In the

United States Court of Appeals For the Seventh Circuit ____________________ No. 21‐1402 JERRY S. WILSON, Petitioner‐Appellant, v.

DAN CROMWELL, Respondent‐Appellee. ____________________

Appeal from the United States District Court for the Eastern District of Wisconsin. No. 2:13‐cv‐01061 — Nancy Joseph, Magistrate Judge. ____________________

ARGUED SEPTEMBER 7, 2022 — DECIDED JANUARY 23, 2023 ____________________

Before SYKES, Chief Judge, and HAMILTON and BRENNAN, Circuit Judges. BRENNAN, Circuit Judge. Melvin Williams was shot and killed on May 23, 2009, and two other men—Robert Taylor and Romero Davis—were injured in the same shooting. A Wisconsin jury found beyond a reasonable doubt that Jerry Wilson was the gunman. He appeals from the district court’s denial of his habeas petition under 28 U.S.C. § 2254, claiming 2 No. 21‐1402

that he received constitutionally ineffective assistance from his trial and postconviction counsel. We do not reach the merits of Wilson’s claims because both are procedurally defaulted. Wisconsin state courts disposed of Wilson’s ineffective assistance of trial counsel claim on ad‐ equate and independent state procedural grounds. And Wilson failed to present his ineffective assistance of postcon‐ viction counsel claim for one complete round of state court review. The default of these claims is not excused by a suffi‐ cient showing of actual innocence, barring federal review of the merits. Accordingly, we affirm the district court’s denial of habeas relief. I The Shooting, Investigation, and Charges. In the early morn‐ ing hours of May 23, 2009, three people were shot during an “after‐set” party1 at a two‐story duplex unit on North 44th Street in Milwaukee. The party was large enough that at‐ tendees were both inside the duplex and outside in the street. Just before gunfire began, two vehicles passed through the crowded street in front of the duplex, and the cars’ occupants exchanged insults with party attendees in the roadway. The drivers parked nearby, and the passengers—who included the three eventual victims—walked back to the party to find the people who had yelled at them. A fistfight broke out in the street, and then the shooting started. Melvin Williams suffered a fatal gunshot wound to the chest and died that day. The two other victims survived. A

1 According to witnesses at trial, an after‐set party is like a house party

or block party, where guests pay an admission fee and alcohol is served. No. 21‐1402 3

bullet struck Robert Taylor in the foot, and Romero Davis re‐ ceived wounds to his stomach and right calf. Neither Taylor nor Davis could identify who shot them. Investigation of the crime scene yielded only a modest amount of physical evidence. Law enforcement recovered five .40 caliber bullet casings, four .38 caliber casings, a .40 caliber bullet, and several bullet fragments at the scene but never lo‐ cated the murder weapon. In general, the .38 caliber casings were damaged and flattened while the .40 caliber casings were in better condition. Police spoke with eyewitnesses early in the investigation. Shakira King attended the after‐set party and identified Wilson as the gunman to law enforcement. She also picked Wilson out of a photo lineup. Antwan Smith‐Currin, who lived in the upstairs duplex unit at the time, also identified Wilson as the gunman in a photo array. According to detective testimony, Samantha Coats and Sanntanna Ross identified Wilson as the shooter as well, alt‐ hough at trial the women either denied having made such identification or sharply qualified their prior statements. Of‐ ficers arrested Wilson in July 2009, and the State charged him with one count of reckless homicide and two counts of reck‐ less endangerment. Smith‐Currin testified at Wilson’s preliminary hearing and identified him as the gunman. When asked whether he saw other gunmen besides Wilson, Smith‐Currin answered, “No, sir,” but acknowledged that “People w[ere] trying to say 4 No. 21‐1402

that I was shooting because I was on the porch.”2 At the hear‐ ing, the trial court found probable cause to believe that Wilson committed a felony and ordered him bound over for trial. Jury Trial. In August 2010, Wilson went to trial with attor‐ ney Glen Kulkoski as his counsel. Given the minimal physical evidence, the case centered on the testimony of four eyewit‐ nesses. Smith‐Currin took the stand and identified Wilson as the gunman, consistent with his previous statements to law enforcement. He testified to seeing Wilson walk between two houses, approach the crowd in the street, and open fire with a handgun. Yet Smith‐Currin’s testimony contained discrep‐ ancies. For instance, he testified to standing on the porch when he saw Wilson open fire, but he was cross‐examined with his prior sworn statement that he had been in the street when he saw the shooting. King also testified at trial and identified Wilson as the shooter. King’s account largely mirrored Smith‐Currin’s: Wil‐ son emerged from between two houses on the same side as the duplex and opened fire. But King also provided certain discrepant details. For example, she was neither consistent in describing her position relative to the gunman, nor certain of the distance between them. At trial, she first suggested that she was two feet from the gunman. But following a courtroom distance demonstration, she changed that estimate to fifteen feet. She also said that the shooter had a ponytail but had pre‐ viously told police that he wore his hair in braids. Finally, King testified she was not involved in the street fight, but pre‐ viously told officers that she had participated.

2 The two‐story duplex has an upper and a lower porch. Smith‐Currin

testified at trial that he was on the lower porch at the time of the shooting. No. 21‐1402 5

The State also called two other eyewitnesses to testify. Sanntanna Ross said she did not see who shot because she was fighting in the street during the shooting. That prompted the State to try to impeach her with her prior statements to law enforcement inculpating Wilson. Per testimony from investi‐ gating detectives, Ross identified Wilson as the shooter and recognized his face in a photograph. In response to the im‐ peachment evidence, Ross claimed she felt pressure from po‐ lice to “get [her] to say things that [she] didn’t want to say.” Samantha Coats testified that, in the seconds before the shooting, she was looking out of a nearby second‐story win‐ dow with a view of the street. She described seeing an indi‐ vidual come into the street near the duplex and start shooting. When asked at trial, she agreed that the gunman’s silhouette fit Wilson’s description, but she did not make an affirmative identification. As with Ross, the State tried to impeach Coats with prior statements. According to police documents and testimony, Coats selected Wilson’s photograph during a photo lineup, indicated he was the shooter, and wrote “I’m sure is the shooter” on the photo lineup paper near her signa‐ ture. In response, Coats explained she was “under a lot of pressure” from law enforcement and believed that she “was going to be taken into custody.” Coats likewise agreed with defense counsel that her statements to police were made to please the detectives and to avoid getting herself in trouble. The State called other witnesses to talk about the physical evidence. Detectives described where they found the different bullet casings and explained that the location of the .40 caliber casings was generally consistent with a gunman firing from an alleyway near the duplex. A firearm examiner opined that 6 No. 21‐1402

the .38 caliber casings were all fired out of one gun while the .40 casings were all fired from a second weapon. After the State rested, Wilson called three witnesses in his defense. Kawana Robinson, Aaron Lee, and Shantell Johnson all testified that they did not see Wilson at the after‐set party the night of the shooting. All in, the accounts of the trial witnesses varied. For in‐ stance, the shooter’s height was described as five‐foot‐three by one witness, and five‐foot‐eleven by another. One witness said the shooter was wearing a fleece‐style top with no hood, while others testified he was either wearing a baseball hat or had a hood up. There was also disagreement about whether the shooter wore his hair in a ponytail or in braids. Finally, at least two witnesses claimed it was too dark to discern any de‐ tails about the gunman. The jury found Wilson guilty on all three counts, and the court sentenced him to 28 years’ imprisonment. Wilson’s § 974.02 Proceedings and Possible New Evidence. Post‐judgment, two events unfolded simultaneously. In the fall of 2010, Wilson obtained postconviction counsel (Thomas Simon)3 and challenged his conviction. Wilson began by pur‐ suing a claim for ineffective assistance of trial counsel which, in Wisconsin, is brought as a § 974.02 motion in the trial court. See WIS. STAT. §§ 809.30, 974.02; Lee‐Kendrick v. Eckstein, 38 F.4th 581, 586 (7th Cir. 2022). Wilson filed that motion in April

3 Throughout we refer to Thomas Simon, who assisted Wilson during

his § 974.02 proceeding, as Wilson’s “postconviction counsel.” “Postcon‐ viction counsel” refers exclusively to Simon and should not be confused with Christopher August, who assisted Wilson with his § 974.06 state col‐ lateral attack, or with Wilson’s current federal habeas counsel. No. 21‐1402 7

2011, arguing that trial counsel had failed to properly investi‐ gate the case, raise a key defense, and thoroughly cross‐exam‐ ine a State witness. Also during the fall of 2010, Wilson had been investigating new evidence. He alleges that three to four months after the trial concluded, he became aware of a new eyewitness through a fellow inmate named Deangelo Harvey. In late 2010, Harvey purportedly told Wilson that a woman living in the duplex was home on the night of the shooting, but he did not provide a name or any other specifics. Nonetheless, Wil‐ son claims he eventually received a letter from that woman— Lakisha Wallace—sometime between March and June of 2011. Per Wilson, Wallace explained in her letter that she had “in‐ formation about what happened that night” but provided no other details. Wilson said he wrote back asking if she would testify on his behalf and requesting her contact information. In a third letter, Wallace allegedly agreed and provided Wil‐ son a post office box number.4 Thereafter, Wilson claims that his mother got in touch with Wallace and that Wallace spoke with his postconviction counsel. Nonetheless, there is no evi‐ dence that Wilson’s postconviction counsel ever obtained an affidavit from Wallace or involved her in the direct appeal. The Wisconsin trial court denied Wilson’s § 974.02 motion on April 18, 2011, and Wilson appealed. In 2012, the Wiscon‐ sin Court of Appeals denied relief, and the Wisconsin Su‐ preme Court declined to grant review, ending Wilson’s direct appeal. Wilson’s § 974.06 Proceedings. Almost a year after Wilson lost his direct appeal, he acquired a notarized statement from

4 Wilson did not keep any of the letters nor did he make copies. 8 No. 21‐1402

Wallace—the same individual with whom he had allegedly exchanged letters in 2011. In her July 1, 2013, statement, Wal‐ lace accused Smith‐Currin of being the shooter and said that Wilson was innocent. Wilson then filed a pro se postconvic‐ tion motion under § 974.06 in Wisconsin state court, alleging ineffective assistance of both trial and postconviction counsel. He also sought a hearing on the “newly discovered” Wallace testimonial evidence. The state trial court denied relief, and the appellate court affirmed. Two years later, though, Wilson’s state collateral challenge gained new life. In September 2016, he renewed his claims by petitioning the Wisconsin Supreme Court for review. That court ordered the State to submit a response, in which the State acknowledged that Wilson was entitled to an eviden‐ tiary hearing on the newly discovered evidence. As the Wis‐ consin Supreme Court summarized, the State conceded in its response that “if the allegation at issue is accepted as true, there is a reasonable probability that a jury, looking at the old evidence and the new evidence, would have a reasonable doubt as to Mr. Wilson’s guilt.” So, the Wisconsin Supreme Court granted the petition for review and remanded on the newly discovered evidence claim. It held in abeyance the other claims, including Wilson’s ineffective assistance of post‐ conviction counsel claim. In August 2017, an evidentiary hearing was held at which Wallace testified to the information in her July 2013 statement. She explained that, on the night of the shooting, she was liv‐ ing on the first floor of the duplex, and there was a big party going on in the upstairs unit where Smith‐Currin lived. In the hours leading up to the shooting, Wallace witnessed Smith‐ Currin drinking, smoking, and ingesting pills on the porch. No. 21‐1402 9

As a result, Wallace believed that Smith‐Currin was under the influence at the time of the shooting: “Yeah, he was very much so under the influence. Like you could tell he was high, you know.” As the party ramped up, Wallace said she noticed commo‐ tion outside her unit and observed Smith‐Currin ask his brother for a firearm. She next saw Smith‐Currin go outside with the handgun and yell that the partygoers should move away from the house. According to Wallace, Smith‐Currin then ran down the front steps and opened fire on the people in the street. During the shooting, Wallace claims to have heard multiple weapons firing: “It wasn’t like it was just one gun. Like you could hear different guns going off. It wasn’t like just one person shooting outside.” Wallace testified fur‐ ther that, once the shooting stopped, Smith‐Currin tried to come inside her unit. She refused him entry but overheard Smith‐Currin tell his brother that he had just shot someone. Wallace also reported hearing Smith‐Currin discuss pinning the crime on Wilson. Wilson took the stand next. He explained how Wallace reached out to him after his conviction in 2011, and he de‐ scribed their alleged exchange of letters. Wilson also testified that, in the hours before the party, he had helped set up a mu‐ sic system for Wallace at the duplex. Yet despite having been to Wallace’s residence just hours before the shooting, Wilson said it never occurred to him that she might have information about the incident. Indeed, Wil‐ son never brought Wallace to trial counsel’s attention or oth‐ erwise reached out to her pretrial. Per Wilson, it was not until Wallace wrote to him that he realized she might have helpful information. And while Wilson claimed he notified 10 No. 21‐1402

postconviction counsel about Wallace during his direct ap‐ peal, he could not explain why his counsel failed to act on the Wallace lead. After the hearing, the state trial court denied Wilson’s re‐ quest for a new trial. In its oral ruling, the trial court found that, “[g]enerally, Miss Wallace’s testimony was credible and worthy of belief.”5 But the judge assessed Wilson’s statements differently, explaining, “Mr. Wilson’s testimony is not credi‐ ble. It is not worthy of belief. I give his testimony zero weight.” The court observed that Wilson had recounted re‐ ceiving letters from Wallace, yet Wallace testified she was il‐ literate. As the court explained, “Miss Wallace doesn’t have the ability to correspond with the defendant. She can’t read. She can’t write.” At bottom, the trial court held that Wilson was negligent in failing to present the newly discovered evi‐ dence to the jury and thus not entitled to a new trial. Wilson then made a strategic decision to streamline his case. He voluntarily dismissed his petition for review (with his ineffective assistance of postconviction counsel claim), which the Wisconsin Supreme Court had held in abeyance, so that he could appeal the denial of his request for a new trial based on new evidence. Nonetheless, Wilson’s appeal of his newly discovered evidence claim failed. The Wisconsin Court of Appeals agreed with the trial court that Wilson was negli‐ gent in not presenting the Wallace evidence earlier and denied relief. Soon after, the Wisconsin Supreme Court declined

5 The trial court qualified this credibility finding somewhat, explain‐

ing, “Miss Wallace does have some limitations that undermine her credi‐ bility, not enormously, but there are areas where her testimony could be more credible.” One such issue was that Wallace “ha[d] some difficulties in sequence of events.” No. 21‐1402 11

review. Having lost on the newly discovered evidence claim and having voluntarily dismissed his other claims pending in the Wisconsin Supreme Court, the doors to state court relief closed for Wilson. Habeas Corpus Petition. Wilson then turned to federal court. He had timely filed an original federal habeas petition on Sep‐ tember 20, 2013, which the district court stayed pending ex‐ haustion of state proceedings. After his state court path was foreclosed, he amended his habeas petition on July 30, 2019, alleging three grounds for relief: (1) ineffective assistance of trial counsel; (2) ineffective assistance of postconviction coun‐ sel; and (3) newly discovered evidence. The district court ruled that Wilson procedurally de‐ faulted his claim for ineffective assistance of trial counsel. Likewise, the court decided that the default was not excused because Wilson failed to make a sufficiently strong showing of actual innocence. On the ineffective assistance of postcon‐ viction counsel claim, the district court did not explicitly en‐ gage with procedural default. Instead, the court found that Wilson could not show constitutionally ineffective assistance on the merits. Finally, the district court disposed of the newly discovered evidence claim, finding that the discovery of new evidence alone does not qualify as grounds for federal habeas relief absent an independent constitutional violation. The dis‐ trict court also denied Wilson a certificate of appealability. At Wilson’s request, we granted a certificate of appealabil‐ ity under 28 U.S.C. § 2253(c)(2) for the following issues:  Whether Wilson has established ineffective assistance of trial counsel; 12 No. 21‐1402

 Whether Wilson has made a strong enough showing of actual innocence to excuse any procedural defaults;  Whether the federal constitutional right to counsel applies to Wisconsin postconviction counselʹs performance; and  Whether, if the federal constitutional right to counsel applies to Wisconsin post‐convic‐ tion counsel, the standard for ineffective as‐ sistance is met here. After reviewing the petition and record, we affirm the district court’s denial of Wilson’s petition for federal habeas relief for the reasons that follow.6 II As noted, the district court dismissed Wilson’s habeas pe‐ tition. “When reviewing a district court’s ruling on a habeas corpus petition, we review the district court’s factual findings for clear error and rulings on issues of law de novo.” Sanders v. Radtke, 48 F.4th 502, 508 (7th Cir. 2022) (quoting Lee‐Kendrick, 38 F.4th at 585–86). As to whether a claim is procedurally de‐ faulted, our review is de novo. Garcia v. Cromwell, 28 F.4th 764, 771 (7th Cir. 2022) (citing Johnson v. Thurmer, 624 F.3d 786, 789 (7th Cir. 2010)).

6 The court thanks Vladimir J. Semendyai, Esq., Andrew P. LeGrand,

Esq., Pooja Patel, Esq., and Zachary T. Reynolds, Esq. of Gibson, Dunn & Crutcher LLP for accepting this appointment and for their fine represen‐ tation of Wilson throughout this appeal. No. 21‐1402 13

A We first consider whether Wilson’s claim for ineffective as‐ sistance of trial counsel is procedurally defaulted. The State contends it is because the state court disposed of Wilson’s claim on an adequate and independent state law ground. Wil‐ son seems to acknowledge this but focuses instead on over‐ coming default through the actual innocence gateway. We hold that Wilson’s claim for ineffective assistance of trial counsel is indeed procedurally defaulted. “[A] state prisoner must exhaust available state remedies before presenting his claim to a federal habeas court.” Davila v. Davis, 137 S. Ct. 2058, 2064 (2017) (citing 28 U.S.C. § 2254(b)(1)(A)). A “corollary” to that rule is that federal courts may not review federal claims that the state court de‐ nied on an adequate and independent state procedural ground. Id. So, we begin by examining the state court’s treat‐ ment of Wilson’s claim for ineffective assistance of trial coun‐ sel. The Wisconsin Court of Appeals was the final state court to evaluate Wilson’s ineffective assistance of trial counsel claim, and it denied that claim as inadequately pleaded under State v. Allen, 682 N.W.2d 433 (Wis. 2004).7 Per Wisconsin law, a defendant claiming ineffective assistance of counsel must plead “sufficient material facts—e.g., who, what, where, when, why, and how—that, if true, would entitle him to the

7 The Wisconsin Supreme Court denied Wilson’s ensuing petition for

review without comment. Therefore, we look to the Wisconsin Court of Appeals’ decision. See Wilson v. Sellers, 138 S. Ct. 1188, 1192 (2018) (holding that federal courts on habeas review look to the “last related state‐court decision that [ ] provide[s] a relevant rationale”). 14 No. 21‐1402

relief he seeks.” Id. at 436; see also id. at 441–42; State v. Bentley, 548 N.W.2d 50, 53–54 (Wis. 1996). State trial courts may deny such a claim without a hearing based on a defendant’s recita‐ tion of “conclusory allegations” or failure to “raise facts suffi‐ cient to entitle the movant to relief.” Allen, 682 N.W.2d at 437; see also Whyte v. Winkleski, 34 F.4th 617, 622 (7th Cir. 2022) (de‐ scribing the Allen pleading standard). Applying that standard, the Wisconsin Court of Appeals determined that Wilson’s ineffective assistance of counsel claim was insufficiently pleaded under Allen: “Despite a lengthy recitation of the standards set forth in Bentley and Al‐ len for a sufficient postconviction motion, Wilson fails to make sufficient allegations to warrant relief.” The state appellate court continued, “Because the allegations in the postconvic‐ tion motion were insufficient under Bentley and Allen, whether to grant a hearing was committed to the [trial] court’s discretion. We discern no erroneous exercise of that discre‐ tion.” In denying Wilson’s claim for ineffective assistance of trial counsel under the Allen standard, the state court of ap‐ peals relied on an adequate and independent state law ground. As stated, federal courts “may not review federal claims that were procedurally defaulted in state court—that is, claims that the state court denied based on an adequate and independent state procedural rule.” Davila, 137 S. Ct. at 2064. The Allen standard at issue here is both adequate and inde‐ pendent. As to adequacy, “For a state‐law ground to be ‘ade‐ quate,’ it must be ‘firmly established and regularly followed.’” Clemons v. Pfister, 845 F.3d 816, 820 (7th Cir. 2017) (quoting Walker v. Martin, 562 U.S. 307, 316 (2011)). The state law ground also “must not have been applied in a manner that No. 21‐1402 15

‘impose[s] novel and unforeseeable requirements without fair or substantial support in prior state law’ or ‘discriminate[s] against claims of federal rights.’” Id. (quoting Walker, 562 U.S. at 320–21). When examining the adequacy of a state law pro‐ cedural ground, our review is limited to whether the proce‐ dural ground “is a firmly established and regularly followed state practice at the time it is applied, not whether the review by the state court was proper on the merits.” Lee v. Foster, 750 F.3d 687, 694 (7th Cir. 2014). We have previously held the Allen pleading standard is a firmly established and regularly followed state practice, and we do so here. In Lee v. Foster, the Wisconsin Court of Appeals denied Lee’s claim for ineffective assistance of counsel and “found that the allegations regarding [Lee’s] postconviction counsel’s performance were conclusory and legally insuffi‐ cient” under the Allen standard. Id. at 693. On federal habeas review, we held that Lee’s claim was procedurally defaulted and that the Allen rule “is a well‐rooted procedural require‐ ment in Wisconsin and is therefore adequate.” Id. at 694. So, the Allen standard functions as an adequate state law ground for denial of Wilson’s ineffective assistance of trial counsel claim. The Allen pleading standard is also independent. A state‐ law procedural ground satisfies the independence prong when “the court actually relied on the procedural bar as an independent basis for its disposition of the case.” Lee‐Kendrick, 38 F.4th at 587 (quoting Garcia, 28 F.4th at 774). Here, the Wis‐ consin Court of Appeals explicitly referenced and relied upon the Allen procedural rule in disposing of Wilson’s claim for ineffective assistance of trial counsel. Thus, the Allen standard served as an independent state law ground for denying 16 No. 21‐1402

Wilson’s claim. We have reached the same conclusion in other cases implicating the Allen standard. See, e.g., Lee, 750 F.3d at 693 (holding that the Allen rule “clearly served as an inde‐ pendent basis for the court’s denial of [petitioner’s] motion”); Triplett v. McDermott, 996 F.3d 825, 829–30 (7th Cir. 2021) (con‐ cluding that the Allen pleading standard is an adequate and independent basis for the state court’s denial of petitioner’s ineffectiveness claim). So, the district court properly ruled that Wilson’s ineffective assistance of trial counsel claim was procedurally defaulted, and we affirm that decision. B Next up is Wilson’s claim that his postconviction counsel rendered ineffective assistance during the § 974.02 proceed‐ ing. This claim implicates the proper classification of § 974.02 proceedings, but in Lee‐Kendrick we already decided that: “[A] claim of ineffective assistance of counsel under [Wisconsin Statute] § 974.02 is part of a direct appeal rather than a request for collateral review.” 38 F.4th at 587. So, 28 U.S.C. § 2254(i), which bars federal habeas relief for the ineffective assistance of counsel at collateral post‐conviction proceedings, does not preclude Wilson’s claim here. With that, we move to whether Wilson procedurally de‐ faulted his claim for ineffective assistance of postconviction counsel. The State argues that Wilson defaulted this claim by failing to present it for one complete round of state court re‐ view. Wilson does not vigorously contest that position, focus‐ ing instead on overcoming default. Under 28 U.S.C. § 2254(b)(1)(A), a petition for federal ha‐ beas relief shall not be granted unless it appears that “the ap‐ plicant has exhausted the remedies available in the courts of No. 21‐1402 17

the State.” Applying that provision, we have held that “[t]o fairly present [a] federal claim, a petitioner must assert that claim throughout at least one complete round of state‐court review, whether on direct appeal of his conviction or in post‐ conviction proceedings.” Richardson v. Lemke, 745 F.3d 258, 268 (7th Cir. 2014) (citing McDowell v. Lemke, 737 F.3d 476, 482 (7th Cir. 2013)). The complete round rule “means that the pe‐ titioner must raise the issue at each and every level in the state court system, including levels at which review is discretion‐ ary rather than mandatory.” Id. (citing Lewis v. Sternes, 390 F.3d 1019, 1025–26 (7th Cir. 2004)). Wilson voluntarily dismissed his claim for ineffective as‐ sistance of postconviction counsel before the Wisconsin Su‐ preme Court ruled on it. That voluntary dismissal effected the same outcome as not filing a petition in the first place—the Wisconsin Supreme Court never evaluated his claim for inef‐ fective assistance of postconviction counsel. As a result, Wil‐ son’s claim for ineffective assistance of postconviction counsel is procedurally defaulted. See Johnson v. Foster, 786 F.3d 501, 504–05 (7th Cir. 2015) (holding that defendant’s failure to file a petition for review with the Wisconsin Supreme Court vio‐ lated the complete round of review rule). Without an entire round of state‐court review, Wilson procedurally defaulted his claim. III Where, as here, a petitioner’s claims are procedurally de‐ faulted, federal habeas review is precluded unless the pris‐ oner demonstrates either of two things. Coleman v. Thompson, 501 U.S. 722, 750 (1991). The petitioner may demonstrate “cause for the default and actual prejudice as a result of the alleged violation of federal law,” or he may “demonstrate that 18 No. 21‐1402

failure to consider the claims will result in a fundamental mis‐ carriage of justice.” Id. Moreover, “[t]he miscarriage of justice exception ‘applies only in the rare case where the petitioner can prove that he is actually innocent of the crime of which he has been convicted.’” Blackmon v. Williams, 823 F.3d 1088, 1099 (7th Cir. 2016) (quoting McDowell, 737 F.3d at 483); see also Sawyer v. Whitley, 505 U.S. 333, 339 (1992). Wilson does not allege cause and prejudice,8 so we focus on the actual inno‐ cence exception. Wilson maintains that he has made a sufficient showing of actual innocence and urges us to review the merits of his claims. First, he suggests that the State has already admitted that statements made during state court proceedings would have given a jury reasonable doubt, and thus conceded the question of actual innocence. Wilson further asserts that Wal‐ lace’s testimony is sufficiently compelling and thus “there can be little doubt that [he] has satisfied the actual innocence standard.” More precisely, Wilson contends that the Wallace testimony is persuasively exculpatory and that Smith‐Cur‐ rin’s preliminary hearing statements corroborate Wallace’s account. He also tries to downplay the probative force of the inculpatory record evidence. The State responds that the Wallace evidence—including when considered with the rest of the trial evidence—falls short of sufficiently establishing actual innocence. It contends Wallace’s testimony is uncorroborated and in tension with other testimonial and physical evidence. It also highlights

8 At oral argument Wilson’s counsel informed us that Wilson was not

pursuing relief on a cause‐and‐prejudice theory. See Oral Arg. at 6:12–7:02. No. 21‐1402 19

that, even if true, Wallace’s account does not technically rule Wilson out as a potential gunman. We start with whether the State conceded that Wilson has made a sufficient showing of actual innocence. Wilson is cor‐ rect that the State previously admitted he was entitled to a hearing on the newly discovered evidence. After Wilson filed a pro se motion about that evidence, both the state trial and appellate courts declined his request for a hearing. Wilson ap‐ pealed to the Wisconsin Supreme Court, and on that court’s direction, the State filed a response conceding that Wilson was entitled to a hearing. Specifically, the State admitted it was “reasonably probable that if a jury were to find Wallace cred‐ ible, her testimony would create a reasonable doubt about whether Wilson was the shooter.” Even so, the federal standard for a showing of actual inno‐ cence demands more than what the State conceded. When we evaluate an actual innocence claim for purposes of federal ha‐ beas review, the appropriate question is whether “it is more likely than not that no reasonable juror would have convicted [Wilson] in the light of the new evidence.” Schlup v. Delo, 513 U.S. 298, 327 (1995). “To be credible, such a claim requires pe‐ titioner to support his allegations of constitutional error with new reliable evidence—whether it be exculpatory scientific evidence, trustworthy eyewitness accounts, or critical physi‐ cal evidence—that was not presented at trial.” Id. at 324. The burden rests on the petitioner to make the requisite showing. Id. at 327. This is a more demanding standard than what is required to merit a hearing. The State’s concession that Wil‐ son was entitled to a state‐court evidentiary hearing does not also serve as an admission that Wilson has shown actual in‐ nocence. Language from Schlup clarifies this point. There, the 20 No. 21‐1402

Supreme Court explained that “[t]he meaning of actual inno‐ cence … does not merely require a showing that a reasonable doubt exists in the light of the new evidence, but rather that no reasonable juror would have found the defendant guilty.” Id. at 329. So, the State did not concede the question of actual innocence. We further hold that Wallace’s testimony does not suffi‐ ciently establish Wilson’s actual innocence. At the outset, we acknowledge that this evidence is both new and credible, which are predicate requirements for the actual innocence gateway. Id. at 324. The evidence is new because it was not presented at Wilson’s trial, and it is credible because the Wis‐ consin Court of Appeals found that Wallace’s testimony was generally worthy of belief. In this appeal, the State also recog‐ nizes as much. Yet the presentation of new and credible evidence does not automatically satisfy the Schlup standard for actual innocence. Instead, the new evidence must be considered along with the existing evidentiary record. “In applying this standard, we must consider all the evidence, both old and new, incriminat‐ ing and exculpatory, without regard to whether it would nec‐ essarily be admitted at trial.” Blackmon, 832 F.3d at 1101 (citing House v. Bell, 547 U.S. 518, 538 (2006)). From there, we make a probabilistic determination about what reasonable jurors would do. House, 547 U.S. at 538. The requisite probability is established only if Wilson shows that “it is more likely than not that no reasonable juror would have convicted him in the light of the new evidence.” Schlup, 513 U.S. at 327. Finally, we always keep in mind that the “Schlup standard is demanding and permits review only in the ‘extraordinary’ case.” House, 547 U.S. at 538 (quoting Schlup, 513 U.S. at 327); see also No. 21‐1402 21

McQuiggin v. Perkins, 569 U.S. 383, 401 (2013) (“We stress once again that the Schlup standard is demanding.”). Adhering to the rigor of the Schlup standard for actual in‐ nocence, we cannot say that the Wallace evidence is so com‐ pelling and unequivocal that no reasonable juror would have convicted Wilson in the light of it. Wallace’s testimony just adds a new voice to a highly complex, and often inculpatory, evidentiary record. For instance, both Smith‐Currin and King still unequivocally identified Wilson as the gunman and de‐ scribed him emerging from an alleyway and opening fire. A reasonable juror could credit their testimony as honest and compelling—especially since a detective testified that the lo‐ cation of the .40 bullet casings was generally consistent with a shooter coming from the alleyway. A reasonable juror could likewise find the State’s impeach‐ ment evidence of Samantha Coats and Sanntanna Ross— which included Coats’ prior identification of Wilson as the gunman during a photo lineup—persuasive. Plus, Wallace’s testimony and the physical evidence do not foreclose the ex‐ istence of multiple shooters. Wallace testified she heard mul‐ tiple guns firing, and detectives recovered two different sets of bullet casings. She explained “[i]t wasn’t like it was just one gun. Like you could hear different guns going off. It wasn’t like just one person shooting outside.” So, a reasonable juror could consider Wallace’s testimony and still find that Wilson was one of two (or more) shooters. Plus, no other witness’s account of the shooting matches Wallace’s. The closest corrob‐ oration of Wallace’s version comes from Smith‐Currin’s pre‐ liminary hearing statement, in which he testified that people thought he was shooting. But that advances the ball little, 22 No. 21‐1402

because Wallace is still the only identified witness to accuse Smith‐Currin of being the gunman. The discrepancies in testimony do not end there. As men‐ tioned, witnesses provided varied accounts of the shooting and the shooter. Whether it is the gunman’s height, hair, or clothing, the witnesses’ recollections differed. Reasonable ju‐ rors could draw different conclusions from this evidence. Even with Wallace’s testimony, we are left with a series of competing eyewitness accounts. When evaluating a claim of actual innocence, our role “is not to make an independent fac‐ tual determination about what likely occurred, but rather to assess the likely impact of the evidence on reasonable jurors.” House, 547 U.S. at 538. As the dissent emphasizes, a state court found Wallace’s testimony to be credible. But that finding does not mean that a reasonable juror would necessarily credit Wallace’s account of the shooting over that of any other witness, such as Smith‐Currin or King. A conflict between trial testimony remains, notwithstanding Wallace’s credibil‐ ity, and we cannot say that it is more likely than not that no reasonable juror would have convicted Wilson in the light of the new evidence. Our conclusion accords with relevant precedent. In Black‐ mon, the court heard competing eyewitness testimony. 823 F.3d 1088. There, two gunmen approached a victim and opened fire. Id. at 1092. The ensuing bench trial focused on the identity of the second gunman, and eyewitness testimony was paramount. Id. at 1092, 1095–96. Approximately two months after the shooting, two eyewitnesses identified Blackmon as one of the triggermen through photo lineups and in‐person lineups. Id. at 1094. Those same witnesses identified Black‐ mon as the gunman at trial. Id. at 1093–95. In response, No. 21‐1402 23

Blackmon called three defense witnesses. Two of those wit‐ nesses provided an alibi for Blackmon; the third claimed to have watched the shooting and testified that Blackmon was not present at the scene. Id. at 1095–96. The presiding judge determined that Blackmon was one of the shooters and found him guilty. Id. at 1096. Like Wilson, Blackmon challenged his conviction through federal habeas and tried to pass through the actual innocence gateway for certain defaulted claims. Id. at 1100–01. To that end, Blackmon provided two new eyewitness affidavits. Id. at 1097. Each of the new witness affidavits claimed that Black‐ mon was not one of the gunmen. Id. Reviewing all the evi‐ dence—old and new—this court concluded that Blackmon’s showing of actual innocence was insufficient. Id. at 1101–02. In reaching that conclusion, this court noted that the new evidence merely contrasted with the State’s two credible eye‐ witness accounts. Id. And the new eyewitnesses did not come forward until eight years after the shooting. Id. at 1102. So, the “balance between inculpatory and exculpatory witnesses [was] not enough to meet the demanding Schlup standard for actual innocence.” Id. The facts here track those in Blackmon. Like Blackmon, Wil‐ son offers new eyewitness testimony into a factual record oc‐ cupied by contrasting eyewitness statements. But as ruled in Blackmon, the introduction of new eyewitness testimony does not amount to a showing of actual innocence when strong and credible testimony to the contrary remains. Just as the two new affidavits in Blackmon merely added to the balance of in‐ culpatory and exculpatory evidence, so too does Wallace’s testimony. Even with the Wallace evidence, we are left with a 24 No. 21‐1402

complex factual record pointing in different directions.9 We therefore hold that Wilson has not satisfied the Schlup stand‐ ard for actual innocence. Other cases from this court also sup‐ port our conclusion. See, e.g., Smith v. McKee, 598 F.3d 374, 387–88 (7th Cir. 2010) (concluding insufficient showing of ac‐ tual innocence where petitioner’s two new affidavits did not sufficiently counter the state’s evidence, which included two eyewitness identifications and a self‐inculpatory statement); Hayes v. Battaglia, 403 F.3d 935, 937–38 (7th Cir. 2005) (holding that a draw between the number of eyewitnesses for and against defendant—six new exculpatory witnesses versus the state’s six inculpatory trial witnesses—“cannot establish that no reasonable factfinder would have found the applicant guilty”) (cleaned up). Finally, Jones v. Calloway, 842 F.3d 454 (7th Cir. 2016), is in‐ structive as a rare case where we concluded that the defend‐ ant had made a sufficient showing of actual innocence. Jones was convicted of murder and sought federal habeas relief. The district court held his claims procedurally defaulted, forcing Jones to rely on the actual innocence gateway to excuse his default. Id. at 459. The new evidence Jones brought to bear on his case was exceptional. Michael Stone, another man present at the murder scene, provided new testimony that he was the lone shooter. Id. at 460. And his testimony was compelling.

9 The dissent observes that, unlike in Blackmon, 823 F.3d at 1093, the

inculpatory witnesses here knew Wilson before the shooting. For our dis‐ senting colleague, that prior knowledge dilutes the weight of the photo lineup identifications by King and Smith‐Currin. But Wallace was not a stranger to Wilson or Smith‐Currin, either. Indeed, at the evidentiary hear‐ ing Wallace testified she had been around Smith‐Currin “plenty of times” before the shooting, and Wilson helped set up the music at Wallace’s apartment on the night of the crime. No. 21‐1402 25

Stone had previously turned himself in for the crime, con‐ fessed to the shooting within days, identified the murder weapon, and given testimony that was consistent with the case’s forensic evidence. Id. at 462. Stone’s story of the shoot‐ ing had also remained consistent for over a decade. Id. at 463. The district court found a sufficient showing of actual inno‐ cence, and this court agreed. Id. at 460, 462. In Jones, the new witness took the stand and personally claimed sole responsibility for the crime. Id. at 462. His testi‐ mony was consistent with the physical evidence as well, whereas the testimony of prosecution witnesses in that case was often in tension with the forensics. Id. The Wallace evi‐ dence is not so forceful. Her eyewitness testimony merely contrasts with that of Smith‐Currin and King (and to a lesser degree, Coats and Ross). Reviewing all the facts, a reasonable juror could still conclude that Wilson was the shooter. Ac‐ cordingly, Wilson has not sufficiently shown actual inno‐ cence. IV Given the unexcused procedural default, we do not reach the merits of Wilson’s ineffective assistance of trial and post‐ conviction counsel claims. In summary, Wisconsin state courts disposed of Wilson’s ineffective assistance of trial counsel claim on adequate and independent state grounds, and he failed to present his inef‐ fective assistance of postconviction counsel claim for one complete round of state court review. So, both of his claims are procedurally defaulted. Wilson attempts to overcome these defaults, but he fails to make a sufficient showing of ac‐ tual innocence. Even considering Wallace’s testimony, we 26 No. 21‐1402

cannot conclude that it is more likely than not that no reason‐ able juror would have convicted Wilson. The Schlup standard for actual innocence is high and reserved for the exceptional case, a threshold Wilson does not clear here. For these reasons, the district court’s denial of Wilson’s pe‐ tition for federal habeas relief is AFFIRMED. No. 21‐1402 27

HAMILTON, Circuit Judge, dissenting. During post‐convic‐ tion hearings in the state courts, Lakisha Wallace testified that the shooter was actually Antwan Smith‐Currin, who was also the state’s chief witness against petitioner Wilson. Ms. Wal‐ lace witnessed the incident from the bottom floor of the du‐ plex where she lived downstairs from Smith‐Currin. She tes‐ tified that she heard Smith‐Currin yell to his brother to give him a gun and then saw Smith‐Currin wave a handgun on the front porch of the duplex, open fire, and run into the crowd while shooting. According to Ms. Wallace, Smith‐Currin im‐ mediately came back inside and shouted to his brother that he had “just offed” someone. Ms. Wallace further testified that in the days after the shooting, she heard Smith‐Currin say that he planned to blame the crime on Wilson. She also offered a plausible motive for the plan to blame Wilson. Smith‐Currin had seen his girlfriend with Wilson on the duplex porch the day before the shooting and was angry about them being to‐ gether. The extraordinary feature of this habeas case is the combi‐ nation of two facts. First, the state agreed during state court proceedings that “[i]t is reasonably probable that if a jury were to find Ms. Wallace credible, her testimony would create a reasonable doubt about whether Wilson was the shooter.” Second, when Ms. Wallace actually testified before a state court judge, that judge found her credible. Under these unu‐ sual circumstances, and given other significant weaknesses in the state’s case, we should find that Wilson has made a show‐ ing of innocence sufficient to excuse his procedural default. We should remand to the district court for an evidentiary hearing on his claims of ineffective assistance of counsel. 28 No. 21‐1402

My colleagues and I agree on all but that one decisive is‐ sue. As the majority opinion explains, under Wisconsin’s un‐ usual procedures for post‐conviction relief, Wilson had a fed‐ eral constitutional right to effective assistance of counsel in post‐trial proceedings under Wisconsin Statute § 974.02. Ante at 16, citing Lee‐Kendrick v. Eckstein, 38 F.4th 581, 587 (7th Cir. 2022). We also agree that Wilson procedurally defaulted his ineffective assistance claims in the state courts. Ante at 17. Where we disagree is whether Wilson has shown “actual in‐ nocence” so as to excuse his procedural default. To avoid the consequences of his procedural default, Wil‐ son offers the testimony of Lakisha Wallace to show that he is actually innocent. See generally Sawyer v. Whitley, 505 U.S. 333, 339 (1992); Blackmon v. Williams, 823 F.3d 1088, 1099 (7th Cir. 2016). To do so, Wilson must come forward with new evidence showing “it is more likely than not that no reasona‐ ble juror would have convicted him in the light of the new evidence.” Schlup v. Delo, 513 U.S. 298, 327 (1995); see also McQuiggin v. Perkins, 569 U.S. 383, 386, 390 (2013). His evi‐ dence must be reliable and may take the form of “exculpatory scientific evidence, trustworthy eyewitness accounts, or critical physical evidence.” Gladney v. Pollard, 799 F.3d 889, 896 (7th Cir. 2015) (emphasis added), quoting Schlup, 513 U.S. at 324. In applying this test, it is essential to remember that the hypothetical jurors would have to examine all the new and old evidence and be convinced of guilt beyond a reasonable doubt. That was, after all, the point of the Supreme Court’s de‐ cision in the canonical Jackson v. Virginia, 443 U.S. 307, 318–21 (1979) (issue in federal habeas review was not whether “any evidence” supported the state conviction but whether evi‐ dence could support finding of guilt beyond a reasonable No. 21‐1402 29

doubt). The Supreme Court has rephrased the relevant stand‐ ard (“to remove the double negative”) as requiring new evi‐ dence making it “more likely than not [that] any reasonable juror would have reasonable doubt.” House v. Bell, 547 U.S. 518, 538 (2006). As I read this record, including Ms. Wallace’s testimony credited by the state court, there is some evidence to support a finding of guilt, but, per Jackson v. Virginia and House v. Bell, any reasonable juror would have a reasonable doubt once Ms. Wallace’s testimony is added to the mix. As the majority opinion presents the facts, Wilson’s trial for the fatal shooting of Melvin Williams presented testimony from four eyewitnesses who identified Wilson as the shooter. From that premise, the majority opinion relies on a portion of our decision in Blackmon where we held that new exculpatory testimony from two eyewitnesses was not enough to over‐ come procedural default. 823 F.3d at 1102. The key to that por‐ tion of Blackmon was that Blackmon had been identified as one of two killers independently, and consistently, by two utterly neutral witnesses. Id. at 1101–02.1 The case here was far shakier. No witness consistently iden‐ tified Wilson as the shooter. The two government witnesses who identified Wilson at trial spoke to police on the night of the shooting. Both knew Wilson at the time. They did not claim

1 We remanded Blackmon for an evidentiary hearing on other grounds,

namely his claim that counsel was ineffective in failing to investigate ade‐ quately his alibi defense. 823 F.3d at 1104–07. After remand, Mr. Blackmon won habeas relief on that basis. Blackmon v. Pfister, 2018 WL 741390 (N.D. Ill. Feb. 7, 2018). 30 No. 21‐1402

that night that they saw Wilson was present, let alone shoot‐ ing.2 That night, Shakira King told police that she had heard an‐ other woman claiming Wilson was the shooter. By the time of trial, however, King’s story had changed. She testified that she herself saw Wilson shooting, and she denied having told an officer on the night of the shooting that it was her friend who claimed to have recognized the shooter as Wilson. King’s trial testimony also contradicted her contemporaneous descrip‐ tion of the shooter’s hairstyle. Her description of the shooter’s clothing did not match that given by any other witness. And at trial King denied being part of the fight that preceded the shooting, though she had previously admitted involvement to police and other witnesses had confirmed her part in the melee. Moving to Smith‐Currin, he did not tell police that he saw Wilson shooting until a month after the crime. On the night of the shooting, Smith‐Currin spoke with police but did not mention Wilson. Smith‐Currin’s trial testimony describing what he saw the shooter wearing was inconsistent. And at a preliminary hearing, Smith‐Currin even testified that some people claimed they had seen him shooting from the duplex’s porch.

2 The fact that both witnesses knew Wilson prior to the shooting is

important. The majority opinion states correctly that King and Smith‐Cur‐ rin identified Wilson as the gunman out of photo lineups. This procedure seems to add credibility to the identifications but its weight is diluted by the fact that both already knew him. No. 21‐1402 31

The two other witnesses who the state argued had previ‐ ously identified Wilson as the shooter strongly refuted or re‐ canted such statements at trial. Sanntanna Ross told the jury that what police construed as her identifying Wilson as the shooter was simply her indicating that she knew Wilson. When asked on the stand whether she saw Wilson shooting, Ross unequivocally said no. Samantha Coats told the jury that her prior identification of Wilson as the shooter was based only on rumors. When Coats was pressed for an identification by police during the investigation, she said, her boyfriend was in custody and she had been threatened with arrest herself. She chose Wilson (whom she knew and recognized) in a photo lineup to avoid arrest and in the hope that the police would release her boyfriend. Unlike the Blackmon case, Wilson has also offered new ev‐ idence that not only exonerates him but identifies a different shooter, the state’s chief witness. In applying the Schlup stand‐ ard, which may be met by “trustworthy eyewitness ac‐ counts,” keep in mind that the state judge who heard Ms. Wal‐ lace testify, subject to lengthy cross‐examination, credited her testimony. If a jury heard all the trial evidence and Ms. Wallace’s tes‐ timony, there would of course still be the trial testimony of Smith‐Currin and King identifying Wilson as the shooter. That’s “some evidence”—but that low bar was the standard rejected in Jackson. Given the problems with their testimony— including their delayed identifications of a person they knew as the shooter they claimed to have seen that night—the lack of any other evidence placing Wilson at the scene, and the 32 No. 21‐1402

consistent and credible testimony of Ms. Wallace, a conscien‐ tious juror could not reasonably find Wilson guilty beyond a reasonable doubt. The majority opinion also suggests that Ms. Wallace’s tes‐ timony does not necessarily exculpate Wilson because there might have been more than one shooter. Perhaps both Smith‐ Currin and Wilson, and even others, were armed and fired shots? The principal problem with this possibility is that it would make it even harder to convince a jury beyond a rea‐ sonable doubt that Wilson was the one who shot the victims. The state prosecuted Wilson on the theory that there was one shooter and that he was the one. The new, more complex, and untested theory of multiple shooters does not offer a solid ba‐ sis for denying relief. The test for actual innocence is demanding, and cases of proven actual innocence are relatively rare. In my view, this is one of those rare cases. I am not saying that Wilson is enti‐ tled to a new trial based on his as‐yet‐unproven claims of in‐ effective assistance of counsel. But I believe he is entitled to a hearing to try to prove them. I respectfully dissent.