. Evensen
administered the Minnesota Multiphasic Personality Inventory (MMPI) on White, delivering a
report to Early three days later with findings undermining any insanity defense. This report
contained some indications that Dr. Evensen might be able to present mitigation evidence. It
described White’s erratic behavior and lack of conflict resolution skills as “consequential to” his
childhood experiences. It also asserted that the murder was likely caused by White snapping
because “murder is an ego alien behavior for which he can feel true remorse.” White’s counsel
never followed up with Dr. Evensen on these findings.

         On February 18, just three days after White’s trial counsel decided to shift defenses, the
court conducted a competency hearing. During the hearing, which the jury did not attend,
defense counsel presented the limited evidence they had collected during their rushed
investigation. Trial counsel adduced testimony from some of White’s family members, including
White’s grandmother, Kitty Neace; White’s mother, Nancy Bowling;2 White’s stepfather, Henry
Bowling; and several of White’s half-siblings.3

         White’s mother and grandmother testified that White was conceived as a result of the
rape of his mother (Nancy Bowling) by her stepfather (Karu Neace); that White had been raised
by Karu Neace; that White’s two-year-old step-brother drowned in White’s presence while
White was a small child; that White witnessed his father (Karu Neace)’s murder at the hands of
an uncle; that there had been several other murders in the family; that one of White’s
grandmother (Kitty Neace)’s romantic partners (Harold Sallee) had threatened White with a gun;
that White’s grandmother had occasionally whipped him; that White wet the bed until he was
fifteen; and that White drank and used drugs as a teenager.




         2
          White’s mother is also referred to as Nancy Moore in some parts of the record due to a later marriage. For
the sake of clarity, I will use “Nancy Bowling,” her name at the time of trial.
         3
          As will become clear throughout this opinion, because of multiple interfamilial rapes, most family
members bear multiple relationships with White (e.g., aunts who are also half-sisters). For the sake of clarity, I will
specify the closest relationships with White for each family member.
 No. 21-5958                             White v. Plappert                               Page 59


       Several witnesses testified about White’s trouble learning in school, issues with memory,
and mental difficulties. Several also testified about head injuries White suffered, including a car
accident where White flew through the windshield while driving around 120 miles-per-hour.

       Counsel also encouraged family members to present testimony that made White seem
dangerous and unstable. Family members testified that White hurt and sexually abused animals,
sexually abused himself to the point that he required attention from a doctor and beat and
attempted to sexually assault his half-sisters. Led by White’s attorney, one of White’s half-
sisters testified that she had been subjected to an attempted sexual attack by White when White
was 16 and she was 10; another half-sister described an additional incident of attempted assault.
White’s grandmother described White threatening to murder his uncles as young as age eight.
His mother described him as a compulsive stealer.            White was described as violent and
frightening.

       Some of this testimony may not have been accurate. Several family members later
revealed that, in his desperation to prove that White was insane, Charters had encouraged the
family to exaggerate, if not fabricate, incidents demonstrating White’s proclivity for bizarre
behavior. Others later cast doubt on trial testimony that White abused animals, testifying that
they never observed such abuse, and that White liked animals. Two of White’s half-sisters later
explained that White had not, in fact, attempted to abuse them and that they had only said he did
because White’s attorneys told the half-sisters that they needed to testify that way in order to
help White. When asked whether she testified that White was “a compulsive stealer” because
“[t]hat is what [White’s attorneys] wanted the jury to hear,” White’s mother (Nancy Bowling)
answered, “Yes.”

       After hearing the testimony, the trial court found that all of the witnesses “testified
without exception that [White] was mentally competent and that he had never had any mental
trouble,” and concluded that there was “no evidence, no actual, valid evidence that this man was
insane either [at the time of the crime] or at the present time.”

       Trial recommenced within the week. At trial, White’s counsel presented an insanity
defense to the jury.     Despite relying on an insanity defense, counsel presented no expert
 No. 21-5958                           White v. Plappert                               Page 60


testimony, instead relying on White’s own testimony and the testimony of seven of White’s
family members. The family testified to largely the same things they had testified to in the
competency hearing, although they did not discuss the head injuries at trial. Several were on the
stand for bare minutes.    One of White’s half-sisters, for instance, was only asked twelve
questions. The jury returned a guilty verdict on March 11.

       The penalty phase began on March 12, just one day after the close of the guilt phase. At
the penalty phase, Charters waived the defense’s opening statement. The court questioned him
about this course, stating, “You don’t want to make an opening statement. Commonwealth is
making an opening statement”; Charters confirmed that the defense waived it. White’s counsel
presented no mitigation evidence at the penalty phase, opting to simply address the jury in
closing argument. In Charter’s closing argument, he pointed to White’s lack of criminal history
as a mitigating factor, and emphasized the “terrible manner” of White’s upbringing relying on
evidence from the guilt phase. Stating, “execution is overkill,” Charters implored the jury to
“give the kid one last shot.” The jury began penalty deliberations around 2:15 PM on March 12,
1980, and continued until approximately 8:25 PM, at which point, they were split five-seven.
Court resumed the next day at 9:15 AM, and the jury returned with a death sentence
recommendation at 12:05 PM. Consistent with that recommendation, the court imposed a death
sentence on March 29, 1980.

               2. State Postconviction Proceedings

       The Kentucky Supreme Court affirmed White’s convictions and sentence on direct
appeal in December 1983, White, 

, and the United States Supreme Court
denied his petition for certiorari, White v. Kentucky, 

 (1984). In December 1984,
White collaterally attacked his sentence by filing a motion under Rule 11.42 of the Kentucky
Rules of Criminal Procedure. In May 1999, almost 15 years after White filed his motion, the
trial court conducted a five-day evidentiary hearing pursuant to Rule 11.42.

       During that evidentiary hearing, White’s postconviction counsel, Kevin McNally,
Margaret O’Donnell, and Vincent Aprille, adduced new mitigating evidence, not presented at
trial. Counsel presented new evidence from the same witnesses who had testified at trial along
 No. 21-5958                             White v. Plappert                                  Page 61


with evidence from witnesses trial counsel had never spoken with. There was testimony that both
Karu Neace, White’s biological father, and Aussie White, White’s uncle, sexually abused
children in the home—White’s half-siblings and aunts (who were raised as step-siblings)—
beginning when they were young children. Several of the children were impregnated as a result
of this abuse. This sometimes happened in the room where White was sleeping. White also
witnessed one of his half-sisters being raped by another uncle.

       One of White’s half-siblings, testified that, following a bad car accident, White’s father
(Karu Neace) got drunk, lined the children (including White) up, and threatened to kill them with
a pistol saying, “Well, I’ll just line you all up and start with the oldest and go to the youngest and
kill you all.” White was also routinely beaten by his mother (Nancy Bowling) and grandmother
(Kitty Neace). His grandmother would fly into rages and beat White so severely that he bled
from his legs. When White wet the bed, he was required to suck the urine out of his sheets with
his mouth.

       Alcohol use was rampant in the home.             One of White’s half-brother’s, provided
additional details about the drowning of White’s younger half-brother, testifying that the night
before his two-year-old brother drowned, the adults “gave [the two-year-old] beer to drink,” and
“the next day he was sick from it.” The child was “suffering from a hangover” when he
drowned. White witnessed the drowning. Multiple witnesses testified during the postconviction
evidentiary hearing that other children in the family, including White, were given alcohol at a
young age because the adults in his life found it amusing to watch the children get drunk. The
family also allowed the children to smoke as young as eight years old.

       Witness testimony also raised concerns about White’s head injuries. White’s father
(Karu Neace) and grandmother (Kitty Neace) beat his mother (Nancy Bowling) and pushed her
down the stairs in an attempt to abort her pregnancy with White. White’s grandmother testified
that White’s mother knocked White’s head on a door handle when he was an infant, an injury
White’s grandmother described as serious, but for which, she testified, no one took White to the
doctor. Several family members testified that White experienced seizures from around age six or
seven into his teenage years. White’s half-brother and codefendant, Tommy Bowling, testified
that he observed White unconscious “probably at least eight, nine, ten times.” Another of
 No. 21-5958                             White v. Plappert                              Page 62


White’s half-brothers recounted seeing White knocked out for fifteen minutes after falling
facedown at his grandmother’s house at 12. At age 17 or 18, White got hit on the head with a
“tire buddy” during a fight and was knocked unconscious for two or three minutes. Other
witnesses recounted White’s multiple—and serious—motorcycle and automobile accidents.
White attempted suicide at least once.

       In 1990, White’s postconviction counsel also deposed Dr. Granacher—the doctor who
had originally advised that White was competent to stand trial. When asked whether trial
counsel had given him “any additional information about Mr. White’s medical, mental,
emotional, family or social history” beyond what Dr. Granacher was able to glean from his own
interview with White, Dr. Granacher testified, “I don’t believe he did.” Dr. Granacher also
stated that, if defense counsel had apprised him of White’s history of seizures, he would have
conducted a CAT scan and neurological exam because “[a]nybody that has a history of seizure
disorder . . . even if it weren’t exacerbated by drugs, you’d still need to do an organic
investigation of the patient.”   Based on postconviction counsel’s representations regarding
perinatal trauma, birth trauma, early alcohol exposure “at age 2 ½,” and early childhood trauma,
Dr. Granacher observed that there were “many, many instances” in White’s life “that may have
caused organic brain dysfunction.” Dr. Granacher testified that the new information he received
from White’s postconviction counsel would not have changed his ultimate recommendation to
the court that White was competent to stand trial but would have prompted a more insistent
conclusion that White needed further evaluation, including of his organic brain function.

       During a 1999 deposition, Dr. Evensen testified that trial counsel gave him an incomplete
picture of the abuse White had suffered. Counsel told Dr. Evensen that White was the product of
incest, that White had been abused by relatives and treated “almost as a non-being” by his
mother (Nancy Bowling), and that White had witnessed the shooting of his biological father
(Karu Neace). Dr. Evensen testified that, even with that information, he concluded that, while
White was criminally responsible, his upbringing would have led him to learn violence as a way
of resolving issues and would have inhibited his development of empathy. Dr. Evensen believed
that White’s upbringing had directly caused his antisocial traits and that White’s substance abuse
had reduced his ability to control those traits. Dr. Evensen further believed, based on White’s
 No. 21-5958                             White v. Plappert                               Page 63


psychological profile, that the murder, while deliberate, was the result of minimally controlled
hostility taught by his family and an intoxicated condition that led him to snap the moment he
felt threatened with discovery rather than a cold, premeditated, and remorseless murder. Dr.
Evensen believed White felt regret for the murder and that White had potential for rehabilitation
in a maximum-security prison. But White’s counsel never met with Dr. Evensen to go over his
conclusions for possible mitigation once he concluded that White was criminally responsible.
Dr. Evensen provided an affidavit, stating that, though he could not testify during the guilt phase
that White was not criminally culpable, he would have been willing offer mitigation testimony
during the penalty phase.

       Dr. Evensen also explained that, after White’s trial, postconviction counsel had provided
Dr. Evensen with the additional information about White’s family history that postconviction
counsel had adduced, including evidence of mental illness and homicidal tendencies in White’s
family, the head trauma White had suffered, the seizures White experienced, White’s learning
difficulties, and the extent of the “lethal” abuse in White’s family. Dr. Evensen emphasized that
this was, to him, “new information,” which would have allowed him to testify that White’s
participation in the murders represented a “resort to violence under stress . . . that he learned
from his family.”

       Early and Charters also testified during postconviction proceedings. When asked, Early
did not know whether Charters had actually tried a death penalty case prior to White’s. When
asked whether Early had “ever presented an insanity defense in a murder trial without an expert
witness,” Early responded, “[n]ot until this case”; when asked if he ever proceeded without an
expert in cases after White’s, Early responded, “[n]o way.”

       Charters described his reasoning for proceeding as he had, testifying that he had no
inkling the defendants were not innocent until a jailer told him that Fisher needed to speak with
him, and intended to cooperate with the Commonwealth. Charters emphasized that he had only
“eleven days” to prepare after pivoting to the insanity defense, and that several of those days
were spent in the courtroom for trial.
 No. 21-5958                            White v. Plappert                                Page 64


       In January 2000, a state trial court denied White’s petition for relief, and denied White’s
motion for reconsideration. In May 2002, the Kentucky Supreme Court concluded White had
failed to demonstrate that his counsel’s performance was deficient or that it prejudiced him and
affirmed the trial court’s denial of relief. White v. Kentucky, No. 94-SC-326-MR, at 5 (Ky. May
16, 2002) [hereinafter “White III”].

       C. AEDPA

       The Antiterrorism and Effective Death Penalty Act of 1996 governs appeals, like
White’s, which challenge, in federal court, a state death sentence.        See 

(outlining initial habeas proceedings); 

 § 2253 (detailing appellate procedures for habeas
proceedings). Prior to AEDPA, the Supreme Court recognized habeas corpus, “the Great Writ,”
as an established feature of Anglo-American jurisprudence, one “[r]eceived into our own law in
the colonial period, given explicit recognition in the Federal Constitution,” and “incorporated in
the first grant of federal court jurisdiction.”   Fay v. Noia, 

 (1963).
Recognizing “that today habeas corpus in the federal courts provides a mode for the redress of
denials of due process of law,” the Court emphasized that habeas corpus’s “root principle is that
in a civilized society, government must always be accountable to the judiciary for a man’s
imprisonment” such that “if the imprisonment cannot be shown to conform with the fundamental
requirements of law, the individual is entitled to his immediate release.” 

.

       AEDPA changed things. “[D]rafted, enacted, and signed in an atmosphere of anger and
fear”—the aftermath of the 1995 Oklahoma City bombing—AEDPA’s proponents claimed the
measure was “necessary to fight domestic terrorism.” Bryan A. Stevenson, The Politics of Fear
and Death: Successive Problems in Capital Federal Habeas Corpus Cases, 

, 701 (2002). In practice, AEDPA “introduced entirely new restrictions on habeas corpus”
including “new restrictions in Section 2254(e) on reviewing state court fact-findings,” and
“limited relitigation opportunities for cases that had been adjudicated on the merits in state
court.” Brandon L. Garrett & Kaitlin Phillips, AEDPA Repeal, 

, 1756
(2022). AEDPA explicitly prohibits a federal habeas court from granting relief on “any claim
that was adjudicated on the merits in State court proceedings unless” the state court decision
“was contrary to, or involved an unreasonable application of, clearly established Federal law, as
 No. 21-5958                           White v. Plappert                                Page 65


determined by the Supreme Court of the United States,” 

(d)(1), or the state
court’s decision “was based on an unreasonable determination of the facts in light of the
evidence presented in the State court proceeding,” 

 § 2254(d)(2). These restrictions on relief
are commonly known as “AEDPA deference.”

       Yet “deference does not imply abandonment or abdication of judicial review,” and it
“does not by definition preclude relief.” Miller-El v. Cockrell, 

 (2003). A
state court decision is contrary to clearly established federal law when it applies a rule that
contradicts the governing law set forth by the Supreme Court or where it confronts a set of facts
that are materially indistinguishable from those underlying a decision of the Supreme Court and
nevertheless arrives at a different conclusion. Carter v. Mitchell, 

 (6th Cir.
2016). An unreasonable application occurs when a state court identifies the correct governing
legal principle from the Supreme Court’s decisions but unreasonably applies that principle to the
facts. Rompilla v. Beard, 

 (2005) (quoting Wiggins v. Smith, 

 (2003)).

       “[T]he lack of a Supreme Court decision on nearly identical facts does not, by itself,
mean that there is no clearly established federal law,” however, “since ‘a general standard’ from
[the Supreme] Court’s cases can supply such law.” Marshall v. Rodgers, 

 (2013)
(quoting Yarborough v. Alvarado, 

 (2004)). AEDPA does not prohibit a
federal court from finding an application of a principle unreasonable when it involves a set of
facts “different from those of the case in which the principle was announced.” Panetti v.
Quarterman, 

 (2007) (quoting Lockyer v. Andrade, 

 (2003)). It
can be an unreasonable application of federal law to decline to extend the Supreme Court’s
precedents to factual scenarios where they should clearly apply. Ramdass v. Angelone, 

 (2000).

       In identifying the State court’s reasoning, we must consider the opinion as a whole to
determine whether the rule applied tracks the standards prescribed by the Supreme Court. Id. at
392. Where the state court overlooks a counterargument, or simply fails to provide reasoning on
a particular point, we may not assume from absence that the state court acted unreasonably or
acted contrary to the law. Harrington v. Richter, 

 (2011); Davis v. Carpenter,
 No. 21-5958                            White v. Plappert                                 Page 66

 (6th Cir. 2015). In such cases, we must ask what arguments could have
supported the state court’s decision and grant relief only if no fair-minded jurist could agree with
those arguments.      Harrington, 

.       Where, however, the state court has
affirmatively applied an incorrect standard—even where the incorrect reasoning is exemplified
by a single sentence making clear the state court’s misapprehension of the law—the federal
courts may step in. See Williams v. Taylor, 

 (2000) (finding that the Virginia
Supreme Court’s decision was both contrary to and an unreasonable application of established
law where a single sentence revealed that the state court had added an additional requirement to
the Supreme Court’s rule in Strickland).

                                           II. ANALYSIS

       Ineffective assistance under Strickland v. Washington, 

 (1984) “is deficient
performance by counsel resulting in prejudice, with performance being measured against an
objective standard of reasonableness under prevailing professional norms.” Rompilla, 

 (internal citations and quotation marks omitted). “[C]ounsel’s duty to investigate,” the
contours of which a court can clarify by “referring to the ABA Standards for Criminal Justice as
guides,” derives from this “‘clearly established’ precedent of Strickland.” Wiggins, 

 (citing Strickland, 

). Applying this review to White’s claim, along with
the AEDPA deference described above, I begin by analyzing counsel’s performance, and then
proceed to the prejudice prong.

       A. Performance

       The Supreme Court has long emphasized “that the Sixth Amendment right to counsel
exists, and is needed, in order to protect the fundamental right to a fair trial.” Strickland, 

. “An accused is entitled to be assisted by an attorney, whether retained or appointed,
who plays the role necessary to ensure that the trial is fair.” 

. “The proper measure”
for attorney performance under this standard is “reasonableness under prevailing professional
norms.” 

.

       As acknowledged in a recent published decision of our en banc court, “the U.S. Supreme
Court has repeatedly faulted trial lawyers who failed to adequately investigate for the penalty
 No. 21-5958                            White v. Plappert                                  Page 67


phase and thereby overlooked substantial mitigating evidence.” Fields v. Jordan, 

 (6th Cir. 2023) (en banc) (emphasis omitted) (collecting cases). This inquiry centers “on
whether the investigation supporting counsel’s decision not to introduce mitigating evidence of
[the defendant’s] background was itself reasonable.”        Wiggins, 

 (emphasis
omitted). As of 1980, defense counsel was on notice that their professional duties included the
“obligation to conduct a thorough investigation of the defendant’s background.” Williams v.
Taylor, 

 (citing 1 ABA Standards for Criminal Justice 4-4.1, commentary, p. 4-55
(2d ed. 1980)).

       The Kentucky Supreme Court concluded that White’s counsel did not perform deficiently
“simply because they did not pursue every conceivable strategy in preparation for the penalty
phase.” White III, at 4. Characterizing White’s arguments as “little more than a lengthy legal
essay on how the case could have been handled with the benefit of hindsight,” the Court
concluded that White’s challenges to counsel’s “penalty performance do not meet the standards
set out in Strickland for ineffective assistance.” 

 The Court credited Early’s and Charters’s
testimony that they both “thought the three defendants were innocent until learning that Fisher
wanted to testify,” and, “[b]ecause the defendants originally claimed they were not guilty,” the
court concluded “there was no reason to investigate White’s background or his physical or
mental health.”   Thus, the court determined that counsel’s performance “was not clearly
unreasonable or unprofessional under all the circumstances.” The court further justified trial
counsel’s performance by emphasizing that in 1980, “a defense attorney had no duty to discover
and provide information about every aspect of White’s entire life or an account of several
generations of his extended family.”

           1. Delay in Investigation.

       I begin with White’s counsel’s decision not to investigate mitigation at all until after trial
had already started. The Kentucky Supreme Court applied the incorrect standard when analyzing
this decision, holding that Charters and Early were not required to investigate mitigation because
their clients claimed they were not guilty.      As the majority notes, “the court might have
articulated a legal standard that contradicted Supreme Court precedent when it stated that
‘[b]ecause the defendants originally claimed they were not guilty, there was no reason to
 No. 21-5958                                 White v. Plappert                                      Page 68


investigate White’s background or his physical or mental health.’” Maj. Op. at 16. In fact,
Kentucky Supreme Court did articulate a legal standard that contradicts Supreme Court
precedent. The Court made clear in Williams that counsel always had an “obligation to conduct
a thorough investigation of the defendant's background.” Williams, 

. And the
Court explained that it is unreasonable to wait until days before trial to begin preparing
mitigation, let alone to wait until after trial has started, even if counsel has otherwise used the
time to competently investigate and handle the guilt phase. 

        Indeed, the Williams Court directed us to the 1980 ABA Standards for Criminal Justice to
determine the scope of an attorney’s responsibility at the time of White’s trial. 

(citing 1 ABA Standards for Criminal Justice 4-4.1, commentary, p. 4-55 (2d ed. 1980)
[hereinafter ABA Standards 1980]). Those standards make clear that reliance on the client’s
assertions regarding their own guilt is no substitute for a thorough factual investigation. ABA
Standards 1980, supra at 4-54.           Rather, an investigation must be prompt, should entail a
thorough investigation of the potential strengths of the state’s case, and should involve an
evaluation of each element the defense counsel might need to prove or disprove. Id. The
investigation must include a search for mitigation evidence that can be used at sentencing beyond
simple emotional appeals.         Id. at 4-55.     It should involve research into “the defendant’s
background, education, employment record, mental and emotional stability, family relationships,
and the like.” Id. Under those standards, it is not reasonable for an attorney to fail to promptly
conduct an investigation to support an entire phase of trial simply because the client asserts that
he is not guilty.

        The majority argues that “this single potential misstep . . . doesn’t provide grounds for
granting White habeas relief.” Maj. Op. at 16. I believe this misreads the law and extends
AEDPA deference beyond its bounds. The court’s reasoning that, because White claimed he was
not guilty “there was no reason to investigate White’s background or his physical or mental
health” was the only reasoning provided for the court’s analysis of the delay in investigation.4 In
short, the Kentucky Court analyzed the delay under a single standard—i.e., that counsel need not

        4
          Although, of course, the court provided other explanations for counsel’s other instances of deficient
mitigation investigation “after there was a change in strategy.”
 No. 21-5958                              White v. Plappert                                   Page 69


investigate the background of a client who claims to be innocent—and it was the wrong standard.
This was not merely a “stray thought.” Maj. Op. at 11 (citing Rogers, 

). Nor
did the Kentucky Court simply use imprecise language to describe an accurate and applicable
legal standard which it correctly framed and applied. See, e.g., Woodford v. Visconti, 

 (2002). The court asserted and then explicitly relied upon an incorrect legal rule, and
that rule was dispositive of the court’s decision on this particular point.

        The majority accurately notes that the Kentucky Court also repeatedly cited, quoted, and
paraphrased Strickland as the overarching standard. Maj. Op. at 16–17. But that does not excuse
the Court’s reliance on a more specific rule that is contrary to clearly established Federal law as
part of the Strickland analysis. A state court may not simply accurately cite Strickland and then
assert a new rule for the application of Strickland that contravenes Supreme Court precedent.
Williams, 

 (granting habeas relief where the state Supreme Court cited and
applied Strickland, but also relied on the erroneous belief that a Strickland analysis should be
guided by a more specific rule set out in Lockhart v. Fretwell¸ 

 (1993) in all cases).
That is what happened in this case. The Kentucky Court correctly cited Strickland. White III, at
3. But the court then went on to make up a new rule—a rule which contravened clearly
established law—for the application of Strickland. White III, at 5. The Kentucky Court acted
contrary to clearly established Federal law.

        Because the Kentucky Court relied upon the incorrect standard in evaluating White’s
attorneys’ performance prior to trial, his performance must be analyzed anew under the correct
standard. Williams, 

. The testimony at the postconviction evidentiary hearing
highlighted the utter failure of trial counsel to reasonably investigate the circumstances of
White’s alleged crime or any potentially mitigating circumstances before Fisher turned state’s
witness. Charters’s testimony during the postconviction evidentiary hearing in state court
underscored his failure to appreciate the gravity of White’s potential death sentence. When
asked, “[s]o this is a death penalty case, right,” Charters testified, “[i]t turned out to be, yes,” and
when counsel responded, “I mean, they [the state prosecutors] were seeking the death penalty,”
Charters said, “[i]n a half-hearted way; they couldn’t have even got a conviction up to that point,
I don’t believe.” This is consistent with Fisher’s testimony during postconviction proceedings
 No. 21-5958                             White v. Plappert                                  Page 70


that he did not understand he was facing the death penalty until the Commonwealth’s attorney
informed him.

       Perhaps because they wholly failed to appreciate the seriousness of their client’s peril,
White’s attorneys “abandoned their investigation of petitioner’s background after having
acquired only rudimentary knowledge of his history from a narrow set of sources” in violation of
Wiggins. 

. Only after Fisher agreed to testify for the Commonwealth did counsel
bother to inquire about White’s supposed “history dating to early childhood of bizarre conduct.”
If it was deficient in Williams to neglect to prepare for mitigation until a week before the trial, it
was surely deficient to neglect to prepare for mitigation at all until after the trial had started.
Williams, 

. Counsel had five months. They chose not to use them to form any
sort of mitigation case. The delay, itself, was deficient performance.

           2. Substandard Investigation.

       Counsel also performed deficiently in failing to investigate once they knew that Fisher
would testify for the state. The Kentucky Supreme Court acted unreasonably under clearly
established Supreme Court precedent in holding otherwise. The Court repeatedly asserted that
“the mere fact that counsel might have presented additional mitigating evidence does not
establish ineffective assistance,” White III, at 3, and “at the time of the 1980 trial, a defense
attorney had no duty to discover and provide information about every aspect of White’s entire
life or an account of several generations of his extended family,” id, at 5. This is true, but
unhelpful. While counsel did not have a responsibility to discover or present all mitigating
evidence, counsel did have an obligation to conduct a reasonable and thorough investigation.
Williams, 

. By the time of the Kentucky Supreme Court’s decision, it was clearly
established federal law that failure to thoroughly inquire into a defendant’s background, acquire
records, and follow up on leads is deficient performance. 

.

       The Kentucky Court distinguished Williams, noting that, in Williams, trial counsel
conducted no investigation into family life or background, whereas White’s counsel did some
investigation in the course of their insanity investigation. As the majority notes, the Kentucky
Court did not act contrary to clearly established Supreme Court precedent in doing so. Maj. Op.
 No. 21-5958                             White v. Plappert                                   Page 71


at 15–16. Williams was different. A court, however, may act unreasonably by declining to apply
a rule established in one case to a new factual scenario. Ramdass, 

.

       In Wiggins, the Supreme Court held that it was an unreasonable application of Williams
and Strickland to excuse counsel for “acquir[ing] only a rudimentary knowledge of [the
defendant’s] history from a narrow set of sources” and for failing to pursue leads that any
reasonably competent attorney would have followed up on. 

. The Court
noted that counsel had turned up some evidence of the defendant’s traumatic background during
their guilt phase investigation. 

. But counsel had focused the mitigation phase only on
evidence pertinent to guilt and conducted no additional mitigation phase investigation. 

. The state court excused Wiggins’s counsel’s decision to abandon their mitigation
investigation on grounds that, while the evidence of Wiggins’s childhood trauma that counsel
had was not as detailed or as graphic as the history uncovered in postconviction, “counsel did
investigate and were aware of appellant's background.” 

 at 518 (quoting Wiggins v. State, 

 (Md. 1999) (emphasis in original)).

       The Supreme Court criticized the state court for failing to consider that, “[i]n assessing
the reasonableness of an attorney’s investigation . . . a court must consider not only the quantum
of evidence already known to counsel, but also whether the known evidence would lead a
reasonable attorney to investigate further.” Id. at 527. It reminded us that “Strickland does not
establish that a cursory investigation automatically justifies a tactical decision [not to investigate]
with respect to sentencing strategy.” Id. The Court addressed ineffective assistance, holding:

       In light of what the PSI and the DSS records actually revealed . . . counsel chose
       to abandon their investigation at an unreasonable juncture, making a fully
       informed decision with respect to sentencing strategy impossible. The [state]
       Court of Appeals’ assumption that the investigation was adequate[] thus reflected
       an unreasonable application of Strickland.

Id. at 527-28 (cleaned up).

       Similarly, in Jells v. Mitchell, we held that “the [state] Court of Appeals unreasonably
applied Strickland when it neglected to find that Jells’s trial counsel’s failure to prepare for the
mitigation hearing” was deficient performance. 

 (6th Cir. 2008). This was
because “Jells’s counsel interviewed only three family members, neglecting to speak with many
 No. 21-5958                            White v. Plappert                                Page 72


other family members who had lived with Jells and were available,” because “[w]hen speaking
with the family members they did contact, their inquiry was brief and they failed to ask
sufficiently probing questions,” and because information that should have “‘prodded’ them into
action was readily available.”     

.     Jells’s counsel had done some background
investigation and was able to discuss Jells’s unstable childhood and academic difficulties. 

. But we held that “counsel’s awareness of Jells’s unstable home environment and academic
difficulties should have alerted them that further investigation by a mitigation specialist might
proved [sic] fruitful.” 

       The Kentucky Court made similar unreasonable errors to those made in Wiggins and Jells
in this case. True, there was some investigation into background. But that investigation was
rudimentary and, as in Wiggins, consisted only of the evidence turned up during counsel’s guilt
phase investigation.      The postconviction testimony demonstrates that White’s counsel
interviewed White’s family on two occasions. First, counsel interviewed White’s grandmother
as a potential character witness for White’s defense of complete innocence. This questioning
was limited to whether White had a character consistent with having committed the crime or a
proclivity for violence. It did not delve at all into family history, childhood trauma, abuse,
injury, or medical conditions.

       Second, counsel hurriedly gathered some of White’s family after they switched to an
insanity defense and asked them to describe “prior conduct that was bizarre.” Counsel explicitly
pushed them to come up with testimony that would make White seem insane. This inquiry
turned up some evidence that was related to mitigation, as demonstrated by the testimony given
by relatives at the competency hearing and, later, at trial. In that sense, as in Wiggins, “counsel
did investigate and were aware of appellant’s background.”          But, just as in Wiggins, the
information counsel collected was not the result of any deliberate mitigation investigation; it was
simply a byproduct of a guilt phase investigation that touched on similar topics. And, as in
Wiggins, this failure to focus at all on a mitigation case separate from the insanity case had
consequences. Counsel missed most of the evidence that was relevant solely for mitigation.

       Further, as in both Wiggins and Jells, counsel had evidence from the guilt phase
investigation that should have prodded them to look further before they wrote off the possibility
 No. 21-5958                            White v. Plappert                                 Page 73


of additional mitigation phase evidence. For instance, counsel knew that there had been sexual
abuse in White’s family, because they knew his mother had been raped by her stepfather. They
knew that White grew up in the same home as that stepfather. Yet they seemingly never
discovered that said abuse carried on into White’s childhood and that White observed it. They
knew that White drank and used drugs as a minor, but they never discovered who produced those
substances for him as a minor or how long he had been using them. They knew that White had
been whipped, but they did not discover the extent of the physical damage that the whipping
caused. They had a report from Dr. Evensen alleging that White’s abuse had affected his mental
processing (even if it did not render him not criminally responsible), but they never followed up
to discuss mitigation with Dr. Evensen or to provide Dr. Evensen with the full family history he
needed to form the conclusions he did postconviction.

       Nor was this the result, as the majority suggests it might be, of counsel asking about all of
these topics and simply not receiving detailed answers. Maj. Op. at 25–28. Several witnesses
testified that they were taken into a room with the attorneys and were only asked about “violent,”
“sexual,” or “bad” things that White had done and not about other parts of his history. One of
White’s half-siblings explicitly testified that the family was not asked about background,
childhood history, or childhood trauma. Dr. Evensen testified that counsel never sat down and
discussed his mitigation related conclusions with him at all. The majority is correct that we
cannot simply assume that counsel never asked about certain evidence because they did not
discover it. But here we have affirmative evidence of what counsel did and did not ask about.
They did not ask about the type of evidence presented at the postconviction hearing.

       Further, as in Jells, there were several key family members that counsel never even
attempted to speak with.     These were not distant family members as the Kentucky Court
suggested when it admonished “a defense attorney had no duty to discover . . . an account of
several generations of his extended family.” Counsel failed to speak to many of White’s half-
siblings—or aunts and uncles of similar ages raised as half-siblings—who had grown up with
him, lived with him, and were available. Jells, 

. Counsel knew that there was, at
a minimum, some abuse in that home, they knew that these relatives existed and lived in the
home. Yet they never followed up to determine what kind of abuse or trauma those children
 No. 21-5958                             White v. Plappert                                  Page 74


might have witnessed or experienced. The relatives that counsel ignored included Kay Hudson,
who could have told counsel about the extent of the whippings White endured, about White
being forced to suck urine from his bed sheets, about White’s father threatening to shoot all of
the children, about White’s seizures, about White’s suicide attempt, and about White observing
sexual abuse. Counsel also ignored Bobbie Hudson, who could have told counsel about the
rampant sexual abuse in the home.

        The majority attempts to distinguish Wiggins on grounds that, “[u]nlike the attorneys in
Wiggins, once White’s counsel learned about his history of abuse and trauma, they chased down
hours of testimony on the matter.” Maj. Op. at 24. But this misreads the legal rule set forth by
Wiggins: “In assessing the reasonableness of an attorney’s investigation . . . a court must
consider not only the quantum of evidence already known to counsel, but also whether the
known evidence would lead a reasonable attorney to investigate further.” Wiggins, 

. Thus, our analysis (and the Kentucky Court’s analysis) of the reasonableness of counsel’s
actions cannot be controlled by the fact that White’s counsel discovered “hours of testimony on
the matter” while conducting their guilt phase investigation. It is controlled by whether that
evidence would have spurred a reasonable attorney to investigate further in advance of the
mitigation phase. And in this case, as in Wiggins, there were numerous, obvious avenues of
needed investigation that counsel failed to pursue.

        The majority attempts to distinguish Jells on grounds that Jells involved counsel who did
not prepare any mitigation evidence until after the guilt phase of trial and, in analyzing that
delay, Jells relied on cases involving “circumstances where a finding of guilty [could not] come
as a surprise.” Maj. Op. at 30 (quoting Jells, 

). It is true that Jells is not controlling
as to the delay in investigation. But Jells held that counsel had performed deficiently in two
ways: first, counsel delayed preparation of mitigation evidence until after the start of trial
(despite it being no surprise that Jells was found guilty), and second, when counsel did conduct
an investigation, they failed to conduct an adequate investigation.         Jells, 

(describing issues of timeliness and then going on to say that “[i]n addition” counsel failed to
conduct an adequate investigation). That second finding, and the court’s reasoning for it, is
highly instructive in this case.
 No. 21-5958                            White v. Plappert                                Page 75


       White’s trial counsel had limited time due to their unreasonable decision not to pursue
any background evidence until after jury selection. But they could and should have spent that
time interviewing White’s half-siblings or pursuing any pure mitigation. They did not. The
record directly contradicts the Kentucky Supreme Court’s assertion that “White’s defense
counsel did the best they could with what they had to work with.” White III, at 7. Counsel were
the authors of their own lack of information. Counsel’s insufficient foundation for mounting a
successful mitigation argument was the direct product of counsel’s failure to conduct a
reasonable investigation; that counsel “had to work with” limited evidence was a direct result of
counsel’s failure to make reasonable inquiries into, let alone do “the best they could” to uncover,
the circumstances of the alleged crime and White’s background. Recognizing these errors,
moreover, does not improperly apply 20/20 “hindsight” to assess counsel’s performance.
Strickland, 

. Thus, on this record, like the one in Wiggins, “the [state court’s]
conclusion that the scope of counsel’s investigation into petitioner’s background met the legal
standards set in Strickland represented an objectively unreasonable application” of clearly
established precedent. Id. at 528-29.

       B. Prejudice

       Counsel’s deficient performance “does not warrant setting aside the judgment of a
criminal proceeding” unless it prejudiced the defense. Strickland, 

. To show
prejudice, the defendant must demonstrate “a reasonable probability”—meaning “a probability
sufficient to undermine confidence in the outcome”—“that, but for counsel’s unprofessional
errors, the result of the proceeding would have been different.” 

. This standard “do[es]
not require a defendant to show ‘that counsel’s deficient conduct more likely than not altered the
outcome’ of his penalty proceeding, but rather that he establish ‘a probability sufficient to
undermine confidence in that outcome.’” Porter v. McCollum, 

 (2009) (brackets
omitted) (quoting Strickland, 

).

       A state court’s weighing of evidence for prejudice is afforded AEDPA deference.
Williams, 

. In this case, the Kentucky Supreme Court reached prejudice. We,
therefore, owe AEDPA deference to the Kentucky Court’s finding of no prejudice. A habeas
petitioner is entitled to relief under AEDPA if “the state court’s decision is so obviously wrong
 No. 21-5958                                  White v. Plappert                                         Page 76


that its error lies ‘beyond any possibility for fairminded disagreement.’” Shinn v. Kayer, 

 (2020) (quoting Harrington, 

). The Kentucky Court held that:

        Even if all the testimony White suggests had been introduced during the penalty
        phase, it would not have been sufficient to counter or explain the brutalities of the
        three murders. See Nguyen v. Reynolds, 

 (10th Cir. 1997). Some of
        the evidence could have been detrimental to his case. As noted in Haight v.
        Commonwealth, Ky. 

 (2001), “the critical issue is not whether
        counsel made errors, but whether counsel was so thoroughly ineffective that
        defeat was snatched from the hands of probable victory.” There is no possible or
        reasonable chance that the omitted testimony would have had such an
        unquestionably favorable impact that it would have changed the sentence of the
        jury.

White III, at 6.

        I begin by concurring with the majority that the Kentucky Supreme Court did not act
contrary to law, but I would be remiss if I did not point out the inaccuracies in the Kentucky
Court’s apprehension of the law. Indeed, the Kentucky opinion bends nearly to the breaking
point the requirement that we read the opinion generously.

        The Kentucky Court did not establish a nexus requirement between the mitigation
evidence and the crime, but it came close to doing so. It stated that the evidence was not
sufficient to “counter or explain” the brutalities of the crimes. As the Supreme Court has been
careful to note, while evidence that cannot explain the crime may have less weight than evidence
which can, Thornell v. Jones, 

, 1311-14 (2024), mitigating evidence may alter the
jury’s selection of penalty even if it does not explain the crime, Williams, 

.
However, reading the Kentucky Supreme Court’s reasoning generously, as we must, I take the
“counter or explain” language to mean that the mitigation evidence did not outweigh the
aggravation (i.e., “counter” it). Thus, I read the Court’s reasoning as an inartful statement of the
correct test. Williams, 

.5

        Similarly, the Kentucky Court’s asserted that there was no “possible or reasonable chance
that the omitted testimony would have had such an unquestionably favorable impact that it would

        5
          I do not agree, however, that this is the same issue presented in Hodge v. Jordan, 

 (6th Cir.
2024) as the majority implies in one of its footnotes. Maj. Op. at 37 n.9.
 No. 21-5958                                    White v. Plappert                          Page 77


have changed the sentence of the jury,” and that to succeed White would have needed to show
that “defeat was snatched from the hands of probable victory.” The Supreme Court has made
clear that mitigation evidence may be mixed and still undermine confidence in a verdict—it need
not be “unquestionably favorable.” Porter, 

. And the Supreme Court has held that
the Strickland standard does not require that the additional evidence would have resulted in
“probable victory”—i.e., “that counsel's deficient conduct more likely than not altered the
outcome.” Strickland, 

. It merely requires a showing that the new evidence
establish “a probability sufficient to undermine confidence in the outcome.” 

.

       A generous reading of the Kentucky Court’s language, however, takes it to mean that,
because the evidence was not unquestionably favorable (i.e., because it was mixed) there was no
reasonable chance that the testimony would change the verdict when weighed against significant
aggravators. This is, again, technically the correct standard.6 Cullen v. Pinholster, 

 (2011). Mitigation evidence that carries a sufficient risk of harming the defendant’s case
rather than helping it may not bear a reasonable possibility of changing the outcome. 

 And
the Supreme Court has warned us against quibbling over the incautious use of the word
“probable” without the “reasonably probable” modifier. Woodford, 

. Thus, the
flaws in the Kentucky Court’s discussion of the appropriate standards do not rise to the level
required to make the opinion contrary to clearly established federal law under AEDPA.

       What remains, however, is whether the Kentucky Court unreasonably applied Strickland.
A failure to investigate and present mitigation evidence can be prejudicial even where some
evidence relevant to mitigation was presented at trial. The Supreme Court has “never limited the
prejudice inquiry under Strickland to cases in which there was only little or no mitigation
evidence presented.” Sears v. Upton, 

 (2010) (internal quotation marks and
citation omitted). Of course, where new mitigating evidence “would barely have altered the
sentencing profile presented,” the fact that the jury sentenced appellant to death under the near
identical sentencing profile suggests a lack of prejudice. 

 (quoting Strickland, 

). But this is not such a case. I agree with the majority that White’s evidence is “different in
both kind and degree from that introduced at trial.” Maj. Op. at 35. Where the new evidence is
       6
           Although, as discussed below, it was not reasonable reasoning in this case.
 No. 21-5958                            White v. Plappert                                Page 78


substantial or different in kind from the evidence presented at trial, courts have “found
deficiency and prejudice in . . . cases in which counsel presented what could be described as a
superficially reasonable mitigation theory during the penalty phase.” Sears, 

.

       Where counsel has presented enough new evidence in postconviction to change the
sentencing profile available to the jury, “we reweigh the evidence in aggravation against the
totality of available mitigating evidence” to determine whether a reasonable juror might have
considered the totality of the evidence differently in light of the new mitigation. Wiggins, 

. “[T]he totality of the available mitigation evidence” includes “both that adduced at
trial, and the evidence adduced in the habeas proceeding.” Williams, 

. The
relevant question is whether the sum total of the mitigating evidence, including both the evidence
counsel presented to the jury and the evidence that only came out in the postconviction
proceeding, undermines confidence in the outcome of the sentencing proceeding.

       White’s case closely resembles cases in which the Supreme Court or this court have
found that the mitigation evidence might undermine the aggravating factors associated with a
particularly brutal murder. In Williams, “dramatic[]” descriptions of “mistreatment, abuse, and
neglect during [the defendant’s] early childhood, as well as testimony that he was ‘borderline
mentally retarded,’ had suffered repeated head injuries, and might have mental impairments
organic in origin,” and evidence of remorse and of improvement in prison, bore a reasonable
probability of outweighing evidence of a lengthy history of burglary, theft, arson, and carjacking;
brutal assaults on an elderly victim leaving her in a vegetative state; and several murders
including the one at issue in which the defendant had beaten the victim to death with a mattock.

 (pre-dating White III).

       In Porter, evidence of cognitive defects caused by a brain abnormality; evidence that
Porter’s father beat his mother in front of him, once causing a miscarriage; evidence that Porter’s
father beat Porter and shot at him; and evidence of “extreme hardship and gruesome conditions”
during military service was “too much mitigating evidence that was not presented to now be
ignored.” 558 U.S. at 33-34 (citation omitted). It was sufficient to undermine confidence in the
outcome of the sentencing proceeding even though “Porter had been ‘previously convicted’ of
another violent felony. . . ; the murder was committed during a burglary; the murder was
 No. 21-5958                            White v. Plappert                                  Page 79


committed in a cold, calculated, and premeditated manner; and the murder was especially
heinous, atrocious, or cruel.” Id. at 32. Indeed, it was so clear that the mitigation undermined
confidence in the outcome of the case that the United States Supreme Court held that the Florida
Supreme Court was unreasonable in holding otherwise. Id. at 44.

       In Mason v. Mitchell, Mason presented evidence that his father ran a prostitution and
drug ring in which his mother was involved and which he witnessed, his father shot his mother,
both of his parents abused him, both of his parents used drugs as did he after the age of eight, and
he had a personality disorder as a result of his home environment. 

 (6th Cir.
2008). This evidence undermined confidence in the outcome of the sentencing proceeding even
though the state had presented evidence that Mason had raped his victim and then beaten her to
death with a board studded with nails. 

. The Ohio Supreme Court was unreasonable in
holding otherwise. 

.

       In Goodwin v. Johnson, Goodwin presented evidence that he “was borderline mentally
retarded and showed signs of organic brain impairment; had a low education level; was involved
in gang and criminal activity from a young age; had an unstable home life and was exposed to
drug use, alcoholism, and prostitution by his mother before she forced him out of the home; was
neglected and abused by his step-father; and had been exposed by his father to sexual abuse,
drug use, and criminal behavior.”      

 (6th Cir. 2011).        This undermined
confidence in the outcome of the sentencing proceeding despite evidence that Goodwin shot a
store clerk through the head during a robbery, causing extensive damage to the clerk’s brain and
skull such that blood and remains of the brain were likely spattered onto Goodwin. 

. It
was unreasonable for the Ohio Supreme Court to hold otherwise. 

.

       In Cauthern v. Colson, Cauthern presented evidence that he had suffered a litany of abuse
which “included physical abuse with brooms, belts and other implements” and emotional abuse
including isolation and verbal insults. 

 (6th Cir. 2013). This undermined
confidence in the outcome of the sentencing proceeding despite evidence that Cauthern had
strangled to death two members of the United States Army Nurse Corps during the course of a
robbery after a violent struggle and rape of the female victim.          

.     It was
unreasonable for the Tennessee Supreme Court to hold otherwise. 

 at 486
 No. 21-5958                             White v. Plappert                                 Page 80


       In this case, as in those cases, the victims were vulnerable, and their deaths were violent
and gory. The murder was committed as part of a robbery. Those facts are not unique to this
case and do not set it apart from the litany of death penalty cases where we have found that the
mitigation evidence undermined confidence in the outcome of the sentencing proceeding. The
mitigation evidence in this case is substantial, comprising nearly all of the types of evidence
presented in not just one of the above cases but all of them together. White was physically,
mentally, and emotionally abused—beaten with every implement that was used on Cauthern and
more. His family members alternated between neglecting him and engaging in extreme violence
towards him including shooting at him, forcing him out of his home, and forcing him to drink
urine. He witnessed the death of a sibling as a result of the same behavior that was directed
towards him on a routine basis and the death of his own father at the hands of an uncle. He
suffered head injuries and showed signs of mental incapacitation. He was exposed to drugs and
deliberately fed alcohol at a young age. He was exposed to inappropriate sexual activity as
Mason and Goodwin were, and much of that sexual activity was violent, including rapes and
beatings of underage family members in his presence, often by his own father. He developed the
same antisocial behaviors that lead to the murders because of this abuse. There was too much
mitigating evidence to ignore.

       Further, there is evidence that the jury was at least persuadable. Evidence of Jurors’
sustained deliberations and initial deadlock supports the inference that presentation of the
mitigation evidence could have changed the outcome. Cf. United States v. Velarde-Gomez, 

 (9th Cir. 2001) (collecting authorities for the proposition that lengthier
deliberations can indicate jurors found it difficult to reach a consensus); Mason, 

(finding it suggestive of prejudice that, as here, the “jury initially reported a deadlock” regarding
the defendant’s sentence).       This record raises “a probability” of prejudice “sufficient to
undermine confidence in [the] outcome.” Porter, 

 (quoting Strickland, 

). The Kentucky Court acted unreasonably in ignoring it.

       The Kentucky Court acted unreasonably in writing off large swaths of the mitigation
evidence as potentially prejudicial. While, in some cases, mitigation evidence may bear no
reasonable possibility of producing a different verdict because it carries too high a risk of being
 No. 21-5958                            White v. Plappert                                  Page 81


taken, instead, as aggravation evidence, Pinholster, 

, the fact that the good
was mixed with some bad does not automatically render the mitigation evidence unable to
outweigh the aggravation evidence. Williams, 

 (finding prejudice even though
“not all of the additional evidence was favorable to Williams”). Indeed, courts have found that
new evidence of substantial abuse and childhood trauma is mitigating to the point of
undermining confidence in a death penalty verdict issued after brutal crimes even where it was
mixed with highly prejudicial new evidence like a history of criminal behavior. Williams, 

.

       In this case, the potential prejudice inherent in the new evidence was that it could make
White seem dangerous and unstable, rendered cruel and incapable of controlling his actions
because of his upbringing. But that prejudice was already in the record and was worse at trial
because of the fabricated evidence of sexual assaults and cruelty to animals. It is not clear that
the new mitigation evidence could be any worse on that front than the old. The new mitigation
evidence countered some of the trial evidence of past misconduct and added some sympathetic
evidence. The new evidence compounded through family testimony—and explained through
expert testimony—the idea that, insofar as White was dangerous, it was because of a truly
horrific upbringing. It also presented, for the first time, the idea that White could be reformed.

       White’s case that the new evidence, while mixed, would move the needle in his favor, is
even stronger than that of Williams. 

. As with White, the mitigation in Williams
suggested that Williams had a tendency to violence. As with White, the mitigation included
evidence that Williams’s “violent behavior was a compulsive reaction.” 

. But, unlike
in this case, that tendency had manifested itself in a history of criminal behavior. 

. If
in Williams, the downsides of the mitigation evidence did not render counsel’s failure to
investigate and introduce it nonprejudicial, then it was unreasonable to hold that the downsides
of the mitigation in this case did.

       The majority offers several counterarguments regarding prejudice. First, the majority
argues that most of White’s postconviction evidence was cumulative of evidence presented at
trial. The majority presents a chart which categorizes broad swaths of postconviction evidence
as the same or similar to trial evidence simply because it could fall under the same broad
 No. 21-5958                            White v. Plappert                              Page 82


headings like “Abuse in White’s Household” or “Exposure to Traumatic Events.” But even if
one were to entirely write off the evidence of “abuse” as cumulative because some evidence of
abuse was presented at trial, the majority concedes that some of the evidence presented in
postconviction was different in “both kind and degree from that introduced at trial”—citing
“episodes of head trauma, car accidents, alcohol consumption, and seizures from his childhood”
as well as the attempt to kill him in utero. Maj. Op. at 35.

        To this I would add, at the very least, that White presented new evidence in
postconviction that a doctor concluded that White’s hostility, anti-social behavior, and lack of
control were the direct result of the trauma he suffered as a child—a fact not included in the
majority’s chart. No expert testimony was presented in mitigation at trial whatsoever, so the
addition of expert testimony on causation was certainly different in kind and degree from that at
trial. I would also add that multiple male family members, including White’s father, raped
White’s half-siblings and aunts who were raised as half-siblings in White’s presence. True, there
was evidence at trial that White’s father raped White’s mother before White was born, but no
evidence was presented at trial that White had personally observed sexual assaults. The first
evidence of White’s observation of sexual assaults was presented at postconviction. And I
would further add that White had suffered from suicidal ideation and had attempted suicide at
least once.    While there was evidence of psychological symptoms at trial, it focused on
childhood terror and bedwetting, rather than psychological symptoms occurring later in life.
These late appearing symptoms were new at postconviction. In sum, a large selection of the
postconviction evidence was entirely new, not cumulative, by any stretch of the imagination.
The evidence introduced at postconviction changed White’s sentencing profile in nontrivial
ways.

        Next, the majority argues that the new evidence would not have diminished White’s
“moral culpability.” Maj. Op. at 35–36. The majority takes a wrong turn from the start by only
considering whether the new evidence (and, only the evidence the majority found noncumulative
at that) alone would diminish White’s moral culpability. This is not the test. Indeed, it adds on
an impermissible extra requirement to the Strickland test. The question of whether mitigating
evidence affects the moral culpability of the defendant is simply a part of the analysis the jury
 No. 21-5958                                     White v. Plappert                                           Page 83


must answer when weighing mitigating and aggravating evidence. Cauthern, 736 F.3d at 486-
87. Under Strickland, when we analyze whether new evidence might undermine our confidence
in the jury’s determination, we do not consider the new evidence in a vacuum; we consider it in
light of the “totality of the available mitigation evidence.” Williams, 529 U.S. at 397.

         The majority compounds this error by ignoring a key link between the mitigating
evidence of abuse, trauma, and head injuries (both the new and the old) and moral culpability.
Dr. Evensen never testified at trial because counsel never followed up with him. But once he
reviewed the evidence gleaned from White’s family, including the seizures, head trauma, and
substance abuse evidence—all of which the majority finds unrelated to culpability—as well as
the extensive abuse, Evensen concluded that White’s upbringing had led directly to his behavior
at the time of the murders. If believed, Dr. Evensen’s testimony teaches that a jury should not
merely be sympathetic to White because of the abuse he suffered, but rather conclude that
White’s horrific history explained his actions on the day of the crime. The cumulative effect of
the new evidence regarding White’s history along with expert testimony linking it to the crime
might well bear on White’s moral culpability in the eyes of a juror.

         The majority argues that the combined mitigation evidence could not outweigh the
aggravating factors. Maj. Op. at 37. The majority concludes this because (a) the mitigation
could not provide a “real link” between the mitigation evidence and the crimes and (b) the
aggravation was particularly severe. Id. (citing Thornell, 144 S. Ct. at 1310-12). The first of
these reasons fails for the same reason the “moral culpability” argument fails. There was a link. 7

         7
           Even if there had not been a link, however, the majority over-reads Thornell. Thornell noted that
mitigation evidence not linked to the crime may have “diminished persuasive value” relative to evidence that is
linked to the crime in the eyes of some courts. 144 S. Ct. at 1310. It did not assert that such evidence has no
persuasive value or even that a sufficiently high quantity of evidence, unlinked from the crime, could never be so
obviously mitigating as to render a state court’s decision to the contrary unreasonable. Id. Thornell does not
overturn Williams’s holding that “even if it does not undermine or rebut the prosecution’s death-eligibility case,”
evidence that is unrelated to the crime may undermine confidence in a verdict. 

-99. Nor does it
overturn Smith v. Texas in which the court held that evidence with no nexus to the crime is relevant (i.e., the jury
“could reasonably deem [it] to have mitigating value”). 

 (2004) (quoting Tennard v. Dretke, 

 (2004)). Evidence not linked to the crime still has persuasive value, and there are instances in which
some quantum of powerful mitigation, unrelated to the crime, is so persuasive that we must grant habeas relief even
with a brutal murder. See, e.g., Williams, 

-99. In Thornell, the Court did not simply jump from the idea
that the evidence was not directly related to the crime to the idea that it would do “little good” for the petitioner, as
the majority suggests. The Court fully analyzed the possible effects of the evidence and concluded that the
underlying Arizona court decision was right that its cumulativeness and attenuation from the crimes rendered it
 No. 21-5958                                    White v. Plappert                                          Page 84


The expert witness provided it. The second fails because, as discussed above, even the types of
particularly gruesome murders present in this case can be undermined by the type of mitigation
evidence also presented.

         The majority also attempts to write off the cases cited above in which substantial
mitigation undermined confidence in the jury verdict after a particularly brutal crime by claiming
that those precedents apply only in “those cases in which ‘defense counsel introduced little, if
any, mitigating evidence.’” Maj. Op. at 37 (citing Thornell, 144 S. Ct. at 1314). But this is a
misreading of Thornell. The Court in Sears explicitly indicated that Williams, Porter, and
Rompilla are applicable to cases where trial counsel presented mitigation evidence, even “a
superficially reasonable mitigation theory.” 561 U.S. at 954-55. At no point in Thornell did the
Court state that this aspect of Sears is no longer good law. The Court does not overrule itself
silently, nor should we adopt readings of its opinions that assume it does so. Rodriguez v. Quijas
v. Shearson, 

 (1989).

         Instead, a closer look at Thornell reveals that the new evidence in Thornell failed to meet
even the threshold requirement of “alter[ing] the sentencing profile presented to the sentencing
judge.” 

 at 1311 (quoting Strickland, 466 U.S. at 699-700). Indeed, that failure was the
primary basis for the court’s holding. Id. The court differentiated Williams, Porter, Rompilla,
and Wiggins on grounds that, in those cases, there was no evidence presented at trial, and
(implicitly) therefore, the sentencing profile had changed on appeal. This reading of Thornell,
unlike the majority’s extrapolation from a single line of that opinion, is consistent with Sears.
561 U.S. at 954 (“[T]here is no prejudice when the new mitigating evidence ‘would barely have
altered the sentencing profile presented’ to the decisionmaker.” (quoting Strickland, 466 U.S. at
700)). But Thornell does not say that Williams, Porter, Rompilla, and Wiggins are never relevant
to any portion of the prejudice analysis in cases where trial counsel presented some mitigation.

         Thornell’s discussion of Williams, Porter, Rompilla, and Wiggins is inapplicable here.
Unlike in Thornell, the question is not whether the new evidence changes the sentencing profile;

unpersuasive in that particular case. Thornell, 144 S. Ct. at 1310-12. We should not draw from that analysis a rule
that we can never grant habeas relief where the mitigation is unlinked to the crime and the crime is brutal. Nor
should we let the fact that the mitigation evidence is unlinked to the crime end our analysis, as the majority does. To
do so ignores the clear example of Williams.
 No. 21-5958                                   White v. Plappert                                           Page 85


it does—as even the majority concedes. Maj. Op. at 35. The question is whether the cumulative
weight of mitigation at trial and at postconviction raises a reasonable possibility that a juror
might have been swayed. As to that question Williams, Porter, Rompilla, Wiggins, and Sixth
Circuit cases in which no mitigation was initially presented are still relevant. And, as above,
they counsel toward granting habeas relief.

        Finally,8 the majority discounts the fact that the jury initially split on grounds that
“sentencing another human being to death in any case is a grave undertaking that might require
prolonged consideration.” Maj. Op. at 38. First, this elides the difference between a deliberation
that took a long time because it was a grave undertaking and not one jurors wanted to rush into
and one in which at least some jurors clearly were not just hesitant and deliberative but actively
disagreed with the eventual holding. One could argue, as the majority does, that the former is
not an obvious indication that an increased quantity of mitigation would move the needle. But
see Velarde-Gomez, 269 F.3d at 1036 (collecting cases that say otherwise). The latter is a much
more clear-cut sign that jurors were unsure and might be persuaded.

        Second, the fact that the delay in the verdict might only be the result of the weighty
deliberations involved in the death penalty process isn’t the issue. The question is whether there
is a reasonable probability that the new evidence would have changed the verdict. For there to
be a reasonable probability that the new evidence would have changed the verdict, there need

        8
           The majority also rejects an additional counterargument by White that the new evidence is not cumulative
because White himself gave some of the evidence at trial. I do not think White’s argument particularly moves the
needle either because of the similarity between White’s testimony and that of his family. But I note the majority’s
argument because the majority claims that “White points to no Supreme Court decision clearly establishing that a
third party’s testimony on the same underlying evidence can establish prejudice. So this argument doesn’t get White
anywhere, either.” Maj. Op. at 38. The majority neglects Skipper v. South Carolina in which the Court held:
        Finally, the State seems to suggest that exclusion of the proffered testimony was proper because
        the testimony was merely cumulative of the testimony of petitioner and his former wife that
        petitioner's behavior in jail awaiting trial was satisfactory, and of petitioner's testimony that, if
        sentenced to prison rather than to death, he would attempt to use his time productively and would
        not cause trouble. We think, however, that characterizing the excluded evidence as cumulative
        and its exclusion as harmless is implausible on the facts before us. The evidence petitioner was
        allowed to present on the issue of his conduct in jail was the sort of evidence that a jury naturally
        would tend to discount as self-serving. The testimony of more disinterested witnesses -- and, in
        particular, of jailers who would have had no particular reason to be favorably predisposed toward
        one of their charges -- would quite naturally be given much greater weight by the jury.

 (1986). Any witness is more disinterested than the defendant himself.
 No. 21-5958                             White v. Plappert                                Page 86


only be a reasonable probability that the weighty deliberations were the result of significant juror
doubts—indicating that new evidence could have swayed a juror. It need not be a sure thing that
the divided jury was the result of doubts for the divided jury to be strong evidence that there is a
“reasonable probability” that things could have gone differently.

        In sum, the Kentucky Supreme Court unreasonably applied Strickland and its progeny
when it concluded that White failed to demonstrate prejudice. Rather, under clearly established
federal law, White has shown that “a reasonable probability exists that, but for counsel’s error,
one juror would have voted against death.” Williams v. Anderson, 

 (6th Cir.
2006). Considering this record in its entirety, as we must, I would “conclude that the available
mitigating evidence, taken as a whole, ‘might well have influenced the jury’s appraisal’ of
[White’s] moral culpability.” Wiggins, 539 U.S. at 538 (quoting Williams v. Taylor, 529 U.S. at
398). So, I would find that White is entitled to relief.

                                       III. CONCLUSION

        This appeal presents a narrow issue: whether White’s counsel met their constitutional
obligations in their penalty-phase representation.         Because this record cannot support “any
reasonable argument that counsel satisfied Strickland,” Harrington, 562 U.S. at 105, I would
conclude that Mr. White is entitled to habeas relief and reverse the district court’s decision.
I, therefore, respectfully dissent.