RECOMMENDED FOR PUBLICATION
                                  Pursuant to Sixth Circuit I.O.P. 32.1(b)
                                         File Name: 23a0196p.06

                     UNITED STATES COURT OF APPEALS
                                    FOR THE SIXTH CIRCUIT



                                                               ┐
 IN RE: DANNY HILL,
                                                               │
                                                  Movant.       >        No. 20-3863
                                                               │
                                                               ┘

                                On Petition for Rehearing En Banc
          United States District Court for the Northern District of Ohio at Youngstown.
                      No. 4:20-cv-01294—John R. Adams, District Judge.

                                  Argued En Banc: June 14, 2023

                                Decided and Filed: August 25, 2023

   Before: SUTTON, Chief Judge; MOORE, CLAY, GIBBONS, GRIFFIN, KETHLEDGE,
               STRANCH, THAPAR, BUSH, LARSEN, NALBANDIAN,
                  READLER, DAVIS and MATHIS, Circuit Judges.*
                                 _________________

                                              COUNSEL

ARGUED EN BANC: Sharon A. Hicks, OFFICE OF THE FEDERAL PUBLIC DEFENDER,
Cleveland, Ohio, for Movant. Benjamin M. Flowers, OFFICE OF THE OHIO ATTORNEY
GENERAL, Columbus, Ohio, for Respondent. ON SUPPLEMENTAL BRIEF: Sharon A.
Hicks, Matthew Gay, Calland M. Ferraro, OFFICE OF THE FEDERAL PUBLIC DEFENDER,
Cleveland, Ohio, for Movant. Benjamin M. Flowers, Zachery P. Keller, OFFICE OF THE
OHIO ATTORNEY GENERAL, Columbus, Ohio, for Respondent. ON AMICUS BRIEF:
Matthew F. Kuhn, OFFICE OF THE KENTUCKY ATTORNEY GENERAL, Frankfort,
Kentucky, Tyler R. Green, Tiffany H. Bates, ANTONIN SCALIA LAW SCHOOL, Arlington,
Virginia, for Amici Curiae.

       NALBANDIAN, J., delivered the opinion of the court in which SUTTON, C.J., and
GIBBONS, GRIFFIN, KETHLEDGE, THAPAR, BUSH, LARSEN, READLER, DAVIS and
MATHIS, JJ., joined. THAPAR, J. (pp. 15–19), delivered a separate concurring opinion in
which SUTTON, C.J., and GRIFFIN, KETHLEDGE and READLER, JJ., joined. CLAY, J. (pp.
20–31), delivered a separate dissenting opinion in which MOORE and STRANCH, JJ., joined.

        *Hon. Eric E. Murphy recused himself from this case, and the Hon. Rachel S. Bloomekatz, who was
confirmed on July 18, 2023, did not participate.
 No. 20-3863                                In re Hill                                     Page 2


                                      _________________

OPINION

                                      _________________

       NALBANDIAN, J., Circuit Judge. In 1986, Danny Hill was convicted of murder and
sentenced to death in Ohio state court. For 37 years, he’s been challenging his conviction and
sentence. The Supreme Court vacated the decision of a panel of this Court to grant federal
habeas relief. After that, we again granted habeas relief based on different reasoning, and an en
banc panel of this Court vacated that opinion and denied Hill’s petition for relief. With the first
petition unsuccessful, Hill has filed a second habeas petition. It’s “second or successive.” So
Hill must meet the gatekeeping provisions of 

(b)(2).

                                                I.

       The brutal facts are familiar by now.         Danny Hill and Timothy Combs kidnapped
12-year-old Raymond Fife as he was riding a bike to a friend’s house before a Boy Scout
meeting. Shoop v. Hill, 

 (2019) (per curiam); State v. Hill, Nos. 3720, 3745,

, at *1 (Ohio Ct. App. Nov. 27, 1989). Hill and Combs beat Fife up. See State
v. Hill, 

, 887–88, 902 (Ohio 1992). They raped him in multiple ways. See 

. They strangled him, bit his genitals, sodomized him, and burned him. See 

 Then, Hill
and Combs left Fife in a field to die. See 

 Fife’s father found him in that field—Fife died two
days later. 

 at 887–88.

       Hill was indicted for kidnapping, rape, aggravated arson, felonious sexual penetration,
and aggravated murder with specifications. 

. At trial, the government’s case was
substantial. Several witnesses testified that Hill was around the crime scene on the day of Fife’s
death. 

. The state introduced Hill’s confession to police that he had watched Combs
beat up and rape Fife. 

; see also Hill, 

, at *2. And it also introduced
Hill’s statement that he had been with the victim when Combs had gone to the Valu-King to get
more materials to torture Fife. Hill, 

, at *3, *33; Hill, 

. The
state introduced a broomstick into evidence—the broomstick, found behind the Valu-King, that
was used to commit felonious sexual penetration. Hill, 

. One of the
 No. 20-3863                                 In re Hill                                    Page 3


state’s witnesses testified that Hill had raped her in the same wooded area where Fife was raped,
and another testified that Hill had raped her multiple times after breaking into her home. 

. And other evidence for the rape showed that anal intercourse had been forcibly performed
on Fife. 

.

       Relevant here, the state’s forensic odontologist testified at trial that in his “professional
opinion, with reasonable degree of medical certainty, . . . Hill’s teeth, as depicted by the models
and the photographs that [he] had, made the bite on Fife’s penis.” 

. Hill called his own
forensic odontologist, who said “that he could not conclude with a reasonable degree of certainty
as to who made the bite marks on the victim’s penis.” 

 He followed up, however, with this:
“What I’m saying is either Hill or Combs, or both, could have left some of the marks but the one
mark that’s consistent with the particular area most likely was left by Hill.” 

       In the end, a three-judge state-court panel convicted Hill of kidnapping, rape, aggravated
arson, felonious sexual penetration, and aggravated murder with specifications. 

. The
trial court then sentenced him to death. 

. Hill appealed to the Ohio Court of Appeals—
challenging, among other things, the imposition of the death penalty based on an insufficiency of
the evidence on the convictions and on the aggravators, which included rape. See Hill, 

, at *3. The Ohio Court of Appeals explained that plentiful evidence, including Hill’s
confession that he had been with Combs at the time of Fife’s rape and torture, supported the rape
conviction: “The evidence of the rape included the biting of the penis, leaving teeth marks; the
pulling of the genitalia, bruising the pelvic area; and the penetrating of the anus with the broom
stick, perforating the rectum and rupturing the urinary bladder.” 

. The Ohio Supreme
Court agreed. See Hill, 

. And Hill’s subsequent state habeas petition proved
unavailing too. See State v. Hill, No. 94-T-5116, 

, at *7 (Ohio Ct. App. June
16, 1995).

       Hill filed his first federal habeas petition in 1996, challenging, among other things, the
denial of expert assistance on the bitemark evidence in his state post-conviction petition under
the Sixth, Eighth, and Fourteenth Amendments because “[t]he state of medical knowledge
regarding bite marks was in flux at the time of the trial.” (1996 Habeas Petition, p. 37.) See Hill
v. Anderson, No. 4:96 CV 0795, 

, at *53, *61, *146 (N.D. Ohio
 No. 20-3863                                         In re Hill                                            Page 4


Sept. 29, 1999). The district court denied the petition, and Hill appealed. While the case was
pending on appeal, the Supreme Court decided Atkins v. Virginia. See 

 (2002)
(holding that it violates the Eighth Amendment to execute intellectually disabled defendants).
Then, we remanded the case so that Hill could exhaust his state remedies on that claim. See Hill
v. Anderson, 

 (6th Cir. 2002).

         The state court determined that Hill was not intellectually disabled, so his death sentence
stood. See State v. Hill, 

 (Ohio 2009) (unpublished table decision); State v. Hill,

 (Ohio Ct. App. 2008). With that, Hill moved to reopen his habeas petition
in federal district court and amended his claims. The district court rejected this petition too.
See Hill v. Anderson, No. 4:96 CV 00795, 

, at *59 (N.D. Ohio June 25, 2014).

         A panel of this Court then reversed the district court’s denial of Hill’s Atkins claim and
granted habeas relief. See Hill v. Anderson, 

 (6th Cir. 2018). Next, the
Supreme Court vacated that decision and remanded for the panel to reevaluate its reasoning.
Hill, 

. The panel again granted habeas relief on the Atkins claim and granted the
writ of habeas corpus as to Hill’s death sentence. See Hill v. Anderson, 

, 295 (6th
Cir. 2020) (per curiam). Our Court, sitting en banc, vacated the panel’s opinion and denied
federal habeas relief. See Hill v. Shoop, 

, 399 (6th Cir. 2021) (en banc), cert. denied,

 (2022) (Mem.).

         In the middle of all this habeas action, Hill moved in state court for leave to file a motion
for new trial under Ohio Rule of Criminal Procedure 33(B)—a mechanism for allowing
defendants to move for a new trial based on “newly discovered evidence” that the defendant
“could not with reasonable diligence have discovered and produced at the trial.”1 This newly
discovered evidence—Hill said—was a 2013 report from the American Board of Forensic




         1The language of this Rule is similar to the evidentiary gatekeeping provision for second-or-successive
habeas petitions under 

(b)(2). See 

(b)(2)(B) (requiring a petitioner to show that
“the factual predicate for the claim could not have been discovered previously through the exercise of due diligence”
to hurdle the second-or-successive bar).
 No. 20-3863                                         In re Hill                                            Page 5


Odontology (“ABFO”) that suggested using bitemarks to identify a specific individual as the
biter might not be reliable. Hill later added a 2016 report backing up the 2013 report.2

        The state trial court allowed Hill to move for a new trial and conducted an evidentiary
hearing. But that court noted that “[i]t is settled practice” that “newly discovered” evidence only
“refers to evidence that existed at the time of trial but of which the moving party was ignorant.”
(R. 1-1, Motion for New Trial Decision, PageID 322.) And the court held that there was “no
probability” that a new trial would lead to a “different outcome” without the bitemark evidence
because the state had proffered so much other evidence of guilt. 3 (Id. at PageID 334.) The state
appellate court agreed, State v. Hill, 

 (Ohio Ct. App. 2018), and the Ohio
Supreme Court denied review, State v. Hill, 

 (Ohio 2019) (unpublished table
decision).

        That brings us to Hill’s second federal habeas petition. In this June 2020 petition, Hill
made two separate but related points: (1) his “conviction and sentence [we]re predicated on
fabricated scientific evidence, violating his due process rights” (R. 1, Habeas Petition, PageID
295.); and (2) the state trial court violated his due process rights by not properly conducting a
materiality review of the bitemark evidence under Ohio Rule of Criminal Procedure 33(B).
Because Hill believed that his petition was second in time but not “second or successive” under

(b), he did not ask for leave to file a “second or successive” petition.




        2“The 2016 ABFO guidelines recognize that even where a patterned injury is determined to be ‘human,’
only the following terms should be used as they relate to [the] questioned dentition [of] a bite mark: A) Excluded as
Having Made the Bitemark, B) Not Excluded as Having Made the Bitemark, C) Inconclusive.” (R. 1-1, Exhibit 1,
PageID 324.)
        3In a response to a different argument Hill made about the voluntariness of his statements to the police, the
trial court noted some of the most striking evidence: In his videotaped interview with the police after Fife was
murdered, Hill “volunteered details about the murder and protracted torture of Raymond Fife,” and “even
demonstrated the location of the parties during the attack on a chalkboard . . . . Hill, without any emotion in his
voice, described the boy’s struggle to run[.]” (R. 1-1, Motion for New Trial Decision, PageID 317.) The trial court
further noted that Hill was found guilty of three specifications for aggravated murder: kidnapping, rape, and
aggravated arson. “A conviction [on] any one of the specifications made Hill eligible to receive a death sentence.”
(Id. at PageID 330.) And even setting aside the other specifications, under Ohio law, “there can be more than one
principal offender in the commission of a crime.” (Id.) Here the trial court quoted the Ohio Supreme Court’s earlier
decision, which held that even absent the bite marks, the other “various sexual acts” performed while Fife “was
rendered unconscious” served as other evidence to meet the rape specification. (Id. at PageID 331.)
 No. 20-3863                                         In re Hill                                            Page 6


         Despite Hill’s characterization of the petition as only second in time, the district court
concluded that Hill’s petition was “second or successive” and thus required our authorization.
So it transferred the petition to a panel of this Court to determine whether Hill’s petition met the
gatekeeping provisions of § 2244(b)(2). See 

(b)(3)(A) (giving the federal
courts of appeals exclusive jurisdiction to determine whether a “second or successive” petition
passes the gatekeeping provisions of § 2244(b)(2)). Hill then filed his habeas petition on appeal.
The case manager informed Hill that he needed to file a corrected petition on the second-or-
successive form. And he did. (See Appellate R. 5, Corrected Petition, p. 1 (“Motion Under

 For Leave to File A Second or Successive Habeas Corpus Petition” (case
altered)).) He maintained that his petition was not “second or successive.” Yet he said that his
claim relied on “newly discovered evidence,” which is a prerequisite to meeting the gatekeeping
provisions of § 2244(b)(2). (See id. at 6.)

         The panel determined that Hill didn’t need to meet the gatekeeping provisions, nor did he
require our authorization to file, because his petition was second in time but not successive. The
panel transferred the petition back to the district court for a merits determination. Next, the state
moved for rehearing on whether Hill’s petition is “second or successive.” With that, we voted to
rehear this case en banc.4 And we review de novo whether Hill’s petition is “second or
successive.” See Lang v. United States, 

 (6th Cir. 2007).

                                                        II.

         The Supreme Court has distinguished two separate questions in these cases.                             The
threshold question is whether an application is “second or successive” under § 2244(b). If
the answer is yes, then the second question is whether to dismiss the successive application.
See Magwood v. Patterson, 

, 336–37 (2010).5                           So determining whether a

         4The decision that a petition is not “second or successive” is reviewable.     See Magwood v. Patterson,

, 330–31 (2010).
          5To be more specific, if we determine that a petition is “second or successive,” then a panel of this Court
must determine whether to authorize the filing of the petition. In doing so, the panel must consider whether the
petitioner has made a “prima facie showing that the application satisfies” the § 2244(b)(2) requirements. 

(b)(3)(C); In re McDonald, 

, 546–47 (6th Cir. 2008). In this context, “prima facie” means
“simply sufficient allegations of fact together with some documentation that would ‘warrant a fuller exploration in
the district court.’” In re McDonald, 

 (citation omitted). If we determine that the prima facie
 No. 20-3863                                          In re Hill                                             Page 7


numerically second petition is “second or successive” is the threshold determination for our
review. Start with the framework. Under the Antiterrorism and Effective Death Penalty Act
(“AEDPA”), “a state prisoner always gets one chance to bring a federal habeas challenge to his
conviction.” Banister v. Davis, 

 (2020). “But after that, the road gets
rockier.” 

         After a petitioner has filed his first petition, any future petition is governed by § 2244(b).
As we must, we start with the text of that provision. A petitioner can’t bring the same claim in a
second petition that was “presented in a prior application.” 

(b)(1). And when a
“claim presented in a second or successive habeas” petition “was not presented in a prior
[petition],” we must dismiss that claim unless one of two circumstances applies: (A) The
petitioner “relies on a new [retroactively applied] rule of constitutional law”; or (B) “[T]he
factual predicate for the claim could not have been discovered previously through the exercise of
due diligence” and “the facts underlying the claim . . . would be sufficient to establish by clear
and convincing evidence that, but for constitutional error, no reasonable factfinder would have
found the applicant guilty of the underlying offense.” 

 In short, it’s much harder to file an
authorized “second or successive” petition than it is for the petitioner to file his first habeas
petition.

         Importantly, though, the Supreme Court has said that not all petitions filed second in time
are “second or successive.” And when a “numerically second petition” is not “second or
successive,” it isn’t “subject to the restrictions of 

(b).” In re Bowen, 

, 705–06 (6th Cir. 2006); In re Hanna, 

 (6th Cir.), cert. denied sub nom.
Hanna v. Shoop, 

 (2021). So there are two kinds of second-filed petitions:
numerically second petitions that are not “second or successive” and numerically second
petitions that are “second or successive.” And the caselaw tells us how to distinguish them.




showing has been made, the “second or successive” application is transferred back to the district court. The district
court then determines whether the § 2244(b)(2) requirements have actually been satisfied before possibly turning to
the merits. 

(b)(4); see In re McDonald, 

 (explaining that the district court must then
“engage in additional analysis in order to ascertain whether but for the constitutional error, no reasonable factfinder
would have found [the petitioner] guilty of the underlying offense”).
 No. 20-3863                                         In re Hill                                            Page 8


        The Court has noted that “second or successive” is a “term of art” that is “not
self-defining.” Banister, 

 (quoting Slack v. McDaniel, 

(2000); Panetti v. Quarterman, 

 (2007)). And in examining what qualifies as
“second or successive,” the Court has looked for general guidance in two places. First, the Court
has “explored historical habeas doctrine and practice.” 

 at 1705–06 (explaining that “[i]n
particular” the Court has considered whether a “later-in-time filing would have constituted an
abuse of the writ” as explained in “pre-AEDPA cases” (cleaned up)). And second, the Court has
“considered AEDPA’s own purposes.” Id. at 1706. Those purposes include “to ‘conserve
judicial resources, reduc[e] piecemeal litigation,’ and ‘lend[] finality to state court judgments
within a reasonable time.’” Id. (quoting Panetti, 551 U.S. at 945–46). “[T]he phrase ‘second or
successive’ must be interpreted with respect to the judgment challenged.” Magwood, 

 (emphasis added).

        More specifically, the Supreme Court has carved out three circumstances when a petition
is second in time but isn’t “second or successive.” First, when a second petition raises a claim
that challenges a new state-court judgment—that is, not the one challenged in the first petition—
the petition is not “second or successive.” See 

. Second, a second petition containing a
claim—whether presented or not in the first petition—that would have been unripe at the time of
the filing of the first petition is not “second or successive.” Panetti, 551 U.S. at 943–45 (claim
presented only in second petition but would have been unripe in the first petition); Stewart
v. Martinez–Villareal, 

 (1998) (pre-AEDPA case where the claim was presented
in both petitions but only became ripe in the second petition).6 Third, when a second petition
contains a claim that, though raised in the first petition, was unexhausted at that time and not




        6It’s worth noting that both Panetti and Martinez-Villareal involved claims under Ford v. Wainwright, 

 (1986). In Ford cases, a competency-to-be-executed claim only ripens once the date for execution is set.
See Panetti, 551 U.S. at 934–35 (“Under Ford, once a prisoner makes the requisite preliminary showing that his
current mental state would bar his execution, the Eighth Amendment, applicable to the States under the Due Process
Clause of the Fourteenth Amendment, entitles him to an adjudication to determine his condition.”). In Panetti, the
Court seemed to cabin the ripeness argument to the “unusual posture” of Ford claims. Id. at 945. Nevertheless, we
have extended it to other ripeness claims. See, e.g., In re Salem, 

, 812–13 (6th Cir. 2011) (holding that
because an entrapment claim was unripe in the first petition, the second petition was not “second or successive”).
Those cases aren’t at issue today.
 No. 20-3863                                         In re Hill                                              Page 9


decided on the merits, the petition is not “second or successive.”7 See Slack, 529 U.S. at 486–87,
489.8

         With all this in mind, in making the threshold determination for a second-in-time
petition—that is, in determining whether the petition is “second or successive”—we need
to define “claim” and “judgment.”               That’s because the Supreme Court and AEDPA have
differentiated between claims and judgments for purposes of “second or successive” petitions.
A judgment is the legal mechanism “authorizing the prisoner’s confinement,” i.e., typically
conviction or sentencing. Magwood, 

 (citation omitted); see 

(explaining that “a district court shall entertain an application for a writ of habeas corpus [o]n
behalf of a person in custody pursuant to the judgment of a State court” (emphasis added)). A
claim is a legal challenge to that judgment. Gonzalez v. Crosby, 

 (2005)
(explaining that “a ‘claim’ as used in § 2244(b) is an asserted federal basis for relief from a state
court’s judgment of conviction”).

         So first, we ask, is the second petition challenging a new judgment or an old judgment?
See Magwood, 561 U.S. at 330–33; In re Caldwell, 

 (6th Cir. 2019); In re
Stansell, 

 (6th Cir. 2016); King v. Morgan, 

 (6th Cir. 2015).
If it’s a new judgment, then the petition is not “second or successive,” and we turn to the merits
of the petition. See King, 

. If it’s the old judgment that the petitioner challenged
in his first petition, we next ask, is the claim presented an old claim or a new claim? See In re
Wogenstahl, 

 (6th Cir. 2018). If it’s an old claim—that is, one that was
presented in the first petition—then it’s a “second or successive” petition that must be dismissed
under § 2244(b)(1). If it’s a new claim, we ask whether it was either unripe or ruled unexhausted
at the time of the first habeas filing. If so, then the petition isn’t “second or successive.” See id.


         7The ripeness and exhaustion exceptions aren’t tied to the text of § 2244.     Instead, as we noted above, the
Supreme Court has looked to the purposes of AEDPA and congressional intent to determine that, although not plain
on the face of the literal text of § 2244, not every numerically second petition is “second or successive.” See
Panetti, 

 (“Our conclusion is confirmed when we consider AEDPA’s purposes.”); 

 (“We conclude,
in accord with this precedent, that Congress did not intend the provisions of AEDPA addressing ‘second or
successive’ petitions to govern a filing in the unusual posture presented here[.]”); Banister, 

(“Here, both historical precedents and statutory aims point in the same direction[.]”).
         8Hill only challenges the second circumstance—whether the claim was previously ripe. See infra Part III.
 No. 20-3863                                 In re Hill                                 Page 10


at 627. If not, under the guidance we have from the Supreme Court, the petition is “second or
successive,” and the claim must meet the gatekeeping provisions under § 2244(b)(2)(B) to
survive.

       Although a mouthful, we can sum it up this way: When a second-in-time petition raises a
new claim purporting to question the previously challenged judgment, the new claim was neither
unripe nor unexhausted the first go-around, and the petitioner nevertheless failed to raise the
claim, it is “second or successive.” See In re Coley, 

, 457–58 (6th Cir. 2017).

                                                III.

       Now to Hill’s second petition. He raises two claims: (1) he challenges his conviction
under the Due Process Clause because professional standards surrounding bitemark evidence
have changed; and (2) he says that the state trial court violated his rights under the Due Process
Clause in denying his motion for new trial by failing to sufficiently evaluate the materiality of
the bitemark evidence to the state’s case. We must determine whether Hill’s petition is “second
or successive” based on these claims.
       Under our roadmap, we start with whether Hill is challenging the same judgment he
challenged in his first petition or a different judgment. Hill’s first ground for habeas relief
challenges the judgment of conviction, so he is challenging the same judgment. Hill’s asserted
second basis for habeas relief doesn’t get him anywhere because it’s attacking a motion-for-new-
trial ruling. By definition, a motion for new trial attacks the original judgment. And a ruling on
that motion doesn’t create a new judgment—it’s just another order. So the motion-for-new-trial
ruling is not a judgment for purposes of § 2244(b).         See In re Stansell, 

(explaining that “new judgments” means “new sentences” or “new convictions”). So we’ll focus
on the first ground—which serves as another challenge to the same judgment he attacked in his
first petition. Next, we ask whether Hill is raising an old claim or a new claim.
 No. 20-3863                                          In re Hill                                           Page 11


         To evaluate Hill’s argument, we have to define the claim at issue. Hill’s claim is that he
was denied his right to due process because the bitemark evidence at trial was unreliable based
on new ABFO reports. As the Warden concedes, Hill didn’t raise this claim in his first petition.9

         Next, we ask whether that claim was unripe or unexhausted at the time he filed his first
habeas petition. He only argues that his claim was unripe. But as Hill’s 1996 habeas petition
makes clear, a challenge to the bitemark evidence has always been available to Hill. So any
claim as to the bitemark evidence’s unreliability was ripe at the time of his first habeas petition.
In other words, “the events giving rise” to Hill’s claim—the introduction of bitemark testimony
at trial—had already occurred when Hill filed his first petition. In re Tibbetts, 

(6th Cir. 2017) (quoting In re Jones, 

 (2007)).

         Hill says that his petition was unripe because the first ABFO report he relies on only
came about in 2013. He points to In re Jones to show why. 

. In that case,
Michigan significantly changed its parole system after Jones was convicted and sentenced, with
the final changes occurring after Jones had filed his first habeas petition. 

 So, when Jones
filed another habeas petition challenging the new parole conditions on ex post facto grounds, we
said that he could not have brought the challenge in his first petition because “the events giving
rise to the claim,” i.e., the changes to parole conditions that “produced a sufficient risk of
increasing the measure of punishment attached to his conviction,” had “not yet occurred.” 

 at
604–05 (cleaned up); see 

 (explaining that Jones was “challeng[ing] the cumulative
effect of amendments to Michigan’s parole system, the last of which took effect . . . two years
after Jones’s initial habeas petition was filed”). As a result, the petition challenging “a parole

         9In his first habeas petition, Hill did raise a claim tangential to his new due-process claim saying,
“Petitioner was denied his rights as guaranteed by the Sixth, Eighth and Fourteenth Amendments to the United
States Constitution when his request for expert assistance in his post conviction petition to reexamine the dental
evidence was denied.” (1996 Habeas Petition, p. 37.) He explained that “[b]etween . . . trial . . . and the time of his
post-conviction much was learned about bite mark analysis [and] . . . [t]he state of medical knowledge regarding bite
marks was in flux at the time of the trial.” (Id.) And he said that while his “counsel on post-conviction requested
expert assistance to reexamine the dental evidence presented at trial,” the “trial court denied said motion without a
hearing.” (Id.)
        This challenge to a denial of post-conviction expert assistance is not the same as his new due-process
claim. And it’s that fact—not just the fact that Hill is putting forth new evidence—that makes his due-process claim
a new claim. See Franklin v. Jenkins, 

 (6th Cir. 2016) (explaining that the new evidence in that
case didn’t give rise to a new claim because that evidence didn’t transform the “gravamen of [the petitioner’s]
argument” (citation omitted)).
 No. 20-3863                                 In re Hill                                    Page 12


determination . . . that occurred after the prisoner’s initial petition was filed” was not “second or
successive.” 

 at 605–06.

         But Jones brought another claim too—one that resembles Hill’s claim in this case and one
that Hill doesn’t mention in his briefing. Jones’s second claim was based on jury selection, and
that made our job “more straightforward.” 

. That’s because, unlike his ex post facto
claim, the jury-selection claim “challenge[d] events that occurred at his trial.” 

 And that
meant the claim had to meet the gatekeeping provisions of § 2244(b)(2) as a “second or
successive” petition. Id.

         Hill’s bitemark-evidence claim is like Jones’s jury-selection claim, not like Jones’s ex
post facto claim. Hill could have made his due-process bitemark challenge in his first petition
because the events giving rise to the claim—the introduction of bitemark testimony at trial—had
occurred at the time of the first petition. In other words, just like Jones’s jury-selection claim,
Hill’s bitemark claim is based on “events that occurred at his trial.” Id. Just as Jones’s jury-
selection claim had to pass the gatekeeping provisions of § 2244(b)(2), so does Hill’s bitemark
claim.

         In a nearly identical case to Hill’s, In re Wogenstahl, we recognized that a petitioner
couldn’t bypass the gatekeeping provisions of § 2244(b)(2) simply by presenting new evidence
for a new Brady claim. In re Wogenstahl, 902 F.3d at 625–28. We said Wogenstahl’s petition
was “second or successive.” Id. at 627–28. In that case, the new evidence Wogenstahl proffered
undermined the testimony the government put on at trial. See id. at 625–26. Still, the petition
was “second or successive” because “the predicates underlying Wogenstahl’s current claims”—
the government introducing testimony at trial—“had already occurred when he filed his
petition.” Id. at 627. Any new evidence undermining the government’s trial testimony could go
toward meeting the gatekeeping provisions under § 2244(b)(2)(B), we said, but couldn’t go
toward showing that the claim wasn’t “second or successive.” Id. at 628.

         Under § 2244(b)(2)(B), when “the factual predicate for the claim could not have been
discovered previously through the exercise of due diligence” and that factual predicate “would
be sufficient to establish by clear and convincing evidence” that “no reasonable factfinder would
 No. 20-3863                                         In re Hill                                           Page 13


have found the applicant guilty of the underlying offense,” a petitioner can avoid dismissal of his
“second or successive” habeas petition. 

(b)(2)(B). It was Wogenstahl’s burden
to meet that standard, given that his petition was “second or successive.”

        And it’s Hill’s burden too.10 Hill presents new evidence—evidence that wasn’t available
to him at the time of trial. And on the new evidence, he raises a new claim—that he was denied
due process because his conviction is based on now-discredited evidence. But that claim has
always been ripe because the factual predicates for the claim occurred at trial. That is, the
bitemark testimony he now challenges occurred at trial, just like the testimony in Wogenstahl’s
case. So Hill must meet the gatekeeping provisions of § 2244(b)(2).

        It’s worth noting that we’re not alone in treating a claim based on newly existing
evidence the same way we treat Brady and Giglio claims in second-in-time petitions. Other
circuits have assumed that claims based on newly existing evidence are “second or successive,”
and they have evaluated whether such claims meet the gatekeeping provisions of
§ 2244(b)(2)(B). See Rhodes v. Smith, 

 (8th Cir. 2020) (holding that “two
peer-reviewed articles published” years after trial did not meet the gatekeeping provisions of
§ 2244(b)(2)); Feather v. United States, 

 (8th Cir. 2021) (rejecting a successive
petition where the petitioner argued that “changes in forensic medical science” meant that “the
government relied on ‘fundamentally defective’ evidence that was ‘inconsistent with sound
scientific methods’”); see also Case v. Hatch, 

 (10th Cir. 2013)
(categorically not allowing newly existing evidence to meet the gatekeeping provisions).

        The question of whether new evidence that was unavailable at trial can meet the
gatekeeping provision of § 2244(b)(2)(B) is not before us, so we leave it to the panel to decide
that question in the first instance and then perhaps the district court. In doing so, we are mindful


        10Hill tries to distinguish his case from In re Wogenstahl.  He points out that the ABFO reports—the new
evidence in this case—were only published after trial. He says that his new evidence is “newly existing
information,” rather than newly discovered evidence, like in In re Wogenstahl. (Hill’s Supplemental Briefing, p.
21.) That difference, he says, bolsters his argument that his due-process claim was unripe when he filed his first
petition, in contrast to Brady and Giglio claims, which he concedes would be ripe in a first habeas petition under In
re Wogenstahl. But Hill’s distinction is without a difference. The factual predicates in his case—the introduction of
bitemark evidence—occurred at trial. And that’s just like the testimony in In re Wogenstahl. So the claim was ripe
at the time Hill filed his first habeas petition.
 No. 20-3863                                 In re Hill                                  Page 14


of the fact that, even if Hill could use that kind of evidence to meet the gatekeeping provision of
§ 2244(b)(2)(B), to prevail on the merits, he must show that the state court’s adjudication
“resulted in a decision that was based on an unreasonable determination of the facts in light of
the evidence presented in the State court proceeding.” 

(d)(2) (emphasis added).
And, as mentioned above, plentiful evidence of the other forms of rape Fife endured—evidence
that was presented at trial—supported the rape conviction.

       We note that if we came out Hill’s way, most convictions involving forensic evidence
would never be final. The end of Hill’s logic is that a petitioner could raise a new due-process
claim, unmoored from the gatekeeping provisions, anytime a relevant scientific standard
changed. If that were the case, we’d be undermining § 2244(b)(2) and AEDPA generally—that
is, we’d be undermining the finality of convictions. See Banister, 

 (“AEDPA
aim[s] to prevent serial challenges to a judgment of conviction, in the interest of reducing delay,
conserving judicial resources, and promoting finality.”).

                                                IV.

       For these reasons, we hold that the petition is “second or successive” and must meet the
gatekeeping provisions of § 2244(b)(2). We return the case to the panel for further consideration.
 No. 20-3863                                In re Hill                                   Page 15


                                     ___________________

CONCURRENCE

                                     ___________________

       THAPAR, Circuit Judge, concurring. I concur in the majority’s thoughtful opinion. I
write separately because the time has come for our court to honor the law’s deadlines.

       In habeas, finality matters. It matters for the victims. It matters for the state. And it
matters for the criminal defendants. Recognizing this, Congress passed the Antiterrorism and
Effective Death Penalty Act with specific time limits. We aren’t free to ignore those limits. Yet
we have—and we should stop doing so.

                                                I.

       After torturing, raping, and killing a twelve-year-old boy, Danny Hill was convicted in
1986. In the last thirty-seven years, Hill has sought review in the United States Supreme Court
three times, Hill’s case has come before a panel of this court four times, and the en banc court
has now heard his case twice. At almost every turn, courts have ruled against him. Hill’s case
will now return to a panel—for the fifth time—to decide whether to let Hill file his second or
successive petition. See 

(b)(3)(B).

       According to AEDPA’s text, the panel must act within “30 days.” 

 § 2244(b)(3)(D).
That law says that “courts of appeals shall grant or deny the authorization . . . not later than 30
days after the filing of the motion.” Id. (emphasis added). It’s hard to imagine a clearer
command.

       Unfortunately, we’ve flouted this command in the past, interpreting the thirty-day
deadline as “advisory or hortatory rather than mandatory.” In re Siggers, 

 (6th
Cir. 1997). Our reasoning? We cited a substantive canon that says Congress’s commands are
mandatory only if they include a penalty for noncompliance. 

 (quoting McCarthney v. Busey,

 (6th Cir. 1992)). Because Congress “failed to specify a consequence for
noncompliance with the thirty-day limit,” we reasoned, “shall” doesn’t mean “shall.” 

 No. 20-3863                                 In re Hill                                   Page 16


       This reasoning falls short. For one, we forgot that parties can seek mandamus from the
Supreme Court to enforce the thirty-day rule. See Cheney v. U.S. Dist. Ct. for D.C., 

, 380–81 (2004). For another, the canon isn’t sound. The First Amendment, for example,
also fails to specify a consequence for noncompliance, but that doesn’t mean Congress can treat
it as advisory or hortatory. Gray-Bey v. United States, 

 (7th Cir. 2000)
(Easterbrook, J., dissenting). Finally, applying a substantive canon here fudges our limited role
in interpreting law. Our role is to enforce the linguistic import of enacted text. See Wilson
v. Safelite Grp., Inc., 

 (6th Cir. 2019) (Thapar, J., concurring in part and in
judgment); Hale v. Johnson, 

 (6th Cir. 2016). But substantive canons don’t
illuminate linguistic meaning.    They favor specific policy outcomes.       Amy Coney Barrett,
Substantive Canons and Faithful Agency, 

, 159–67 (2010). For that reason,
we don’t use substantive canons when the text is otherwise clear. See Miller v. French, 

, 340–41 (2000); Moskal v. United States, 

, 107–08 (1990). And here, the text is
clear. “Shall” means “shall.” See Miller, 

; Lexecon Inc. v. Milberg Weiss
Bershad Hynes & Lerach, 

 (1998).

       Worse yet, many circuits have fallen into the same error, with several adopting our
reasoning. See Rodriguez v. Bay State Corr. Ctr., 

 (1st Cir. 1998), abrogated
on other grounds by Bousley v. United States, 

 (1998); Galtieri v. United States, 

 (2d Cir. 1997), overruled in part on other grounds by Magwood v. Patterson, 

 (2010); In re Hoffner, 

, 307 n.11 (3d Cir. 2017); In re Williams, 

 (4th Cir. 2003); Gray-Bey, 201 F.3d at 867–68; Ezell v. United States, 

 (9th Cir. 2015); Browning v. United States, 

, 1263–64 (10th Cir. 2001) (en
banc). As far as I can tell, only the Fifth and Eleventh Circuits treat the rule as mandatory. In re
Williams, 

, 1102–03 (11th Cir. 2018) (Wilson, J., concurring) (discussing In re
Johnson, 

 (11th Cir. 2016) (en banc)); In re White, 

, 956 n.2 (5th
Cir. 2015).

       Though our circuit has treated the thirty-day rule as difficult to follow, it isn’t. Congress
gave us a simple task that should be easy to perform within thirty days: decide, using a prima
facie standard, whether a petitioner is entitled to file a second or successive petition. 28 U.S.C.
 No. 20-3863                                  In re Hill                                 Page 17


§ 2244(b)(3)(C). The criteria for filing a second or successive petition are spelled out right in
AEDPA. Id. § 2244(b)(1)–(2). Because these conditions are difficult to meet, it’s easy to judge
whether they’ve been met. See id. Accordingly, we don’t need an exhaustive opinion addressing
the petition’s merits. Id. § 2244(b)(3)(C).

       Indeed, AEDPA envisions that most second or successive petitions will be denied. Id.
§ 2244(b)(2). There are only two narrow exceptions. Id. § 2244(b)(2)(A) (new, retroactive rule
of constitutional law); id. § 2244(b)(2)(B) (newly discovered evidence so compelling that no
rational juror would find the defendant guilty). That’s why most of these petitions are routinely
denied in a brief order. And because our denial is unappealable, that order ends the case. Id.
§ 2244(b)(3)(E).

       Even in the extraordinary case when a petitioner makes the prima facie showing, our
order can be brief. That’s because AEDPA requires district courts to decide for themselves
whether a petition meets AEDPA’s stringent filing requirements. Id. § 2244(b)(4). And because
our decision to let a petitioner file a second or successive petition isn’t even binding on the
district court, there’s little reason for a panel to write more than a few pages.

       It should be easy to perform this narrow, gatekeeping function within thirty days,
especially when our decision isn’t binding or subject to later review. But even if not, we could
adopt procedures to ensure we meet the deadline. We could start by expediting the briefing. If,
for some reason, a panel needs to write more than a brief order, it could issue an order within
thirty days and release an opinion later. See Galtieri, 

. Or, true to the expectation
implicit in AEDPA’s text, we could develop a default rule that motions are automatically denied
after thirty days.   See Gray-Bey, 

 (Easterbrook, J., dissenting); 

(b)(2).

       In fact, we’ve already test run a procedure to ensure compliance with the thirty-day rule.
In 1996, our circuit relied on staff attorneys to draft proposed orders, and we’d wait to start the
thirty-day clock until the motion and draft order were submitted to a panel. In re Sims, 

, 48 n.1 (6th Cir. 1997). Unfortunately, that procedure conflicted with the text just as much as
 No. 20-3863                                 In re Hill                                   Page 18


blowing past the deadline: AEDPA gives us thirty days from “the filing of the motion,” not from
the submission of a proposed order. 

(b)(3)(D).

       Whatever procedure we choose, the point is that we must comply with the thirty-day rule.
That rule is anything but “advisory or hortatory.” Cf. Siggers, 

.

                                                II.

       Why does the thirty-day rule matter? First, Congress made it mandatory, and that should
be good enough for us. But also, the rule reflects the careful balance that Congress struck when
enacting AEDPA.

       Habeas plays “a vital role in protecting constitutional rights.” Slack v. McDaniel, 

 (2000). But AEDPA balances this role against at least three concerns: finality of
state convictions, abuse of habeas, and delay. See Woodford v. Garceau, 

(2003) (quoting Terry Williams v. Taylor, 

 (2000) (opinion of Stevens, J.)
(“Congress wished to curb delays, to prevent ‘retrials’ on federal habeas, and to give effect to
state convictions to the extent possible.”)). First, because habeas is a last-ditch challenge to
another sovereign’s conviction, AEDPA indulges a strong presumption that the conviction will
be upheld. 

(b)(2), 2254(d). This furthers “comity, finality, and federalism.”
Michael Williams v. Taylor, 

 (2000). Second, because federal habeas comes
after several rounds of review, AEDPA limits habeas to the most compelling cases. 

(b)(2).   This protects petitioners, for “[i]t must prejudice the occasional meritorious
application to be buried in a flood of worthless ones.” Brown v. Allen, 

 (1953)
(Jackson, J., concurring in judgment).       Finally, because delays keep states, victims, and
petitioners “in limbo,” AEDPA demands that petitions be resolved quickly.               

(b)(3)(D), (d)(1); Baze v. Rees, 

, 81 n.17 (2008) (Stevens, J., concurring)
(quotation omitted) (noting victims’ families are “left in limbo,” to relive a crime “for decades”);
Knight v. Florida, 

 (1999) (Breyer, J., dissenting from denial of cert) (stating
that long delays are “inhuman, degrading, or unusually cruel” for the prisoner). These concerns
are weighty, and AEDPA’s balance is “delicate.” Michael Williams, 529 U.S. at 436.
 No. 20-3863                                 In re Hill                                  Page 19


        The thirty-day rule is integral to this balance. It furthers federalism and finality by
keeping us from meddling in another sovereign’s conviction for longer than necessary. Because
second or successive petitions are the last stage in a long review process (thirty-seven years
here), the thirty-day rule has us decide quickly whether to let petitioners file them, curbing
abuse and focusing us on the petitions most likely to be meritorious. And, most obviously, the
thirty-day rule also prevents delay. Indeed, even the legislative record confirms that AEPDA’s
deadlines are essential to the statute’s purpose: to solve “the acute problem[] of unnecessary
delay,” a committee report explained, “[c]ourts are directed” to decide petitions “within specified
time periods.” H.R. Report 104-518, at 111 (Apr. 15, 1996).

        We have neither the authority nor the expertise to second-guess AEDPA’s delicate
balance. And we don’t have discretion to upset the balance by disregarding one of the statute’s
key components. But treating the thirty-day rule as “advisory or hortatory” does just that. Cf.
Siggers, 

.

                                         *       *        *

        This case has gone on for thirty-seven years. But AEDPA, correctly understood, ensures
that it will not go on forever. Our court should honor—not ignore—the balance Congress struck
in the thirty-day rule.
 No. 20-3863                                        In re Hill                                           Page 20


                                             _________________

DISSENT

                                             _________________

        CLAY, Circuit Judge, dissenting. Danny Hill was tried on capital charges in 1986. Hill
was convicted and sentenced to death, and his subsequent direct appeals and post-conviction
proceedings were unsuccessful. Decades later, based on a discovery in forensic science, Hill
moved for a new trial in Ohio state court. He was denied relief and now seeks to file a habeas
corpus petition challenging the admission of recently debunked bitemark evidence at his trial and
the state court’s subsequent denial of his motion for a new trial as a denial of due process. The
sole issue before the en banc court is whether Hill’s proposed habeas petition is “second or
successive” under 

(b). Hill’s habeas petition is not second or successive and
should be remanded to the district court. I therefore respectfully dissent.

                                               BACKGROUND

        The majority’s recitation of the facts presented by this case is in many ways misleading.
Hopefully, the following description of the factual background of the case will place the relevant
events in proper perspective.

        Hill’s trial began before a three-judge state-court panel in 1986. See State v. Hill, 

 (Ohio 1992) (“Hill II”).                The prosecution argued that Hill had actively
participated in the rape, murder, and other offenses committed against Raymond Fife. 

 The
majority’s version of facts would lead the reader to believe that Hill was the primary perpetrator
in the offenses committed against Fife. However, Tim Combs, now deceased, according to
Hill’s confession, “did everything to the victim.”1 

. Hill, who was 18 years old at the
time of the offense, merely admitted he was present during the commission of the offense. 

The critical evidence in proving that Hill was more than a spectator to these attacks was Dr.



        1Combs was sentenced to multiple life imprisonments and not death, even though he was convicted of the
same crime as Hill and considered the “principal offender.” Hill II, 595 N.E.2d at n.1; see State v. Combs, 

 (Ohio Ct. App. 2005).
 No. 20-3863                                       In re Hill                                          Page 21


Curtis Mertz’s testimony at trial that Hill had left a bite mark on Fife’s penis.2 Id. at 889. Dr.
Mertz, now deceased, was then an influential forensic odontologist who had previously served as
the president of American Board of Forensic Odontology (ABFO), a membership-based
organization formed in 1976 that issues guidelines for the collection, analysis, and reporting of
bite mark evidence. See National Academy of Sciences, Strengthening Forensic Science in the
United States: A Path Forward 173 (2009) [“NAS Report”]3; American Board of Forensic
Odontology, About the ABFO, https://abfo.org/ (last accessed August 1, 2023).

        In his opening statement, the prosecutor previewed that Dr. Mertz would identify Hill as
the person who had left a purported bite mark on Fife’s penis. Dr. Mertz then testified that it was
his “professional opinion, with reasonable degree of medical certainty, that Hill’s teeth, as
depicted by the models and the photographs that [Dr. Mertz] had, made the bite on Fife’s penis.”
Hill II, 

. The prosecutor returned to Dr. Mertz’s testimony in his closing
statement, emphasizing that the evidence showed that Hill, not Combs, had left the supposed bite
mark on Fife. Hill was ultimately convicted of kidnapping, rape, aggravated arson, felonious
sexual penetration, and aggravated murder with specifications, and was sentenced to death. 

        At the time of trial, Dr. Mertz’s testimony fell within acceptable limits set by the ABFO.
In 1986, and indeed until 2013, ABFO guidelines permitted odontologists like Dr. Mertz to
relate, “to a reasonable degree of dental certainty,” a “suspected biter to a bite mark.” See 2012
ABFO Guidelines, R. 1-12, Page ID #432. This meant that odontologists could, consistent with
best practices, testify that a person was “the biter,” “the probable biter,” “not excluded as the
biter,” or “excluded as the biter,” or that it was “inconclusive” whether the person was or was not
the biter. 

 The ABFO guidelines, of course, did not resolve whether a suspected “biter” had
left a given bite mark, but the guidelines endorsed the prevailing belief that odontologists were
capable of accurately and reliably identifying who had left a particular bite mark. See 

        2The majority highlights the weakness of its argument by focusing on prior rape allegations against Hill
and the horrendous nature of the crime, instead of confining its discussion to the legal issues presented.
Furthermore, although the majority claims that Hill may have committed other rapes, that allegation purportedly is
based on less than credible witness testimony.
        3Available at: https://www.ojp.gov/pdffiles1/nij/grants/228091.pdf.
 No. 20-3863                                 In re Hill                                    Page 22


Dr. Mertz’s identification of Hill as the bite mark contributor was thus consistent with the
contemporaneous view of odontologists’ abilities.

       More than two decades after Hill was sentenced to death and had begun pursuing his
initial round of collateral relief, forensic science underwent a seismic shift. In 2009, the National
Academy of Sciences (NAS) published a landmark report concluding that a wide swath of
forensic evidence routinely introduced in criminal trials lacked sufficient support to be
scientifically valid.   See NAS Report, supra, at 4–8.          Odontology shared many of the
shortcomings that pervaded other forensic disciplines.          Critically, odontologists had not
“establish[ed] the uniqueness of bite marks” and thus lacked a valid basis to make conclusive
bite mark identifications.    Id. at 174.   The ABFO guidelines exacerbated the issue.           The
guidelines failed to provide criteria for determining “whether the bite mark can be related to a
person’s dentition and with what degree of probability.” Id. The report stated that “[e]ven when
using the guidelines, different experts provide widely differing results and a high percentage of
false positive matches of bite marks using controlled comparison studies.” Id. The unreliability
of bite mark evidence is compounded, according to the report, because “second expert[s] are not
widely used” and because, “[b]ite marks often are associated with highly sensationalized and
prejudicial cases, and there can be a great deal of pressure on the examining expert to match a
bite mark to a suspect.” Id. at 175. The NAS report concluded that there was no known
“evidence of an existing scientific basis for identifying an individual to the exclusion of all
others” and that further “research [was] warranted in order to identify the circumstances within
which the methods of forensic odontology can provide probative value.” Id. at 176.

       The ABFO responded by substantially limiting the scope of acceptable bite mark
identification testimony. In 2013, the ABFO issued amended guidelines that continued to permit
odontologists to relate a suspected biter to a bite mark, but warned that the ABFO no longer
“support[ed] a conclusion of ‘the biter’ in an open population case”—which is to say, where the
universe or total number of possible bite mark contributors is unknown. 2013 ABFO Guidelines,
R. 1-12, Page ID #434. The ABFO set even more restrictive limitations in 2016, this time
amending its guidelines to preclude odontologists from making conclusive bite mark
 No. 20-3863                                    In re Hill                                   Page 23


identifications—beyond a broad claim that an individual could not be eliminated as a potential
biter—in any cases, open population or not.

        Given the substantially limited scope of the 2013 ABFO guidelines and their successors,
were Dr. Mertz to have testified in accordance with these guidelines, he would have been
precluded from identifying Hill as the person responsible for the alleged bite mark found on Fife.
Had Hill been tried in 2013 or in the decade since, any odontologist called on to identify Hill as
the source of Fife’s purported bite mark, notwithstanding the ABFO guidelines that preclude
such a determination, would have had to confront during cross-examination the widespread
consensus that such testimony is unsupportable and unscientific—in the unlikely event that the
odontologist were permitted to testify at all. See Daubert v. Merrell Dow Pharms., Inc., 

 (1993) (imposing an obligation on a trial judge to “ensure that any and all
scientific testimony . . . is not only relevant, but reliable).

        In 2014, Hill brought these scientific developments to the attention of an Ohio trial court
when he moved for leave to file a delayed motion for a new trial under Ohio Rule of Criminal
Procedure 33(B). Hill asserted in the motion that the revised ABFO guidelines constituted new
evidence demonstrating that the bite mark identification evidence used at his trial was invalid,
and that he was convicted in violation of his Fifth, Eighth, and Fourteenth Amendment rights.
The state trial court found that Hill was unavoidably delayed in discovering his new evidence
because the ABFO guidelines were not in existence at the time of his trial and granted Hill
permission to file a motion for a new trial.

        Hill filed his motion for a new trial in June 2016. In it, he recounted the scientific
discoveries made since Dr. Mertz took the stand in 1986. He also submitted an affidavit from
forensic odontologist Dr. Franklin Wright, another former president of the ABFO who had
reviewed the evidence from Hill’s trial. Based on his examination, Dr. Wright concluded that
Fife’s injury was “not a human bite mark.” 2014 Wright Aff., R. 1-4, Page ID #368. Dr. Wright
further opined that even ignoring his finding that the injury was not a human bite mark, Dr.
Mertz’s finding was not scientifically supportable because “[i]t is not possible to determine from
the evidence” presented at Hill’s trial that Fife’s injury “was created by [Hill] to the exclusion of
millions of other individuals in the open universe of possible biters.” 

 Page ID #368.
 No. 20-3863                                 In re Hill                                   Page 24


       The state trial court found that advances in forensic odontology impeached and
contradicted the trial testimony that marks on Fife could be linked to Hill. But the trial court
held that it was bound by the Ohio Supreme Court’s findings on the sufficiency of the evidence
of aggravated murder and its independent weighing of the sufficiency of the rape, kidnapping,
and aggravated arson specifications. The trial court concluded that, even if the bite mark
evidence were excluded, there was no probability of a different outcome in a new trial because
there was overwhelming evidence in support of Hill’s convictions for the kidnapping
specification. The Ohio Court of Appeals affirmed, State v. Hill (“Hill VI”), 

(Ohio Ct. App. 2018), and the Ohio Supreme Court denied further review, State v. Hill (“Hill
VIII”), 

 (Ohio 2019) (table), cert. denied, 

 (2020).

       Hill filed a habeas petition under § 2254 in the district court. He alleged that Ohio Rule
of Criminal Procedure 33 created a liberty interest under the Fourteenth Amendment, and that
the state trial court violated his due-process rights by not conducting an appropriate materiality
review of his new evidence as required under Rule 33. He further asserted that the Ohio Court of
Appeals did not properly analyze his due-process claims and made factual and legal errors. Hill
concluded that the state courts’ failure to follow the established process for reviewing his claim
violated his constitutional right to due process and a fair proceeding and was an unreasonable
application of clearly established federal law and an unreasonable application of the facts.

       The district court found that Hill’s habeas petition was second or successive and
transferred it to this court. A panel of this court held that Hill’s petition was second in time but
not second or successive, and transferred it back to the district court for further proceedings. In
re Hill, No. 20-3863, 

 (6th Cir. Aug. 30, 2022) (order), reh’g en banc
granted, opinion vacated, 

 (6th Cir. 2023) (order). The State petitioned for
rehearing en banc, which the en banc court granted. 

.
 No. 20-3863                                  In re Hill                                 Page 25


                                           DISCUSSION

       As the majority correctly identifies, we review de novo whether Hill’s petition is “second
or successive.” See Lang v. United States, 

 (6th Cir. 2007).

       Hill claims that his conviction and sentence violated the Constitution because the
prosecution relied on bite mark evidence that was discredited as scientifically unreliable in 2013
and 2016, and the Ohio state courts violated his due process rights by failing to follow
established procedures when adjudicating his motion for a new trial. Hill seeks to raise that
claim in a habeas petition under § 2254. The issue is whether he must satisfy § 2244(b)(2)’s
gatekeeping requirements before he may file his petition in the district court. Because Hill’s
petition is merely second in time, not “second or successive,” the petition is not governed by
§ 2244(b) and Hill may file the petition directly in the district court.

       Section 2244(b) limits the ability of habeas petitioners to file second or successive habeas
petitions. See 

(b). Although Hill filed an initial habeas petition in 1996 and
seeks to file a second petition now, the Supreme Court has repeatedly affirmed that a habeas
petition is not “second or successive” merely because it was “filed second or successively in
time[.]” Magwood v. Patterson, 

 (2010) (quoting Panetti v. Quarterman,

 (2007)); see also In re Cook, 

 (6th Cir. 2000) (order)
(“[T]he fact that [the defendant] has already filed one § 2254 application . . . is not necessarily
determinative of whether the current attempt is a ‘second or successive’ application.”). Indeed,
the Supreme Court has “declined to interpret ‘second or successive’ as referring to all § 2254
applications filed second in time, even when the later filings address a state-court judgment
already challenged in a prior § 2254 application.”         Panetti, 

 (citing Slack
v. McDaniel, 

, 486–87 (2000)); Stewart v. Martinez-Villareal, 

(1998)).

       The ripeness of a defendant’s argument at the time of his initial petition is critical in
determining whether a claim presented in a second habeas petition is considered second in time,
or instead second and successive. In Panetti, the Supreme Court considered how to categorize a
petitioner’s second habeas petition that raised a competency claim under Ford v. Wainwright,
 No. 20-3863                                       In re Hill                                         Page 26

 (1986), whereby the petitioner’s “first federal habeas application, which was fully
and finally adjudicated on the merits, [had] failed to raise a Ford claim.” Id. at 942 (citation and
quotation marks omitted). The Court found that the defendant’s claim was second in time, not
second and successive, notwithstanding the petitioner’s failure to raise a Ford claim in his first
petition, because the petitioner filed his “Ford-based incompetency claim . . . as soon as that
claim [became] ripe.” Id. at 945; see also Martinez-Villareal, 523 U.S. at 643–46 (holding that a
Ford claim, raised for the second time after petitioner’s first claim was dismissed as premature,
was not a “second or successive” petition under § 2244(b)). Our Court has applied Panetti
outside the context of a Ford claim, including in In re Salem, where we held that the defendant’s
entrapment claim was not ripe at the time of the initial petition, and “should therefore be treated
as a continuation of [the defendant’s] earlier, undoubtedly reviewable, application for federal
habeas relief.” 

 (6th Cir. 2011) (quoting Martinez-Villareal, 

).
The majority agrees that a claim not ripe at the time of an initial petition is not second and
successive, but for inexplicable reasons finds that Hill’s petition is nonetheless “second or
successive.”

        When determining whether a habeas petition is second or successive, the Supreme Court
and this court have “looked for guidance in two main places.” Banister v. Davis, 

 (2020). First, the courts have considered whether the habeas petition would have
historically “constituted an abuse of the writ, as that concept is explained in [the Supreme
Court’s] [pre-Antiterrorism and Effective Death Penalty Act (AEDPA)] cases.” 

 at 1706
(quoting Panetti, 

). “If so, it is successive; if not, likely not.” 

 Second, the
courts have looked to the purposes of AEDPA, including “conser[ving] judicial resources,
reduc[ing] piecemeal litigation, [and] streamlin[ing] federal habeas proceedings.”4 Panetti, 

.

        Hill’s petition does not constitute an abuse of the writ. A petitioner abuses the writ “by
raising a claim in a subsequent petition that he could have raised in his first, regardless of


        4The   majority unnecessarily delves into a distinction between judgments and claims in relation to
§ 2244(b). Hill’s petition is not “second and successive” even accepting the majority’s view that the petition is
based on a purported “new claim.”
 No. 20-3863                                 In re Hill                                   Page 27


whether the failure to raise it earlier stemmed from a deliberate choice.” McCleskey v. Zant, 

 (1991). This court has previously held that a claim that could have been raised in
an earlier petition but was not, either because the petitioner chose not to, or because the
petitioner was not aware of existing facts supporting the claim, is “second or successive” under
§ 2244(b). See 

(b)(2)(B)(i); In re Wogenstahl, 

, 627–28 (6th Cir.
2018). By contrast, a petitioner does not abuse the writ by raising a claim that would have been
unripe if it had been presented in an earlier petition. Panetti, 

. Such is the case
where “the events giving rise to the claim had not yet occurred” when the petitioner filed his
earlier habeas petition. In re Jones, 

 (6th Cir. 2010). The majority opines that
Hill’s new claim was ripe during his initial habeas petition because the bite mark evidence was
presented during trial and was appealable from that moment on.             This is a fundamental
misunderstanding of Hill’s claim.

       Hill’s claim arises out of events that occurred decades after he filed his first federal
habeas petition in 1996. His claim arose in 2013 and continued into 2016, when the scientific
community repudiated the bitemark evidence offered at his trial and the ABFO revised its
guidelines to account for that scientific revolution. Prior to those developments, Dr. Mertz’s
testimony “fell within acceptable limits” set by the ABFO and subscribed to by odontologists.
2020 Wright Decl., R. 1-11, Page ID #410. After the change, his testimony was deemed by the
relevant scientific community to lack a valid scientific basis. Given the timing of these events,
Hill could not have brought his claim in his initial petition. Hill was aware of the bite mark
evidence during his trial; however, his claim disputing the reliability of the evidence was not ripe
until the scientific community invalidated the evidence.

       Further, the majority claims its opinion preserves AEDPA’s objectives, because Hill
could have brought his bite mark claim in his initial petition, which would have streamlined his
proceedings.    However, it makes little sense to demand that Hill argue that bite mark
identification lacked a scientific basis at a time when bite mark identification was still widely
considered to be a valid form of forensic evidence, solely on the off chance that the science
might eventually change. Cf. Panetti, 

 (finding that “[a]n empty formality
requiring prisoners to file unripe Ford claims neither respects the limited legal resources
 No. 20-3863                                   In re Hill                                      Page 28


available to the States nor encourages the exhaustion of state remedies”). Rather, allowing Hill
to raise his claim in a second-in-time habeas petition once the scientific community reached its
consensus and his claim became ripe serves a greater judicial purpose, and is not contrary to the
purposes of the AEDPA. See id.; see also Jones, 

.

        Moreover, beyond placing an impossible standard on defendants that requires defendants
to predict scientific discoveries in their initial petition, the majority also directly flouts this
Circuit’s and the Supreme Court’s precedent. Hill’s inability to challenge the bite mark evidence
before the evidence was scientifically discredited is directly comparable to a petitioner’s inability
to challenge an execution that has not yet been set, Martinez-Villareal, 523 U.S. at 644–45, or
laws that have not yet been passed, Jones, 

.

        In particular, Hill’s claim in this case is directly analogous to a claim presented in Jones.
In Jones, the petitioner moved for authorization to file a second or successive habeas petition
based in part on his claim that changes to Michigan’s parole system since his conviction
constituted an ex post facto law. Jones, 

. Jones’ “ex post facto claim . . .
challenge[d] the cumulative effect of amendments to Michigan’s parole system, the last of which
took effect . . . two years after Jones’s initial habeas petition was filed.” 

. Given the
timing of these events, this Court held that Jones’s claim was “unripe when his initial petition
was filed” and that his claim was therefore second in time rather than second or successive. 

        Like the newly enacted change to Michigan’s parole system which rendered Jones’
petition second in time, the seismic shift in forensic odontology and newly formed scientific
consensus that such evidence lacks validity renders Hill’s petition second in time, and not second
or successive. As the Jones Court held, “no useful purpose would be served by requiring
prisoners to file ex post facto claims in their initial petition as a matter of course, in order to leave
open the chance of reviving their challenges in the event that subsequent changes to the [law]
create an ex post facto violation.” 

 Similarly, no useful purpose would have been served by
requiring Hill to file his claim challenging the scientific validity of the bite mark evidence in his
initial petition, to then lose on the basis that no scientific evidence supported his claim. Just as
we do not require petitioners to predict changes in the law, we should not require petitioners to
predict changes in science. The majority’s repeated assertion that Hill could have brought this
 No. 20-3863                                 In re Hill                                   Page 29


claim initially is misguided, and would have required Hill to mount a scientific challenge that
was not supported by the scientific evidence available at that time. In view of such enormously
consequential errors, this Court should acknowledge that Hill’s petition is merely second in time,
not second or successive, and hold that § 2244(b) does not apply to Hill’s petition.

        While the majority agrees that Jones’ ex post facto claim was second in time, the majority
attempts to argue that a second claim brought by Jones, which the Jones Court held was second
or successive, is more analogous to Hill’s petition. Jones’ second claim “contend[ed] that the
jury pool at his trial systematically excluded African-Americans, in violation of the Sixth
Amendment’s fair cross-section requirement.” Id. at 604. Because Jones’ claim “challenge[d]
events that occurred at his trial,” the Jones Court found that the claim fell “squarely within the
scope of § 2244(b).” Id. at 606. Attempting to analogize, the majority opines that Hill’s
bite-mark claim, like Jones’ cross-section claim, challenges events that occurred at his trial, and
thus must pass the gatekeeping provisions of § 2244(b).

        The majority is plainly wrong. In bringing his cross-section jury claim, Jones alleged no
new discovery or new law that changed the legal landscape regarding jury selection practices—
thus Jones’ jury claim was ripe at the time Jones brought his first petition. Clearly, Jones’
cross-section jury claim is completely unlike Hill’s claim, which is based on new scientific
evidence that did not exist when Hill filed his initial petition—new scientific evidence that
completely invalidates the bite mark evidence—and therefore was unripe at the time of the initial
petition.

        Further, the majority insists that “the factual predicates for the claim occurred at trial,”
thus any claim that is a derivative of those factual predicates would necessarily be “second and
successive” if not brought in the initial petition—relying on Wogenstahl, 902 F.3d at 627–28, in
which a Brady violation had occurred but the petitioner did not learn of it until later. See Brady
v. Maryland, 

 (1963). However, unlike in Wogenstahl, the factual predicate for
Hill’s claim did not exist at the time he filed his first habeas petition: it was not until the
scientific community repudiated the odontological evidence introduced at Hill’s trial, and the
ABFO manual was later reissued, that Hill’s claim became ripe. In other words, if Wogenstahl
had access to all then-available information at the time of his first habeas petition, he could have
 No. 20-3863                                 In re Hill                                   Page 30


brought a Brady claim. In contrast, if Hill had access to all then-available information at the
time of his first habeas petition, he still could not have brought a viable claim that the bite mark
evidence was invalid because the evidence had not yet been repudiated, the prior ABFO
guidelines still controlled, and his argument would have been contrary to the prevailing scientific
consensus.

       The majority also points to out-of-circuit cases in which courts have “assumed” that
claims based on newly existing evidence are “second or successive.” Critically, these cases do
not analyze whether the petitions are “second or successive.” See Rhodes v. Smith, 

 (8th Cir. 2020) (noting that the only question raised on appeal was whether the
district court erred in holding that the petitioner failed to satisfy the § 2244(b) requirements);
Feather v. United States, 

 (8th Cir. 2021) (treating the § 2255 motion as an
initial petition and reviewing it on its merits); Case v. Hatch, 

 (10th Cir.
2013) (assuming without analysis that the § 2254 application was “second or successive”). Thus,
these cases shed no light on the question presented to the Court in this case.

       To be clear, the only issue before the en banc court is whether Hill’s proposed habeas
petition is second or successive under § 2244(b). The merits of Hill’s challenge to the state trial
court’s decision, the strength of the evidence he has introduced, and all other matters bearing on
the merits of his claim are irrelevant to that issue and are not before the en banc court. Even
taking the merits into account, however, the majority grossly overstates and overestimates the
evidence against Hill that would qualify him for the death penalty. As described above, the bite
mark evidence is paramount in determining the level of Hill’s participation in the offense.
Likewise, the amount of time that has passed since Hill was convicted and sentenced should have
no bearing on the en banc court’s determination. Those considerations are already reflected in
the limitations § 2244(b) places on second or successive habeas petitions and in the Supreme
Court’s decisions interpreting the provision.      Hill did not unduly delay his claim, as the
invalidating evidence did not come to light until 2013. Nor should Hill’s efforts to litigate his
other claims, which have generated substantial judicial disagreement, influence the Court’s
decision here.
 No. 20-3863                               In re Hill                                  Page 31


       The en banc court should not alter the balance Congress and the Court have struck.
Because the majority’s approach cannot be reconciled with more than two decades of Supreme
Court precedent recognizing that a second-in-time habeas petition is not second or successive
when the petitioner could not have “receive[d] an adjudication of his claim” when he filed his
earlier petition, Martinez-Villareal, 

, the majority’s approach is grievously
misguided.

       For the aforementioned reasons, Hill’s habeas petition is second in time but not second or
successive, and should therefore be remanded to the district court for further proceedings as a
second in time petition. Because the majority declines to follow that straightforward procedure
based on its misreading of precedent and its mischaracterization of Hill’s claim, I respectfully
dissent.