dissenting.
I.
Although I agree with the bulk of the majority’s analysis, I believe that Benge has presented one meritorious claim that should entitle him to a writ of habeas corpus. When Benge’s attorney failed to object to the jury instructions regarding the lesser included offense of voluntary manslaughter, resulting in a jury charge that the Ohio Supreme Court later acknowledged was erroneous, he failed to provide Benge effective assistance of counsel. Because I believe a habeas writ should issue regarding that claim under Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), I respectfully dissent.
In order to make out an ineffective assistance of counsel claim under Strickland, “a defendant must show that counsel’s performance fell below an objective standard of reasonableness, and that the defendant was prejudiced by the attorney’s error.” Dando v. Yukins, 461 F.3d 791, 798 (6th Cir.2006). The district court correctly determined that trial counsel’s failure to object to the jury instruction fell below an objective standard of reasonableness, due largely to the fact that the instruction regarding voluntary manslaughter was clearly erroneous, rendering the failure to lodge an objection to the instruction objectively unreasonable.1 Benge v. Johnson, *251312 F.Supp.2d 978, 988 (S.D.Ohio 2004). Moreover, as Benge points out on appeal, trial counsel’s strategy involved calling Benge to testify and admit to killing Gab-bard, while contending that he acted in a passion or sudden fit of rage after she provoked him by attempting to run him over with the car. After establishing this record, it was imperative to Benge’s defense that the jury charge include a correct instruction regarding voluntary manslaughter as an affirmative defense. Thus, trial counsel put all of Benge’s eggs in the basket of voluntary manslaughter based on the mitigating circumstances of sudden passion or fit of rage, but then dropped the basket (and maybe even stepped on the eggs) by not even seeking a jury instruction consistent with this theory of the case. This abandonment of Benge’s trial theory at the jury instruction stage clearly falls below the objective standard of reasonableness required of counsel under Strickland.
The more contentious point involves the second prong under Strickland, which examines whether Benge was prejudiced by counsel’s mistake. The district court concluded that Benge could not establish prejudice, reasoning that because the jury convicted Benge of aggravated robbery in addition to aggravated murder, it necessarily rejected his version of events, including his testimony that Gabbard had provoked him into a fit of rage. Benge, 312 F.Supp.2d at 991. Under this view, even though the jury instructions effectively precluded the jury from considering the lesser included offense of voluntary manslaughter, the omission could not have prejudiced Benge because “the jury necessarily rejected his defense of sudden passion and provocation.” Id. The majority similarly concludes, based on its independent assessment of the evidence at trial, that there is no reasonable probability of a juror believing that Benge was seriously provoked. Maj. Op. at 254. I respectfully disagree with this analysis.
The majority correctly notes that based on Strickland, “the issue for the state courts was whether, but for defense counsel’s failure to object to the erroneous jury instruction, there was a reasonable probability that the outcome of Benge’s case would have been different.” Maj. Op. at 247. The majority is also correct that AEDPA’s deferential standard of review is inapplicable here, given the shortcomings of the state appellate courts’ review of this issue. Considering Benge’s claim de novo, as we are required to do by the omissions of the state courts, I would find that he is entitled to a habeas writ.
Based on the evidence at trial, a reasonable juror could have accepted aspects of both the prosecution and defense arguments, and determined that Benge was first provoked by Gabbard, and subsequently killed and robbed her. A conviction for aggravated robbery does not as a matter of law preclude an affirmative defense of provocation with regard to a related murder charge.2 Because there is no *252planning or premeditation required for a robbery conviction, the robbery conviction does not defeat the possibility that there was also provocation. Based on the evidence presented in this case, Benge could have had no plan to rob or murder Gab-bard when they got in the car together. He could have then been provoked by her fighting with him, and attacked her in response, consistent with his testimony at trial. Upon concluding his attack on Gab-bard, it could have occurred to him to take the ATM card from her before disposing of her body in the river. Under this set of facts, Benge could establish the affirmative defense of provocation with regard to the murder charge, even though he would still be guilty of aggravated robbery, for inflicting serious harm on another while committing a theft offense and/or for using a dangerous weapon in committing a theft offense.
Given the trial court’s instruction, however, the very real possibility of the jury finding both aggravated robbery and provocation was foreclosed. I disagree with the district court’s conclusion that “it is impossible to conclude that the mistake had any bearing on the conclusions reached by the jury in this case.” Benge, 312 F.Supp.2d at 991. Because as a matter of law the jury could have found both that Benge was provoked and was guilty of robbery, I believe that there is a reasonable probability that the mistake affected the verdict by eliminating the possibility of such a finding and the consequent verdict that Benge was guilty of voluntary manslaughter instead of aggravated murder. This probability caused Benge prejudice under the second prong of Strickland.
I also disagree with the majority’s assessment of the evidence, and its conclusion that “there is no reasonable probability that a juror would have accepted Benge’s voluntary-manslaughter defense, even if the jury had been properly instructed.” Maj. Op. at 248. Whatever doubts we as judges may have about Benge’s testimony, the Sixth Amendment prohibits the substitution of a court’s judgment for that of the jury. See Barker v. Yukins, 199 F.3d 867, 874 (6th Cir.1999) (“[T]he Michigan Supreme Court’s determination that the erroneous jury instruction was harmless necessarily means that the court believed some evidence but discredited other evidence. This, however, it cannot do and remain in compliance with our constitutional guarantees.”).3 Al*253though we must assess the evidence presented at trial to evaluate the prejudice created by ineffective assistance of counsel, I believe that like the district court, the majority is too dismissive of the possibility that the jury may have partially believed Benge had it been properly instructed. Foreclosing this possibility necessarily required credibility determinations by the reviewing judges, including today’s majority, which substitute a judge’s view of the veracity of Benge’s story over that of a properly instructed jury. Instead of receiving a jury’s judgment on the most critical issue in his case — whether there was sufficient provocation to establish voluntary manslaughter — Benge has been given a death sentence based on the speculation of several judges as to how a hypothetical, properly instructed jury would have viewed the evidence.
The government’s evidence that contradicted Benge’s testimony is itself far from conclusive — Shields’s testimony is subject to questions about credibility, and the significance of the tire running through blood is not entirely clear. Although Benge’s testimony was not consistent with his statements after the incident, it does not automatically follow that his explanation of the events in his trial testimony would have necessarily been disregarded by the jury. I cannot agree with the majority’s characterization of the evidence against Benge as “overwhelming,” and I am not convinced that a properly instructed jury would have relied on this to disbelieve Benge’s testimony regarding the fight. I am not under any delusion that Benge could be mistaken for the young George Washington of the cherry tree story, who could not tell a lie. But viewing the evidence; including Benge’s testimony, in its entirety, there is a reasonable probability that the jury would have found that Benge and Gabbard indeed fought before the murder, and that this provocation was sufficient to render his offense voluntary manslaughter rather than aggravated murder.
Benge was in fact prejudiced by trial counsel’s failure to object to the instruction because of the likely effect of the erroneous instruction on jury deliberations. Benge therefore meets the prejudice requirement and has established ineffective assistance of counsel under Stñckland, as there was “a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” 466 U.S. at 694, 104 S.Ct. 2052. Because there can be little dispute that the failure to object to the erroneous jury instruction was deficient, and because it led to a prejudicial disregard of the lesser degree offense, I would find that Benge was denied the effective assistance of counsel on this claim, and that the habeas *254writ should issue on this ground.4 For this reason, I respectfully dissent from the majority’s holding.
II.
I also continue to adhere to my belief that the arbitrary enforcement of the death penalty, in Ohio and elsewhere in this country, violates the Eighth Amendment’s prohibition of cruel and unusual punishment and the Fourteenth Amendment’s Due Process Clause. See Moore v. Parker, 425 F.3d 250, 270 (6th Cir.2005) (Martin, J., dissenting). The indisputably faulty jury instructions in this case only reinforce those concerns. While the version of the murder to which Benge confessed — and indeed, any version of any murder — was heinous and deserving of extreme punishment, it is troubling that his conviction and death sentence were returned by a misinstructed jury that was precluded from convicting him of a lesser included offense in contravention of state law.
Additionally, the only legal hook on which Benge’s death sentence hangs is the jury’s finding that he also committed aggravated robbery by stealing Gabbard’s ATM card in the process of killing her. I recognize that this is an aggravating factor under Ohio law, which the Ohio legislature likely requires in order to comply with Gregg v. Georgia, 428 U.S. 153, 96 S.Ct. 2909, 49 L.Ed.2d 859 (1976), and its progeny so as to attempt to combat the arbitrary enforcement of the death penalty. Even so, the imposition of a death sentence in this case on the basis of that factor strikes me as promoting, rather than preventing, the arbitrary application of the death penalty. Had Benge impulsively and fatally hit his common law wife in the head with a tire iron in an abhorrent act of extreme domestic violence, instead of killing her to gain access to her ATM card, as the prosecution alleged and the jury supposedly found, would his conduct somehow be less heinous and reprehensible? Such a murder would be at least as revolting as the one that occurred here, yet as far as I can tell, would have presented none of the aggravating factors required for a death sentence under Ohio law. Benge’s actions can certainly not be taken lightly in any respect, but his “theft” of an ATM card — to which it appears he had shared access with Gabbard in the past — as a means of accessing money to support his drug habit, is better characterized as a pathetic act of a sick and miserable man than as a factor that makes this murder more heinous or deserving of the death penalty than any other. In fact, in the same month this panel heard oral argument in this case, I sat on a panel in another habeas corpus case stemming from an Ohio state court conviction where *255a defendant who planned and led the firebombing of a house that caused the deaths of five people, four of whom were children, was not given the death penalty. See Williams v. Haviland, 467 F.3d 527 (6th Cir.2006). Of this admittedly small sampling, any neutral observer would be hard pressed to identify Benge as the defendant more deserving of execution.
I fully recognize that the ability of the jury to sentence one defendant to death while sentencing another, convicted of an arguably more heinous crime, to life in prison, is a natural function of the Supreme Court’s ruling that the Sixth Amendment requires a jury to determine the presence of aggravating factors that warrant the death penalty. See Ring v. Arizona, 536 U.S. 584, 589, 122 S.Ct. 2428, 153 L.Ed.2d 556 (2002). I also believe that the Supreme Court has generally made an earnest effort to require states to apply the death penalty in accord with the Constitution both through its Sixth Amendment holding in Ring, and by condemning the arbitrary enforcement of the death penalty under the Eighth and Fourteenth Amendments. See Gregg, 428 U.S. at 195, 96 S.Ct. 2909; Furman v. Georgia, 408 U.S. 238, 92 S.Ct. 2726, 33 L.Ed.2d 346 (1972). Even so, it appears to me that this case provides one of likely many examples of the validity of Justice Blackmun’s observations in Callins v. Collins, 510 U.S. 1141, 1144, 114 S.Ct. 1127, 127 L.Ed.2d 435 (1994) (Blackmun, J., dissenting from denial of certiorari), in which he acknowledged that “the constitutional goal of eliminating arbitrariness and discrimination from the administration of death can never be achieved without compromising an equally essential component of fundamental fairness — individualized sentencing.”
Prior to Callins, Justice Blackmun had concurred in the results of Supreme Court opinions that affirmed death sentences, under the belief that certain procedural safeguards could eliminate arbitrariness in death sentencing. See id. In Callins, however, Justice Blackmun argued that it had become apparent that the Court could not have it both ways. He explained his revised view on the death penalty as follows:
From this day forward, I no longer shall tinker with the machinery of death. For more than 20 years I have endeavored— indeed, I have struggled — along with a majority of this Court, to develop procedural and substantive rules that would lend more than the mere appearance of fairness to the death penalty endeavor. Rather than continue to coddle the Court’s delusion that the desired level of fairness has been achieved and the need for regulation eviscerated, I feel morally and intellectually obligated simply to concede that the death penalty experiment has failed. It is virtually self-evident to me now that no combination of procedural rules or substantive regulations ever can save the death penalty from its inherent constitutional deficiencies. The basic question — does the system accurately and consistently determine which defendants “deserve” to die?- — cannot be answered in the affirmative. It is not simply that this Court has allowed vague aggravating circumstances to be employed, see, e.g., Arave v. Creech, 507 U.S. 463, 113 S.Ct. 1534, 123 L.Ed.2d 188 (1993), relevant mitigating evidence to be disregarded, see, e.g., Johnson v. Texas, 509 U.S. 350, 113 S.Ct. 2658, 125 L.Ed.2d 290 (1993), and vital judicial review to be blocked, see, e.g., Coleman v. Thompson, 501 U.S. 722, 111 S.Ct. 2546, 115 L.Ed.2d 640 (1991). The problem is that the inevitability of factual, legal, and moral error gives us a system that we know must wrongly kill some defendants, a system that fails to deliver the fair, consistent, *256and reliable sentences of death required by the Constitution.
Callins, 510 U.S. at 1145-46, 114 S.Ct. 1127. The conclusion Justice Blackmun reached was that “the proper course when faced with irreconcilable constitutional commands is not to ignore one or the other, nor to pretend that the dilemma does not exist, but to admit the futility of the effort to harmonize them. This means accepting the fact that the death penalty cannot be administered in accord with our Constitution.” Id. at 1157, 114 S.Ct. 1127.
In accord with Justice Blackmun’s comments, I do not believe that Benge’s death sentence, or for that matter many of the death sentences that this Court has reviewed, reflects the product of “the system accurately and consistently determin[ing] which defendants ‘deserve’ to die.” It is equally likely that Benge received a death sentence while other potentially more culpable convicted murderers in Ohio did not for entirely arbitrary reasons. One arbitrary and constitutionally troubling possibility is that Benge’s death sentence turned more on the ability (or inability) of his trial counsel than on the facts of his crime. See Moore, 425 F.3d at 270 (“One of the most clear examples of the arbitrariness of the death penalty is the common knowledge that those defendants with decent lawyers rarely get sentenced to death.”). This possibility is particularly likely here in light of counsel’s failure to object to a jury instruction that contravened the entire theory of Benge’s case, as discussed above in Part I, as well as the adverse consequences resulting from counsel’s concurrent representation of a potential defense witness in a drug case, and counsel’s failure to object to several prejudicial statements during the penalty phase of the trial.5 See Benge, 312 F.Supp.2d at 994-95, 1008-09.
Some judges view the situation quite differently, believing that the Sixth Amendment right to counsel, and the jurisprudence of this Court and the Supreme Court that require effective assistance of counsel, actually create incentives for defense counsel to intentionally provide constitutionally deficient representation in capital cases so that the resulting death sentences can later be thrown out on appeal. See Poindexter v. Mitchell, 454 F.3d 564, 588 (6th Cir.2006) (Boggs, J., concurring) (speculating that the Sixth Amendment jurisprudence of this Court and the Supreme Court creates a “moral hazard” by encouraging intentional ineffective assistance of counsel); Id. at 589 (Suhrhein-rich, J., concurring) (“I agree with Judge Boggs.”). As I have written elsewhere, see Keith v. Mitchell, 466 F.3d 540, 547 (6th Cir.2006) (Martin, J., dissenting from denial of rehearing en banc), I believe this view is simply out of touch with the realities of criminal trial practice. It would be a high-risk and misguided wager for an attorney to entrust her client’s Sixth Amendment rights to a reversal on appeal by a federal habeas court in light of the ever-growing deference that is shown to “strategic” decisions by defense counsel and legal decisions of state courts, and the apparent trend of the federal judiciary becoming increasingly willing to play fast and loose with the individual protections guaranteed by the Constitution simply to *257avoid temporarily standing in the way of a state’s rush toward death.6
The frequent findings of ineffectiveness of counsel in capital cases that Judge Boggs has documented have more to do with the fact that there is insufficient support, financial and otherwise, for attorneys representing capital defendants than it does with some scheme of providing intentionally deficient representation. See also Poindexter, 454 F.3d at 590 (Daughtrey, J., concurring) (concluding “contrary to Judge Boggs’s intimations, not that capital defense attorneys are engaged in a demented, premeditated game of ‘gotcha’ with the courts, but rather that those lawyers representing the absolute pariahs of society are frequently hamstrung by a critical lack of relevant experience, an obvious lack of time and resources, or both.”) (emphasis in original). Regrettably, the observations that Judge Daughtrey and I have made on this problem are nothing new, and have been documented, but not effectively remedied, for many years. See McFarland v. Scott, 512 U.S. 1256, 1256, 114 S.Ct. 2785, 129 L.Ed.2d 896 (1994) (Blackmun, J., dissenting from denial of certiorari) (“Without question, ‘the principal failings of the capital punishment review process today are the inadequacy and inadequate compensation of counsel at trial.’ ”) (quoting Ira Robbins, Toward a More Just and Effective System of Review in State Death Penalty Cases, Report of the American Bar Association’s Recommendations Concerning Death Penalty Habeas Corpus, 40 Am. U.L.Rev. 1, 16 (1990)). In our capitalist society you get what you pay for. We are yet to show a willingness to adequately compensate members of many professions (public school teachers, military and emergency response personnel, social workers, and yes, attorneys who represent indigent defendants, to name a few) whose competent performance is most important to the functioning of our democracy.
It is also very possible that the constitutionally impermissible factor of the race of Benge’s victim played a role in his death sentence. See Andrew Welsh-Huggins, Race, Geography Can Mean Difference Between Life, Death, The Assooiated PRESS, May 7, 2005 (explaining that a 2005 Associated Press study of Ohio death sentences found that “[o]ffenders facing a death penalty charge for killing a white person were two times more likely to go to death row than if they had killed a black victim. Death sentences were handed down in 18 percent of cases where the victims were white, compared with 8.5 percent of cases where victims were black.”); David Baldus and George Woodworth, Race Discrimination and the Legitimacy of Capital Punishment: Reflections on the Interaction of Fact and Perception, 53 DePaul L.Rev. 1411, 1423-255 (2004) (concluding that nationwide “defendants with white victims are at a significantly higher risk of being sentenced to death and executed than are defendants whose victims are black, Asian, or Hispanic.”); see also McCleskey v. Kemp, 481 U.S. 279, 286, 107 S.Ct. 1756, 95 L.Ed.2d 262 (1987) (noting that among capital murder cases in Georgia during the 1970’s, “defendants charged with killing white persons received the death penalty in 11% of the cases, but defendants charged with killing blacks received the death penalty in only 1% of the cases”). Benge could also have been sentenced to death by virtue of the entirely *258arbitrary factor of the location of his trial within Ohio. See Welsh-Huggins, supra (noting a significantly higher rate of death sentences in capital trials held in Southern Ohio compared to Northern Ohio). All of these possibilities underscore the merit of Justice Blackmun’s prediction that “death will continue to be meted out in this country arbitrarily and discriminatorily.” Callins, 510 U.S. at 1157, 114 S.Ct. 1127; see also Alley v. Little, 447 F.3d 976, 978 (6th Cir.2006) (Martin, J., dissenting from denial of rehearing en banc).
As I have previously stated, “I know my place in the judiciary,” Moore, 425 F.3d at 270, and I recognize that unless and until the Supreme Court deems it necessary to address what I (like Justice Blackmun and others) view as the inherent arbitrariness of the death penalty, my reflections on this topic will only be observations without the force of law. In the meantime, I add my voice to those dissenters who have hoped that the Supreme Court “eventually will conclude that the effort to eliminate arbitrariness while preserving fairness ‘in the infliction of [death] is so plainly doomed to failure that it — and the death penalty— must be abandoned altogether.’ ” Callins, 510 U.S. at 1159, 114 S.Ct. 1127 (Blackmun, J., dissenting from denial of certiorari, quoting Godfrey v. Georgia, 446 U.S. 420, 442, 100 S.Ct. 1759, 64 L.Ed.2d 398 (1980) (Marshall, J., concurring in the judgment)).
. The trial judge instructed the jury "[i]f your verdict is guilty [on the charge of aggravated murder], proceed to Specification One and Two and do not consider lesser included charges.” State v. Benge, 75 Ohio St.3d 136, 661 N.E.2d 1019, 1024 (Ohio 1996) (emphasis added). As the Ohio Supreme Court determined, the instruction was incorrect as a matter of Ohio law because voluntary manslaughter is a lesser included offense of aggravated *251murder, and consequently "the jury should have been instructed to consider the mitigating evidence to determine whether appellant proved voluntary manslaughter.” Id.
. The Ohio aggravated robbery statute provides as follows:
§ 2911.01. Aggravated robbery
(A) No person, in attempting or committing a theft offense, as defined in section 2913.01 of the Revised Code, or in fleeing immediately after the attempt or offense, shall do any of the following:
(1) Have a deadly weapon on or about the offender's person or under the offender's control and either display the weapon, brandish it, indicate that the offender possesses it, or use it;
*252(2) Have a dangerous ordnance on or about the offender’s person or under the offender’s control;
(3) Inflict, or attempt to inflict, serious physical harm on another.
. The majority states that Barker is inapposite because "it arose in a completely different context.” Specifically, in Barker, this Court was evaluating a harmless error analysis undertaken by a state court on direct review, as opposed to the question here of whether a defendant was prejudiced by counsel’s failure to request a correct jury instruction. I primarily cite to Barker for illustrative purposes, rather than as controlling precedent on the issue before us today. It is relevant authority for the point that whenever a court makes credibility determinations or otherwise overreaches in concluding that the evidence of guilt was overwhelming despite a significant mistake in instructing the jury — either a mistake by the trial court, which is subsequently deemed harmless, or by counsel, which is subsequently deemed non-prejudicial — it is invading the province of the jury. In both types of cases, assessing the evidence is an objective exercise, and it is not the role of a reviewing court to make credibility determinations. This point from Barker is equally applicable here, despite the fact that the issue before the Court was not identical to the present question.
Further, I am not convinced that the distinction between the issue presented in Barker and the issue here is as significant as the majority suggests. The standard for this Court's review of a state court's harmless error determination that was at issue in Bark*253er is whether the error in question "had a substantial and injurious effect or influence in determining the jury’s verdict and resulted in actual prejudice.” 199 F.3d at 873. For purposes of our review of the state court's determination of the prejudice prong of an ineffective assistance of counsel claim, which is at issue in this case, we look to whether there is “a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different.” Hodge v. Hurley, 426 F.3d 368, 376 (6th Cir.2005). The standard in both types of cases requires us to assess the evidence and make some hindsight judgment as to the likelihood of guilt in a hypothetical trial where the error in question did not occur. The illustrative point from Barker, that applies with equal force here, is that where the error had a substantial effect (or there is a probability that the result of the proceeding would have been different but for the error), a judge's view of guilt is no substitute for that of the jury, and cannot be used to reason away’ the significance of the mistake.
. The majority primarily frames the ineffective assistance of counsel issue as creating cause and prejudice for Benge’s procedural default, whereas I have primarily addressed his freestanding ineffective assistance of counsel claim. There are nuanced differences between these two analytical approaches. See Joseph v. Coyle, 469 F.3d 441, 459 (6th Cir.2006) ("Although Joseph must satisfy the AEDPA standard with respect to his independent [ineffective assistance of counsel] claim, he need not do so to claim ineffective assistance of counsel for the purpose of establishing cause.”). I do not believe that these differences are particularly relevant here, given that both the majority and I address Benge's Strickland, claim de novo. I would thus grant the writ either on the ground that Benge established cause and prejudice with regard to his erroneous jury instruction claim, or the roughly related independent ineffective assistance of counsel claim. See id. ("[Petitioner] has established his [ineffective assistance of counsel] claim under the AEDPA standard, which necessarily means that he has also established ineffective assistance of counsel for the purpose of establishing cause.”).
. Although I would agree with the district court's conclusion that there is an insufficient showing of prejudice for these latter two shortcomings to support viable habeas claims on their own, one cannot help but wonder whether the same result would have issued without the cumulative effect of the incorrect jury instruction, the cooperation of the witness that Benge claims was prevented by the concurrent representation, and the inflammatory comments during the penalty phase, all of which a fully competent defense attorney presumably would have prevented.
. See Herrera v. Collins, 506 U.S. 390, 446, 113 S.Ct. 853, 122 L.Ed.2d 203 (1993) (Black-mun, J., dissenting) (“I have voiced disappointment over this Court’s obvious eagerness to do away with any restriction on the States' power to execute whomever and however they please.”).