COA Denial Where Appellate-IAC “Cause” Fails: Sixth Circuit Rejects “Dead-Bang Winner” Framing and Applies Reasonable-Probability Standard

Case: Robert Dennis v. Michael Burgess (6th Cir. Mar. 17, 2025) (recommended for publication)
Posture: Motion for Certificate of Appealability (COA) from denial of a 28 U.S.C. § 2254 petition

1. Introduction

This Sixth Circuit decision addresses a familiar habeas bottleneck: when a petitioner’s federal claim is procedurally defaulted under state postconviction rules, federal review is barred unless the petitioner can show “cause and prejudice” (or a miscarriage of justice). Robert Lee Dennis, a Michigan prisoner, attempted to use ineffective assistance of appellate counsel as “cause” to excuse his defaulted claim that the prosecutor breached a plea agreement.

Dennis faced four counts of first-degree criminal sexual conduct involving a victim under 13, each carrying a mandatory minimum of 25 years and a maximum of life. On the trial date, the prosecutor offered a deal: in exchange for a guilty/no-contest plea to all four counts, she would not ask for more than 25 years on the minimum and would not bring charges regarding a second victim. Dennis pleaded no contest after repeated advisements that the court was not bound by the prosecutor’s recommendation. The court imposed 30-to-50-year terms and made two counts consecutive.

In state postconviction proceedings Dennis reframed his challenge: he asserted the prosecutor “covertly” advocated for consecutive sentences, breaching the agreement, but the state courts rejected relief under Michigan Court Rule 6.508(D)(3) (claims that could have been raised on direct appeal). In federal habeas, the district court held the claim procedurally defaulted and found no adequate “cause” to excuse it; Dennis sought a COA.

2. Summary of the Opinion

The Sixth Circuit denied a COA. Applying the COA standards of Miller-El v. Cockrell and Slack v. McDaniel, the court concluded that reasonable jurists could not debate the district court’s ruling that Dennis failed to establish “cause” because appellate counsel was not ineffective for omitting a meritless plea-breach claim.

Critically, the court rejected any suggestion that the governing ineffective-assistance standard requires showing the omitted issue was a “dead-bang winner.” Although that phrase appears in some case discussions, the Sixth Circuit emphasized it is not the circuit’s approved test and is not an appropriate description of the standard. Instead, the court applied the established appellate-IAC framework: whether the omitted issue was clearly stronger than those raised and whether there is a reasonable probability the appeal would have succeeded had it been raised. On the record, the prosecutor did not recommend consecutive sentences; she responded to defense counsel’s objections by explaining consecutive sentencing was within the court’s discretion while reiterating she was not asking for more than 25 years. Thus, the underlying breach claim lacked arguable merit, defeating “cause,” and the COA was denied.

3. Analysis

A. Precedents Cited

  • Santobello v. New York, 404 U.S. 257, 262 (1971)
    Dennis invoked Santobello for the constitutional proposition that the prosecution must honor material promises inducing a plea. The Sixth Circuit did not reach the merits of a Santobello claim because the case turned on procedural default and “cause.” Nevertheless, Santobello provided the substantive backdrop for evaluating whether the omitted issue had enough merit that competent appellate counsel should have raised it.
  • Miller-El v. Cockrell, 537 U.S. 322, 327 (2003) and Slack v. McDaniel, 529 U.S. 473, 484 (2000)
    These cases supply the COA gatekeeping framework. Miller-El articulates the general “substantial showing” threshold; Slack adds the two-part requirement when the district court denies relief on procedural grounds: debatability of both (i) the constitutional claim and (ii) the procedural ruling.
  • Coleman v. Thompson, 501 U.S. 722, 749-50 (1991) and Murray v. Carrier, 477 U.S. 478, 488-89 (1986)
    Coleman provides the default rule: federal habeas review is barred by adequate and independent state procedural grounds absent cause and prejudice (or miscarriage of justice). Murray recognizes that ineffective assistance can constitute “cause” if it satisfies constitutional standards.
  • Strickland v. Washington, 466 U.S. 668, 688, 694 (1984)
    The foundational two-prong test for ineffective assistance—deficient performance and prejudice—frames whether appellate counsel can serve as “cause.” The court emphasized the strong presumption of effective performance and the need for a “reasonable probability” of a different result.
  • Chase v. MaCauley, 971 F.3d 582, 592, 595 (6th Cir. 2020), Webb v. Mitchell, 586 F.3d 383, 399 (6th Cir. 2009) (quoting Smith v. Robbins, 528 U.S. 259, 288 (2000))
    These cases supply the Sixth Circuit’s appellate-IAC formulation: the omitted issue must be “clearly stronger” than issues raised, and the petitioner must show a reasonable probability of prevailing had the issue been presented. The Dennis court treated this as the controlling standard, explicitly choosing it over “dead-bang winner” rhetoric.
  • Rush v. King, No. 23-1770, 2024 WL 4043118, at *17 (6th Cir. Sept. 4, 2024) (per curiam), cert. denied, 2025 WL 247473 (U.S. Jan. 21, 2025)
    Cited to underscore that the Sixth Circuit has not adopted “dead-bang winner” terminology. Dennis uses Rush to critique the district court’s language; the panel agrees with the critique but finds the outcome unchanged under the correct standard.
  • McMeans v. Brigano, 228 F.3d 674 (6th Cir. 2000); Fields v. Forshey, No. 22-3031, 2023 WL 5093481, at *3 (6th Cir. Aug. 8, 2023); Carpenter v. Mohr, 163 F.3d 938, 947 (6th Cir. 1998), rev'd on other grounds, 529 U.S. 446 (2000) (quoting Banks v. Reynolds, 54 F.3d 1602, 1515 (10th Cir. 1995))
    These citations map the history of the “dead-bang winner” phrase. The court’s key move is doctrinal: it treats “dead-bang winner” as, at most, a sufficient example of ineffective assistance (if counsel omitted an obvious winner), not as a necessary condition.
  • Shaneberger v. Jones, 615 F.3d 448, 452 (6th Cir. 2010)
    Provides the decisive principle for this record: appellate counsel is not ineffective for failing to raise a meritless claim. This links directly to the “cause” analysis under Coleman/Murray.
  • People v. Dennis, 913 N.W.2d 306 (Mich. 2018) (mem.) and People v. Dennis, 961 N.W.2d 147 (Mich. 2021) (mem.)
    These citations situate the case procedurally: the Michigan appellate courts denied leave on direct appeal and later denied leave in postconviction review, reinforcing that the plea-breach theory was not properly raised on direct review.

B. Legal Reasoning

The decision proceeds in three linked steps: (1) COA standard; (2) procedural-default framework; (3) whether appellate counsel’s omission supplies “cause.”

  1. COA inquiry focuses on debatability.
    Because the district court denied relief on a procedural ground, Slack v. McDaniel required Dennis to show that reasonable jurists could debate both the procedural ruling (default not excused) and the viability of the constitutional claim.
  2. Procedural default was conceded; only “cause” was disputed.
    Dennis did not contest that Michigan Court Rule 6.508(D)(3) barred his claim because it could have been raised on direct appeal. He therefore needed “cause and prejudice” under Coleman v. Thompson.
  3. Appellate-IAC can be “cause,” but only if the omitted claim had real merit.
    The panel accepted the correct doctrinal path—ineffective assistance can be “cause” (Murray v. Carrier)—but held Dennis could not satisfy Strickland v. Washington as applied to appellate advocacy (Webb v. Mitchell, Smith v. Robbins, Chase v. MaCauley).
    • Clarifying the standard: The court expressly disapproved using “dead-bang winner” as the governing test, relying on Rush v. King to emphasize that the circuit has not adopted that terminology. It then applied the “reasonable probability of success” approach from Chase.
    • Record-based assessment of breach: The core factual/legal question was whether the prosecutor breached the plea agreement by implicitly urging consecutive sentences. The court read the sentencing transcript as showing:
      • Defense counsel challenged the presentence report’s note that consecutive sentencing was discretionary.
      • The prosecutor responded to that dispute by stating the court “could” sentence consecutively under Mich. Comp. Laws § 750.520b(3), but immediately reiterated she was not asking for more than the 25-year mandatory minimum.
      • In follow-up questions, the prosecutor again framed the matter as the court’s discretion, and even limited how many counts could be consecutive.
      • The trial judge stated the plea agreement did not bind the court and that this sentencing choice was the court’s exercise of discretion.
      The court concluded that none of this amounted to an explicit or implicit recommendation for consecutive sentences; it was an explanation of legal discretion in response to defense argument and judicial inquiry.
    • Meritlessness defeats appellate-IAC: Because the underlying breach claim “lacked arguable merit,” the omission could not be prejudicial under Strickland, and under Shaneberger v. Jones, counsel could not be ineffective for omitting it. Without ineffective assistance, Dennis lacked “cause,” so the procedural default stood.

C. Impact

1) Doctrinal clarification in the Sixth Circuit: The opinion reinforces that “dead-bang winner” is not the Sixth Circuit’s operative standard for appellate ineffective assistance. Courts and litigants should frame the inquiry using the established tests: whether the omitted issue was “clearly stronger” and whether there is a “reasonable probability” of success on appeal (Webb/Smith/Chase), not whether it was an obvious slam dunk.

2) Practical guidance for plea-breach claims at sentencing: The decision draws a line between (i) a prosecutor’s breach (advocating contrary to the bargain) and (ii) a prosecutor responding to defense objections or judicial questions by describing the court’s lawful discretion—especially where the prosecutor reaffirms adherence to the negotiated recommendation. Future habeas petitioners alleging “covert advocacy” will likely face heightened scrutiny of the transcript to identify a concrete recommendation rather than a legal clarification.

3) Procedural-default gatekeeping remains decisive: Even where a plea-breach claim invokes a prominent constitutional case like Santobello v. New York, the Sixth Circuit’s analysis shows that federal review can turn entirely on whether the petitioner can satisfy “cause” under Strickland and related appellate-IAC precedents.

4. Complex Concepts Simplified

  • Certificate of Appealability (COA): Permission to appeal a federal habeas denial. The petitioner must show the issues are debatable among reasonable judges—not that he will win.
  • Procedural default: If a state court rejects a claim because the prisoner failed to follow state procedural rules (e.g., didn’t raise it on direct appeal), federal habeas courts usually cannot review the claim.
  • Cause and prejudice: A narrow escape hatch from procedural default. “Cause” is a legally acceptable reason the claim wasn’t properly raised; “prejudice” is meaningful harm from the error. Ineffective assistance can be “cause,” but only if it satisfies Strickland.
  • Ineffective assistance (Strickland): The lawyer’s performance must be objectively unreasonable and there must be a reasonable probability the result would have been different without the error.
  • “Dead-bang winner”: Informal shorthand sometimes used to describe a claim so strong that failing to raise it is clearly deficient. This opinion stresses that the Sixth Circuit does not require a petitioner to meet that label; the real test is reasonable-probability-of-success under the governing appellate-IAC precedents.
  • Plea agreement breach (Santobello): If the prosecution makes a promise that induces the plea, it must honor it. But a breach typically requires advocacy inconsistent with the promise, not merely acknowledging the judge’s independent sentencing discretion.

5. Conclusion

Key takeaways: (1) The Sixth Circuit denied a COA because Dennis could not show “cause” to excuse his procedural default; (2) appellate counsel was not ineffective for omitting a plea-breach claim that the record did not support; and (3) the court expressly rejected “dead-bang winner” as the governing standard, reaffirming that appellate-IAC claims turn on the established “clearly stronger issue” and “reasonable probability of success” framework.

In broader context, the opinion strengthens doctrinal precision in Sixth Circuit ineffective-assistance analysis and underscores that plea-breach allegations—particularly those premised on “covert” advocacy—rise or fall on careful transcript-based proof of an actual recommendation inconsistent with the bargain.