NEAL v. UNITED STATES (2025)

Tyree M. Neal, Jr. pleaded guilty to conspiracy to distribute cocaine in violation of 21 U.S.C. §§ 841(a)(1), (b)(1)(C), and 846. At his sentencing, the district court found that Neal's prior Illinois cocaine conviction supported a recidivism enhancement under 21 U.S.C. § 841(b)(1)(C) that raised his maximum sentence from 20 years to 30 years' imprisonment. The district court then sentenced Neal to the 30-year maximum. Neal appealed and argued that he should have been allowed to withdraw his guilty plea because it was involuntary and lacked a factual basis; his appeal did not challenge the enhancement. We ruled against the arguments Neal raised on direct appeal. United States v. Neal, 907 F.3d 511 (7th Cir. 2018).

Two years after Neal's appeal, in United States v. Ruth, 966 F.3d 642 (7th Cir. 2020), we applied the “categorical approach” and held that because Illinois's definition of cocaine is broader than the federal definition, an Illinois conviction for cocaine delivery could not support a recidivism enhancement under 21 U.S.C. § 841(b)(1)(C). Accordingly, if Neal were sentenced today, his maximum sentence would be 20 years.

In the proceedings below, Neal sought Section 2255 relief for ineffective assistance of counsel, arguing that his appellate, sentencing, and plea counsel were all constitutionally deficient in failing to raise the argument that succeeded in Ruth. The district court denied his petition. We affirm.

I. Background

A. Neal's Conviction, Sentencing, and Appeal

In 2014, after Neal was indicted on one count of conspiracy to distribute cocaine in violation of 21 U.S.C. §§ 841(a)(1), (b)(1)(C), and 846, he evaded arrest and led officers on a 17-mile car chase at sustained speeds exceeding 100 miles per hour. Neal eventually abandoned his vehicle and carjacked a mother who was in a van with her three children. Neal forced her to drive to another vehicle that was waiting for him. With the help of that getaway car, Neal escaped from police pursuit. He was finally arrested in February 2015.

Following Neal's arrest, the government filed a Section 851 notice seeking a recidivism enhancement under 21 U.S.C. § 841(b)(1)(C) because Neal had a 2007 conviction for unlawful delivery of cocaine under 720 ILCS 570/401(c)(2). If the district court agreed that Neal had a “prior conviction for a felony drug offense,” it would increase Neal's statutory maximum term of imprisonment from 20 years to 30 years. 21 U.S.C. § 841(b)(1)(C).

In August 2016, Neal pleaded guilty. During plea negotiations, he was represented by four successive attorneys, three of which withdrew due to disagreements with Neal. The final attorney was Heather Winslow, who recommended a guilty plea.

At the sentencing stage, Neal fired Winslow because they disagreed about how to address a potential career-offender classification. The district court appointed J. Christian Goeke in February 2017. Neal attempted to hire private counsel, but Goeke ultimately represented him through the end of sentencing.

As relevant to his sentencing, Neal had an extended criminal history that included convictions for 15 other offenses, including aggravated battery. And in March 2016, a jailhouse informant told authorities that Neal had said he paid someone to kill a witness. The Probation Office's Presentence Investigation Report (prepared using the 2015 United States Sentencing Guidelines) concluded that Neal was subject to an offense level of 35, criminal history category VI, and a 292–365 month sentencing range.¹ It found that Neal had accepted responsibility, but that enhancements applied for the drug amount, the use of violence, obstruction of justice, and reckless endangerment.

With the help of an investigator to interview witnesses, Goeke contested the use of violence, obstruction of justice, reckless endangerment, and career offender enhancements. However, Goeke did not object to the Section 841 recidivism enhancement, nor did he discuss such an objection with Neal.

After a two-day sentencing hearing on September 18 and 19, 2017, the district court found that the recidivism enhancement applied, as well as the other challenged enhancements. It sentenced Neal to the statutory maximum, 30 years.

Neal appealed on September 25, 2017 and Assistant Federal Public Defender Johanna Christiansen was appointed to represent him. Christiansen argued that the district court erred in not allowing Neal to withdraw his guilty plea. Neal, 907 F.3d at 514–16. She asserted that Neal did not understand the conspiracy charge, making the plea involuntary, and that the plea was not supported by a sufficient factual basis. Id. On October 22, 2018, about three weeks after oral argument, our court rejected these contentions and affirmed Neal's conviction. Id. at 516–17. Christiansen did not argue against the Section 841 recidivism enhancement, nor did she tell Neal that any such argument existed.

B. The Categorical Approach to Recidivism Enhancements

To determine if a Section 841 recidivism enhancement applies, we use the “categorical approach” that looks solely to the elements of a defendant's offense and not to the specific facts underlying the conviction. Ruth, 966 F.3d at 646; see Taylor v. United States, 495 U.S. 575, 110 S.Ct. 2143, 109 L.Ed.2d 607 (1990). “If, and only if, the elements of the state law mirror or are narrower than the federal statute can the prior conviction qualify as a predicate felony drug offense.” Ruth, 966 F.3d at 646 (quoting United States v. De La Torre, 940 F.3d 938, 948 (7th Cir. 2019)).

Our use of the categorical approach in assessing challenges to recidivism enhancements under Section 841 originated in 2018, contemporaneous with Neal's direct appeal. We first encountered “the question of whether the ․ categorical approach applies to the term ‘felony drug offense’ for the purpose of” Section 841 in Brock-Miller v. United States, where in April 2018, we reserved an answer “for another day.” 887 F.3d 298, 307 (7th Cir. 2018). That day came four months later in August 2018, when we adopted the categorical approach for Section 841 enhancements in United States v. Elder, 900 F.3d 491 (7th Cir. 2018). Because the defendant in Elder was previously convicted under an Arizona statute that swept more broadly than federal law by including additional substances (propylhexedrine and scopolamine) in its definition of “dangerous drugs,” that conviction could not serve as a predicate offense for the Section 841 enhancement. Id. at 501.

We decided our next categorical approach case in 2019, after Neal's appeal concluded. In United States v. De La Torre, we observed that both Indiana and federal law regulated methamphetamine and its isomers—an isomer shares the same number and type of atoms but with a distinct arrangement in space, which can give rise to different effects when used. 940 F.3d at 951. There are three relevant types of isomer: optical, positional, and geometric. Id. Consequently, despite a “seeming match” at first glance, “because the federal definition of methamphetamine includes only its optical isomers whereas the Indiana definition includes something more than just optical isomers of methamphetamine, the mismatch renders the Indiana statute overbroad” for purposes of supplying a predicate offense under Section 841. Id. We further stated that “our decision [was] not solely dependent on the definition of methamphetamine” because Indiana law also included two other substances (parahexyl and Telazol) not covered under federal law. Id. at 952.

That brings us to United States v. Ruth, a key case for this appeal.² In that 2020 decision, we remarked that the Illinois statutory definition of cocaine included “positional isomers” of cocaine, whereas the federal definition includes only “optical isomers” and “geometric isomers.” Ruth, 966 F.3d at 647. “On its face, then, the Illinois statute is categorically broader than the federal definition.” Id. We were uninterested in the government's contention that positional isomers of cocaine did not actually exist in the drug trade because “it is not the province of the judiciary to rewrite Illinois's statute to conform to a supposed practical understanding of the drug trade.” Id. at 648. That led to the “inescapable conclusion that the plain language of the state statute categorically covers a larger swath of conduct than its federal counterpart.” Id. Thus, the defendant's Illinois conviction under 720 ILCS 570/401(c)(2) was not a predicate “felony drug offense” that triggered Section 841(b)(1)(C)'s sentencing enhancement. Id. at 650.

C. Neal's Petition

This proceeding began when Neal filed a timely 28 U.S.C. § 2255 motion to vacate, set aside, or correct his sentence. Neal asserted that his appellate (Christiansen), sentencing (Goeke), and plea (Winslow) counsel each provided ineffective assistance in violation of the Sixth Amendment. According to Neal, all three attorneys were ineffective because they failed to argue that Neal's Illinois cocaine conviction did not qualify as a predicate “felony drug offense” supporting a recidivism enhancement under Section 841(b)(1), which we hereafter call the Ruth argument.

The district court denied Neal's claims of ineffective assistance from Goeke and Winslow and ordered an evidentiary hearing on Christiansen's performance. The district court acknowledged that if Neal had been sentenced after Ruth, his statutory maximum sentence would have been a decade shorter than his actual sentence. Still, the district court concluded that Goeke and Winslow were not deficient for failing to anticipate a change in the law.³

For appellate counsel Christiansen, the district court observed that Ruth was preceded by several cases (especially Elder and De La Torre) that applied the categorical approach to determine whether state drug offenses were “felony drug offenses” under Section 841(b)(1). Because Neal's appeal was closer in time to that wave of cases than his sentencing and plea, the district court said that it was “possible that a reasonable appellate counsel may have been required, at a minimum, to consider the possibility of making the [Ruth] argument in Neal's appeal.” The district court ordered an evidentiary hearing into whether Christiansen considered the Ruth argument for Neal's appeal and, if she did, why she chose not to raise it.

At the hearing, Christiansen testified that she initially considered making a categorical challenge to the use of Neal's Illinois cocaine conviction as a predicate for the recidivism enhancement. But she did not compare the Illinois and federal definitions of cocaine, so she failed to identify the isomer statutory mismatch. Christiansen stated that if she had gone deeper into the statute, the overbreadth would have been “obvious” to her. Christiansen added that the review would be for plain error and that she would probably not raise an issue of first impression under plain error. Notwithstanding that caveat, Christiansen accepted that she “should have” made the argument.

The district court held that Christiansen did not provide ineffective assistance of counsel. It stated that while Christiansen had the “building blocks” to make the Ruth argument in Neal's appeal, “they had not been brought together such that a competent attorney must have recognized their significance.” And the Ruth argument “was neither obvious nor clearly stronger than the arguments ․ raised on appeal, especially considering it would have faced plain error review.”

Although the district court held that Christiansen's performance was not deficient, it addressed the prejudice prong “so that, if the Court of Appeals disagrees with [its] performance assessment, the Court's prejudice assessment will be clear.” The district court determined that Neal had shown prejudice because there was a reasonable chance that Christiansen could have successfully argued that it was plain error to apply the recidivism enhancement based on Neal's Illinois cocaine conviction, and if so, Neal's sentence would have been ten years shorter.

The district court granted a certificate of appealability on Neal's ineffective assistance claim regarding Christiansen, and the Seventh Circuit expanded the certificate of appealability to include Neal's claims that Goeke and Winslow were also ineffective.

II. Discussion

Section 2255 entitles a federal prisoner to collateral relief if his sentence “was imposed in violation of the Constitution or laws of the United States, or [if] the court was without jurisdiction to impose such sentence, or [if] the sentence was in excess of the maximum authorized by law․” 28 U.S.C. § 2255(a). This “extraordinary remedy” is available only in limited situations. Almonacid v. United States, 476 F.3d 518, 521 (7th Cir. 2007).

When a district court denies a Section 2255 petition, we review legal conclusions de novo and factual findings for clear error, and its decision to deny an evidentiary hearing for abuse of discretion. Bridges v. United States, 991 F.3d 793, 799 (7th Cir. 2021). “A decision that rests on an error of law is always an abuse of discretion.” Id.

The Sixth Amendment establishes the right to effective assistance of counsel. U.S. Const. amend. VI; Strickland v. Washington, 466 U.S. 668, 686, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). Under Strickland's test for ineffective assistance of counsel, a petitioner will not receive collateral relief unless he can establish that (1) his counsel's “performance was deficient” and (2) “the deficient performance prejudiced the defense.” Strickland, 466 U.S. at 687, 104 S.Ct. 2052. Counsel's performance is not deficient unless it falls “below an objective standard of reasonableness.” Id. at 688, 104 S.Ct. 2052.

The general Strickland standard also governs claims of ineffective assistance of appellate counsel, “but with a special gloss when the challenge is aimed at the selection of issues to present on appeal.” Makiel v. Butler, 782 F.3d 882, 897 (7th Cir. 2015). Because appellate advocacy involves “careful selection” of a small number of issues for appeal, in such challenges “appellate counsel's performance is deficient under Strickland only if she fails to argue an issue that is both ‘obvious’ and ‘clearly stronger’ than the issues actually raised.” Id. at 897–98 (quoting Brown v. Finnan, 598 F.3d 416, 425 (7th Cir. 2010)).

A. Ineffective Assistance of Counsel Claims After Ruth

This is the third time we address a claim for ineffective assistance of counsel related to our adoption of the categorical approach for Section 841 enhancements. In Harris v. United States, a Section 2255 petitioner claimed ineffective assistance from his counsel who represented him during plea negotiations and sentencing in 2017. 13 F.4th 623, 626–27 (7th Cir. 2021). The petitioner argued that his counsel was deficient because he did not use the categorical approach to argue that a prior Indiana cocaine conviction was not a predicate “felony drug offense.” Id. In assessing the petition, we cautioned that “[i]n some circumstances, defense counsel may be required to anticipate arguments foreshadowed but not yet adopted by existing case law.” Id. at 629 (quoting Bridges, 991 F.3d at 802). We noted that the “categorical approach was well-established by 2017” and the Ninth Circuit “had already applied the categorical approach to the federal ‘felony drug offense’ definition.” Id. (citing United States v. Ocampo-Estrada, 873 F.3d 661, 667 (9th Cir. 2017)). Plus, “comparing statutory definitions is part of competent representation.” Harris, 13 F.4th at 629. We concluded that “[a]t the time of Harris's sentencing, his counsel should have known about a possible categorical challenge to the predicate offenses.” Id.

Nevertheless, we affirmed the denial of the petition. “At the time of sentencing, no case law held that any state defined isomers more broadly than the federal government.” Id. at 630. While it was unclear whether the petitioner's counsel considered a categorical approach argument based on the statutory definition of isomers, we held that either way, it was objectively reasonable for counsel to pursue a plea deal over a “novel challenge to the predicate offenses with the risk of a mandatory life sentence.” Id. at 631.

Our decision in Coleman v. United States went further in describing counsel's obligations. 79 F.4th 822 (7th Cir. 2023).⁴ In Coleman, much like Neal, the petitioner argued that he received ineffective assistance because his counsel did not make the argument that succeeded in Ruth. Id. at 831. More specifically, though, Coleman involved counsel's 2014 failure to challenge a sentence enhancement that used a prior Illinois cocaine conviction to increase the mandatory minimum to life imprisonment. Id. In response to the petitioner's argument that this omission was deficient performance, we advised that “the groundwork for such an argument was, at the very least, foreshadowed by numerous decisions issued before 2014 ․ that applied the categorical approach to predicate offenses in other contexts.” Id. So we were “unpersuaded by the government's argument that application of the categorical approach to cocaine isomers was too novel in 2014 to have been recognized by competent defense counsel.” Id. at 832.

We thus held that “it would have been objectively unreasonable for Coleman's defense counsel to have not even considered a categorical challenge to the government's reliance on prior Illinois cocaine convictions to enhance Coleman's sentence.” Id. (emphasis in original). If proven, that claim “would entitle [Coleman] to relief.” Id. “Of course, if counsel did consider the argument but had credible strategic reasons for not raising it, that would be a different question.” Id. We remanded to the district court for an evidentiary hearing on whether counsel “considered the possibility of a categorical challenge” and if he did, “what reasons he had for not raising it.” Id.

Neal argues that Coleman controls his ineffective assistance claim and entitles him to relief. After all, Coleman instructs that by 2014, an attorney encountering a Section 841 enhancement based on an Illinois cocaine conviction should have considered making a categorical challenge based on the inclusion of positional isomers in the state definition of cocaine.⁵

But Neal also takes Coleman to say that an attorney who failed to consider such a challenge necessarily provided constitutionally ineffective assistance. We do not read Coleman so broadly. Neal's interpretation takes Coleman out of context and is misaligned with our Strickland jurisprudence.

“Strickland ․ calls for an inquiry into the objective reasonableness of counsel's performance, not counsel's subjective state of mind.” Harris, 13 F.4th at 630 (quoting Harrington v. Richter, 562 U.S. 86, 110, 131 S.Ct. 770, 178 L.Ed.2d 624 (2011)). Under that framework, counsel's inadvertent failure to raise an issue does not automatically trigger relief. Id. (citing Yarborough v. Gentry, 540 U.S. 1, 8, 124 S.Ct. 1, 157 L.Ed.2d 1 (2003)). “Judicial scrutiny of counsel's performance must be highly deferential.” Strickland, 466 U.S. at 689, 104 S.Ct. 2052.

At first glance, one could see tension between Coleman's holding that a remand was necessary and Harris's holding that counsel's failure to make a categorical challenge was objectively reasonable and provided no grounds for habeas relief. A closer look shows the consistent reasoning underlying both cases. Coleman cites Harris approvingly and did not overrule any aspect of it. Indeed, Harris's language on objective performance comes from the Supreme Court. To show why Harris and Coleman are in harmony, we highlight Coleman's facts.

The holding in Coleman was “particularly true” because “absent the district court's application of the Section 851 enhancement, [the defendant] would not have been subject to a mandatory life sentence.” 79 F.4th at 832–33. When defeating the enhancement was necessary to prevent such a dire outcome for the client, we would expect a competent attorney to research and consider all the viable arguments in his arsenal and, absent some compelling reason, make use of a defense against a mandatory life sentence. Cf. Wilson v. Gaetz, 608 F.3d 347, 352 (7th Cir. 2010) (“Especially because of the severity of the penalty that [defendant] was facing, [counsel] should have done more[.]”). Not every recidivism enhancement case is analogous to Coleman. When a defendant is subject to discretionary sentencing, like Neal was, an attorney could reasonably choose between several strategies designed to achieve a better result for the client. Coleman's harsh binary—life or not—presents a crucial distinction from any case where a wider set of outcomes is possible. See, e.g., United States v. Liscano, No. 02 CR 719-16, 2021 WL 4413320, at *4 (N.D. Ill. Sept. 27, 2021) (“A life sentence is different in kind, and not merely degree, from nearly all other sentences. It is the end of the road.”).

Moreover, in Harris, counsel had negotiated a plea agreement that guaranteed the defendant would avoid a life sentence. Harris, 13 F.4th at 630. But if counsel rejected the plea and pursued a statutory mismatch challenge to the defendant's predicate offenses, the defendant faced a mandatory life sentence if the challenge failed. Id. Given these specific “options available to the defense”—the prospect of a mandatory life sentence should a novel argument fail—it was reasonable for counsel to advise the defendant to take the plea, and it was unnecessary for counsel to consult with his client about the then-untested mismatch argument. Id. at 630–31.

Also, Coleman (and Harris) dealt with alleged ineffective assistance at the sentencing stage. We did not need to ask, as we do for Neal's appellate counsel claim, whether the unraised challenge to the Section 841 enhancement was “clearly stronger” than the arguments that counsel did present. Makiel, 782 F.3d at 898. Neither did we confront the array of strategic considerations that counsel might make when a client is deciding whether to plead guilty or proceed to trial.

We additionally call attention to White v. United States, where we held that a defendant who did not raise a Ruth argument during his sentencing and direct appeal was procedurally defaulted from bringing that claim in a Section 2255 petition. 8 F.4th 547, 557 (7th Cir. 2021). Said differently, White declined to make Ruth “retroactive on collateral review․” United States v. Vaughn, 62 F.4th 1071, 1072 (7th Cir. 2023). Neal's reading of Coleman would effectively make Ruth retroactive to his sentence, but only because his counsel did not grasp the argument. That would be an unusual framework.

In short, we reject Neal's reading of Coleman. That decision does not say that starting in 2014, every defense attorney, at every stage, in every criminal case involving a Section 841 enhancement for a prior Illinois cocaine conviction, provided constitutionally ineffective assistance if they did not consider the Ruth argument. While Coleman identified one circumstance where it would be ineffective assistance to miss the Ruth argument, we reiterate that the inquiry is an objective assessment of “counsel's performance [ ] considering the options available to the defense.” Harris, 13 F.4th at 630.

With the appropriate test clarified, we address Neal's three ineffective assistance of counsel claims, beginning with Christiansen and then working backward chronologically through Neal's criminal case. Our analysis is restricted to the first prong of Strickland, performance.

B. Appellate Counsel Christiansen

We start with appellate counsel Christiansen because it is the more difficult part of the analysis, at least in one aspect. Christiansen represented Neal later in time, when the components of the Ruth argument were coming into clearer focus.

Unlike in Coleman, we already have the benefit of an evidentiary hearing on whether Christiansen considered making the Ruth argument. She did not. While Christiansen initially considered a categorical challenge to the enhancement, she did not contemplate that “Illinois might define cocaine differently than the federal government defined cocaine” nor did she look at the “definition of isomers.” Her view was that “cocaine is cocaine.” In Christiansen's words, the isomer argument “was either not on my radar or I completely missed it.” Because Christiansen did not realize that the Ruth argument was available, she could not have had “credible strategic reasons” for not raising it. Coleman, 79 F.4th at 832.

The government contends that the Ruth argument was not “obvious” at the time of Neal's appeal. See Makiel, 782 F.3d at 898. Coleman tells us otherwise. 79 F.4th at 832 (“As we have ‘repeatedly’ made clear, ‘comparing statutory definitions is part of competent representation,’ and application of the categorical approach to drug isomers is ‘not complex.’ ” (quoting Harris, 13 F.4th at 630)). So does Harris. 13 F.4th at 629 (“At the time of Harris's sentencing, his counsel should have known about a possible categorical challenge to the predicate offenses. ․ By 2017, when Harris was negotiating his guilty plea, the categorical approach was well-established.”). There is no need to retread that territory. The unresolved question is whether the Ruth argument was “clearly stronger” than the arguments Christiansen raised. Makiel, 782 F.3d at 898–99.

“[T]he reasonableness of counsel's performance must be assessed ‘in the context of the law’ at the time.” Harris, 13 F.4th at 629 (quoting Lilly v. Gilmore, 988 F.2d 783, 786 (7th Cir. 1993)). “Proving that an unraised claim is clearly stronger than a claim that was raised is generally difficult ‘because the comparative strength of two claims is usually debatable.’ ” Makiel, 782 F.3d at 898 (quoting Shaw v. Wilson, 721 F.3d 908, 915 (7th Cir. 2013)).

We start with the strength of the unraised Ruth argument. Elder was argued (as it happens, by Christiansen herself) in May 2018 and decided in August 2018.⁶ See 900 F.3d at 491. To again situate Elder in context, our decision in Brock-Miller had planted the seeds of a categorical challenge, but our case law had not developed further. See id.; Brock-Miller, 887 F.3d at 313–14 (approving of a categorical challenge to a predicate felony regarding drug paraphernalia). Neal's appeal was roughly concurrent with Elder: Christiansen filed Neal's opening brief in May 2018 and reply brief in July 2018, argued the case in October 2018, and we decided it that same month. Although Elder likely came too late for Christiansen to have reframed Neal's appeal before oral argument, if she had made the Ruth argument to begin with, Elder—where we approved a categorical challenge to a predicate felony drug offense under Section 851—would have been potent additional authority. Whether to apply the categorical approach would be answered, and the case would come down to isomers. And, as discussed above, we accepted an isomer mismatch argument in De La Torre without additional cases after Elder. De La Torre, 940 F.3d at 948. De La Torre made Ruth's isomer analysis easier, but it was not a required building block. Finally, Christiansen testified to her concern about bringing a novel argument on plain error review, but Ruth itself was decided on plain error. 966 F.3d at 650. All things considered, if Christiansen had raised the Ruth argument in Neal's appeal, it plausibly could have succeeded and reduced his maximum sentence by ten years.

Christiansen, by contrast, sought to overturn Neal's conviction. She argued that Neal should have been allowed to withdraw his plea because it was involuntary and not supported by a sufficient factual basis. Neal, 907 F.3d at 514. Christiansen testified that she thought that she raised “strong” issues because aside from one comment at the change of plea hearing about Neal fronting drugs, the record otherwise indicated to her that it was a buyer-seller case rather than a conspiracy. We see why she might have found this angle promising. Back then, our buyer-seller case law was marked by “tension and inconsistency.” United States v. Page, 123 F.4th 851, 858–59 (7th Cir. 2024) (en banc) (stating that our “conspiracy and buyer-seller jurisprudence has drifted far from the Supreme Court's guidance” and “correcting course” by overruling precedent); see also United States v. Brown, 726 F.3d 993, 1000–01 (7th Cir. 2013) (“There is disagreement in our case law, however, over what other evidence, when combined with a credit arrangement, is sufficient to infer conspiracy.”).

All the same, the arguments Christiansen made on appeal were “ ‘an uphill battle’ ․ in overcoming the ‘presumption of verity’ that the law attaches to a guilty plea.” Neal, 907 F.3d at 514 (quoting United States v. Patterson, 576 F.3d 431, 437 (7th Cir. 2009)). Neal's plea colloquy evinced that his plea was “knowing and voluntary” and that the “government's factual proffer was sufficient.” Id. at 516; see Hutchings v. United States, 618 F.3d 693, 699 (7th Cir. 2010) (“[A] defendant is normally bound by the representations he makes to a court during the colloquy.”). Neal admitted that he “oftentimes” bought resale quantities of cocaine from the same supplier on credit, which gave rise to an inference that he shared a common purpose with another person to distribute the drug. Neal, 907 F.3d at 513–14. Christiansen may have been disappointed when we held that “Neal understood the charge against him and his plea had an adequate factual basis,” but she could not have been shocked. Id. at 516.

Looking back, it is fair to say the Ruth argument was stronger than those Christiansen made (if that is not already apparent from the outcome of Ruth compared to Neal's direct appeal). But we cannot conclude the Ruth argument was “clearly stronger” at the time of Neal's appeal.

A significant obstacle to comparison is that the arguments aimed at different goals. As an appellate lawyer “must,” Christiansen “curtail[ed] the number of issues” by focusing on Neal's conviction.⁷ Knox v. United States, 400 F.3d 519, 521 (7th Cir. 2005). If Christiansen's arguments prevailed, Neal would have had a second opportunity for trial and acquittal. A successful Ruth challenge would have left Neal with a 20-year sentence. Those different ends do not affect the likelihood of success for their different means, but they would influence an attorney's professional judgment about the “strongest” issues for appeal. Although we can imagine situations where an argument against a defendant's sentence is strong enough that it must be raised even if in the alternative, appellate attorneys are not always required to pick issues with a higher likelihood of success but a lower reward.

Holding that Christiansen provided ineffective assistance would thus encourage a kitchen-sink approach to advocacy, the implicit logic being that Christiansen should have made the Ruth argument in addition to the issues she raised, not instead. Would an attorney still think we would find the Ruth argument persuasive if she knew it shared a limited page-count with Neal's plea arguments? See Makiel, 782 F.3d at 899 (underscoring the “emphasis in appellate practice on not raising too many different issues and thus not diluting or burying an appellant's strongest points”).

Even in De La Torre, we qualified our decision as “not solely dependent on the definition of methamphetamine and which of its isomers do or do not exist.” De La Torre, 940 F.3d at 952. Only in Ruth did we rest our judgment squarely on the isomer question, supported in part (but not made obvious) by De La Torre. We are loath to imply that the Sixth Amendment frowns on the incremental development of the law. Attorneys need discretion to sometimes save their ideas for a later day.

“To avoid the distorting effects of hindsight,” we must put our later decisions out of mind. Shaw, 721 F.3d at 915 (quoting Strickland, 466 U.S. at 689, 104 S.Ct. 2052). Before Elder, Christiansen or any other attorney could have done nothing more than speculate whether we would endorse the categorical approach. After Elder, the forecast on our next steps would still be hazy. Recall that in Ruth the government argued that the supposed overbreadth of Illinois law was illusory because positional isomers do not actually “exist in the drug trade.” Ruth, 966 F.3d at 648. We were unpersuaded, but a competent attorney could have predicted a different reaction, especially when weighing the likelihood of that outcome versus other arguments. Equally true is that reasonable minds could have disagreed on whether unclear buyer-seller precedent would accrue to Neal's benefit. We do not grant that a competent attorney must have made those twin guesses, compared the odds, and judged the Ruth argument to be “clearly stronger.”

“This is not a case like Shaw v. Wilson, for example, where the single claim actually raised was so weak that pursuing it ‘was the equivalent of filing no brief at all.’ ” Makiel, 782 F.3d at 899 (quoting Shaw, 721 F.3d at 915). Nor is it like Cates v. United States, where counsel made one argument that was “certain to fail,” leaving unraised that a critical jury instruction was “flatly” against the law at the time of the appeal. 882 F.3d 731, 737 (7th Cir. 2018). Christiansen raised unsuccessful but colorable arguments rather than an untested novel argument. That the Ruth argument eventually succeeded is irrelevant, lest our Strickland jurisprudence crumble into retrospection. Based on the chronology of Neal's appeal, Christiansen was not constitutionally obliged to predict our years-later ruling on an issue of first impression, and therefore did not provide ineffective assistance.

C. Sentencing Counsel Goeke

Goeke's affidavit explains that he never discussed a Ruth challenge with Neal, but it does not say if he ever considered one. Assuming he did not, we nevertheless conclude that Goeke did not provide ineffective assistance. Considering the options, and Neal's preferences for how to handle the case, Goeke's performance was “within the wide range of reasonable professional assistance.” Strickland, 466 U.S. at 687, 104 S.Ct. 2052.

Goeke represented Neal in 2017, before Elder applied the categorical approach to Section 841 enhancements (and before Brock-Miller broached the idea). The Ruth argument may have been foreseeable, yet Goeke had little to indicate that we would receive it well.

Of course, that was also true in Coleman, but we repeat the difference: Neal's recidivism enhancement did not lead to a mandatory life sentence. The district court could have found that the enhancement applied and sentenced Neal to 20 years. In that discretionary sentencing regime, Goeke was not necessarily required to explore every avenue for defeating the enhancement.

In fact, Neal's preferred strategy required Goeke to focus his attention elsewhere. On Neal's request, Goeke attacked the enhancements for use of violence, obstruction of justice, and reckless endangerment. After Neal provided “the names of a number of witnesses that he wanted [Goeke] to interview,” Goeke retained an investigator to interview those witnesses, subsequently spoke with several, and subpoenaed them to appear at Neal's sentencing hearing.

Notably, Goeke also negotiated an agreement that stipulated an offense level of 31 and capped the government's sentencing recommendation at 20 years. Neal initially signed it before changing his mind. If the agreement had gone forward, and the district court followed the recommendation, Neal would have the same term of incarceration as under a successful Ruth challenge. Despite the district court's discretion to issue a higher sentence, and Neal's eventual decision to walk away, Goeke reasonably took the course of negotiating with the government. See, e.g., Harris, 13 F.4th at 631 (explaining that counsel is “not deficient for declining to press an unsettled legal question when doing so would have detracted from defendant's efforts to obtain a lesser sentence” (describing Harris v. United States, 366 F.3d 593, 596 (7th Cir. 2004))).

After Ruth, we can comment that Goeke's attempts to lower Neal's offense level and convince the district court to exercise its discretion leniently were not “the best possible approach” to reducing Neal's exposure. Williams v. Lemmon, 557 F.3d 534, 538 (7th Cir. 2009). But “it is difficult to establish ineffective assistance when counsel's overall performance indicates active and capable advocacy.” Harrington, 562 U.S. at 111, 131 S.Ct. 770. Such is the case here. Neal's claim regarding Goeke is unsuccessful.

D. Plea Counsel Winslow

Lastly, Neal argues that he pleaded guilty because of Winslow's ineffective assistance during plea negotiations. As always, under step one of Strickland, he must first establish that Winslow's representation fell below the standard of a reasonably competent attorney. United States v. Barnes, 83 F.3d 934, 939 (7th Cir. 1996). For step two, he must show a reasonable probability that but for Winslow's errors, he “would not have pleaded guilty and would have insisted on going to trial.” Lee v. United States, 582 U.S. 357, 364–65, 137 S.Ct. 1958, 198 L.Ed.2d 476 (2017) (quoting Hill v. Lockhart, 474 U.S. 52, 59, 106 S.Ct. 366, 88 L.Ed.2d 203 (1985)).

Winslow's affidavit states that she did not discuss a Ruth challenge with Neal, and like with Goeke, we assume she did not realize it was possible. That does not mean Winslow's performance was deficient. As we explained when discussing Goeke, Coleman is about (1) the sentencing stage and (2) a mandatory life sentence. Representing Neal during plea talks did not require Winslow to accurately predict the future in anticipating the precise argument that carried the day in Ruth. However, Winslow did need to consider favorable arguments that might have affected Neal's ultimate sentence.

Still, Neal argues Winslow provided ineffective assistance because he had a right to know that the Ruth argument existed. True, sometimes an attorney can be ineffective when she does not “consult with the defendant about a potential challenge to a sentence enhancement, thus failing to provide him ‘with the information necessary for a knowing and voluntary guilty plea.’ ” Harris, 13 F.4th at 631 (quoting Gaylord v. United States, 829 F.3d 500, 507 (7th Cir. 2016)). But counsel does not need “to raise with their client every potential sentencing challenge.” Id.

In 2016, when Winslow represented Goeke, a Ruth challenge was novel. See Harris, 13 F.4th at 631 (noting that in Gaylord “[c]ounsel advised the defendant to plead guilty in 2011, without mentioning a viable challenge to his sentence enhancement that had been established more than a year prior”). Imagine the conversation that Neal says should have occurred. There, Winslow would explain that she could argue the categorical approach should apply to the recidivism enhancement, but that the Seventh Circuit had never discussed the issue and she could not provide a concrete prediction on whether we would be amenable to it. She would note that if successful, a challenge changes his statutory maximum but might not lower his actual sentence, which could still be up to 20 years. On these facts, Winslow's failure to have a discussion defined by uncertainties was not so far outside professional norms as to be incompetent.⁸

Before we declare Winslow's conduct objectively reasonable, we also answer Neal's contention that Winslow should have used the prospect of a categorical challenge as leverage in plea negotiations with the government. Winslow's affidavit suggests that tactic could have backfired. Neal was charged under 21 U.S.C. § 841(b)(1)(C), which does not require the government to prove a minimum amount of cocaine. According to Winslow, the government considered seeking a superseding indictment—expanding the dates of the alleged conspiracy and including additional drug amounts—that could have exposed Neal to a statutory maximum of life imprisonment. See 21 U.S.C. § 841(b)(1)(A). If Winslow had informed the government that she was going to challenge the recidivism enhancement, the government may well have followed through with the superseding indictment and Neal might have ultimately received a sentence longer than 30 years. Given that possibility, even if Winslow had contemplated the Ruth argument, she would have been within reason to avoid raising it with the government. That is all to say that Neal fails to meet his burden of showing counsel's performance was not objectively reasonable. See Walker v. Litscher, 421 F.3d 549, 558 (7th Cir. 2005). Neal's ineffective assistance claim regarding Winslow fails at step one of Strickland.

III. Conclusion

We AFFIRM the district court judgment.

FOOTNOTES

1.   The Presentence Investigation Report concluded based on Neal's age, prior convictions, and the characteristics of the instant offense that Neal was a career offender, which carries an offense level of 34 under U.S.S.G. § 4B1.1. But because Neal's calculated offense level was 38, his guidelines were computed using that higher offense level. Neal also received a three-level decrease for acceptance of responsibility, resulting in a total offense level of 35. U.S.S.G. § 3E1.1(a)–(b).

2.   Another categorical approach case between De La Torre and Ruth, United States v. Garcia, offered a break from chemistry. 948 F.3d 789 (7th Cir. 2020). Garcia was “easy” because Indiana law prohibited salvia but federal law did not. Id. at 793.

3.   Neal's petition also alleged Winslow was ineffective for several other reasons, including because she purportedly misinformed him of his maximum sentencing exposure prior to his guilty plea. The district court found that Winslow adequately communicated with Neal about his likely sentence and that Neal was additionally informed about his possible 30-year maximum sentence during his guilty plea colloquy. These issues are not before us, as the expanded certificate of appealability is limited to whether Winslow was “ineffective for failing to challenge the recidivism enhancement.”

4.   We decided Coleman after the district court opinion in this case.

5.   Viewing Harris together with Coleman, the same is also true for Indiana cocaine convictions. See Coleman, 79 F.4th at 832 & n.9.

6.   The fact that Christiansen argued Elder does not impact whether the Ruth argument was objectively “clearly stronger” than the arguments she raised in Neal's appeal.

7.   Christiansen did not challenge the substantive reasonableness of the sentence or any of the offense-level enhancements.

8.   We therefore need not address Neal's claim that he would have gone to trial if he knew about the Ruth argument, which goes to step two of Strickland.

Kolar, Circuit Judge.