MEMORANDUM AND ORDER

NANCY F. ATLAS, District Judge.

Adonius Collier was murdered in 1992. A surviving victim led the police to suspect Nanon McKewn Williams' involvement in the crime. While Williams fled to California, his co-participant in the shootings, Vaal Guevara, turned himself into the police. Guevara's self-serving confession became the backbone of a flawed prosecution that led to Williams' capital-murder conviction three years later. The issue now before the Court is whether trial counsel's failure to seek expert assistance prejudiced Williams. In light of significant post-trial evidence showing deep flaws in the State's case against Williams, the Sixth Amendment requires that the Court vacate his capital conviction.

I. BACKGROUND

Williams' trial in 1995 provides the vantage point for evaluating his ineffective-assistance claims. This Court has elsewhere extensively outlined the crime, trial testimony, and post-trial factual development. [Doc. # 41 at 20-69]. For the purposes of this Memorandum and Order, the Court will only recount the factual background briefly.

In 1992, Williams and Vaal Guevara conspired to steal from two drug dealers while feigning interest in purchasing cocaine. The two dealers—Adonius Collier and Emmade Rasul—traveled to Hermann Park with Rasul's girlfriend, Stephanie Anderson. Williams and Guevara showed up with Guevara's neighbor Elaine Winn and their friend Patrick Smith, who until well after trial was only known by the name "Xavier." While the others waited in the parking lot, Williams, Guevara, Collier, and Rasul went across the street to transact their drug deal unobserved. Those who remained in the parking lot could not see the events that led up to Collier's murder.

After the men crossed the street, they paired off. Within minutes, Williams and Guevara began shooting at the other men. Rasul fled the scene after Williams shot him in the face and foot with his .25 caliber pistol. Guevara briefly chased Rasul, but apparently rejoined Williams. The police later found Collier lying dead with a massive shotgun wound in his head.

Williams fled the state. Accompanied by his attorney, Guevara turned himself in to the police shortly thereafter. Guevara admitted in his police interviews that he fired his .22 caliber derringer in Collier's direction.

A less-than-thorough autopsy and a careless forensic investigation led to the controversy now before the Court. When Dr. Tommy J. Brown, an Assistant Medical Examiner for the Harris County Medical Examiner's Office, performed the autopsy, he removed some shotgun pellets from Collier's cranial cavity. A mutilated bullet slug—eventually designated "EB-1"—came into evidence with the shotgun pellets. The trajectory and entrance of EB-1 have been perplexing because (1) medical examiners never observed a bullet in the now-destroyed X-ray of Collier's head and (2) the autopsy does not mention an entrance wound. The police department's handling of forensic evidence compounded the problems in determining the cause of Collier's death. Robert Baldwin, a criminalist with the Houston Police Department Firearms Laboratory, identified EB-1 as a .25-caliber slug. Mr. Baldwin would eventually confess that this was an error, but he did so much too late. The autopsy findings and forensic mistakes allowed the prosecution to argue that Williams fired both shots into Collier's head.

The State of Texas charged Williams with capital murder committed in the course of a robbery. Ms. Loretta Muldrow represented Williams at trial, which took place in 1995. Four of the people who traveled to Hermann Park— Anderson, Rasul, Winn, and Guevara—testified at trial about the events of that evening. Testimony from Anderson, Rasul, and Winn provided some context to the drug deal, but none of them saw Williams shoot Collier. Collectively, these three witnesses clarified who carried which weapon and identified who shot Rasul. Their testimony, however, did not answer significant questions about the events surrounding the murder.1 Importantly, the witnesses could not identify who fired EB-1. Their testimony could not conclusively provide a sequence for the two shots that went into Collier's head. They could not give information about Collier's physiological state before the shotgun injury occurred. The prosecution relied on Williams' co-defendant Guevara to fill in those gaps.

Guevara had been incarcerated for three years before trial. Although originally accused of capital murder, he pleaded guilty to an illegal investment of drugs charge. Guevara had not received a sentence at the time of trial, but the prosecution had agreed to ten years of incarceration in return for his truthful testimony. Tr. at 28:184-85. Guevara's testimony minimized his role in the planning and commission of the murder. In fact, Guevara claimed that he first thought that the two other men were robbing him when the shooting first started. Tr. at 28:240. Guevara admitted that he fired shots that night, though he tried to excuse himself as only having acted in self-defense. Tr. at 28:240, 243-44. Only with reluctance did Guevara testify that he had fired shots at both Rasul and Collier. Importantly, Guevara admitted that when he "saw [Collier] coming, he panicked and shot at him." Tr. at 29:84, 88. Guevara testified that he was certain that he was not the one who killed Collier. Tr. at 28:274-75. The prosecution's theory of the case, in fact, depended on Guevara's shot completely missing Collier. Guevara explained that, while chasing Rasul, he heard a loud noise, presumably the shotgun blast, and turned around to see Williams standing over Collier. Tr. at 28:251-52. Guevara provided two critical pieces of information about Collier's death: after Collier ended up on the ground, Guevara heard him mumbling, Tr. at 29:84, and after the shotgun blast, Guevara saw Collier's feet move, Tr. at 28:279-81; 29:83. The prosecution maintained that Williams first shot Collier with the .25 caliber pistol and, because he was not yet dead, finished him off with the shotgun.

Guevara's self-serving testimony, therefore, provided the most important basis for excluding himself as the killer, making the shotgun wound the only fatal injury. Guevara, however, was a difficult witness. He was a criminal, he had fired shots that night, and he obviously tried to minimize his role in the murder. Given Guevara's problematic testimony, the prosecution adduced some forensic evidence to show that Williams killed Collier. Dr. Brown, the assistant medical examiner who performed the autopsy, provided the State with a basis to argue that the shotgun was the only fatal wound. Dr. Brown testified at trial that, because "[t]here was a red margin around the entrance wound" instead of a "yellow margin," Collier "had blood pressure at the time he was shot with the shotgun." Tr. at 29:101-02. Because his autopsy did not reveal any bullet path for EB-1, and the shotgun blast "shredded" Collier's brain, Dr. Brown could not provide a sequence for the two shots. Tr. at 29:101-02. Dr. Brown opined that the shotgun blast was the cause of Collier's death. Tr. at 29:123.

To discount further the chance that Guevara was the killer, Mr. Baldwin identified EB-1 as a .25-caliber slug. With the testimony showing that Williams shot EB-1, the autopsy did not have to factor into the jury's deliberations. The jury convicted Williams of capital murder and answered Texas' special issues in a manner requiring a death sentence.2

Post-trial investigation has shown that the police misidentified EB-1. On January 15, 1998, a prosecutor asked Mr. Baldwin to test fire the two recovered slugs in this case, including EB-1. Mr. Baldwin test fired the .22-caliber pistol linked to Guevara and compared it with the two bullets recovered from the crime scene. Mr. Baldwin opined that the .22 pistol which Guevara had carried on the night of the murders had fired EB-1. F.E.H., PX-3(3) (letter signed by Robert D. Baldwin, dated January 15, 1998).3 Testing by Ronald Lopez. Singer, Chief Criminalist of the Tarrant County Medical Examiner's Office, confirmed the results of Mr. Baldwin's retesting. Mr. Singer opined that EB-1 and the bullets fired from Williams' gun "are easily distinguishable from one another, particularly if examined with the aid of a comparison microscope, and should have presented no problem to a competent firearms examiner." The misidentification "at best demonstrates extreme carelessness on [Mr. Baldwin's] part and at worst calls into question his expertise." F.E.H., PX3(5) (Affidavit of Ronald Lopez. Singer, dated April 15, 1998).

Williams' state habeas challenge turned on the fact that Guevara, not Williams, shot Collier with a .22 bullet. After full presentation of the evidence, the lower habeas court recommended granting him a new trial. In a tersely worded and opaque opinion, the Texas Court of Criminal Appeals adopted part and rejected part of that recommendation. When Williams came to federal court, the Anti-Terrorism and Effective Death Penalty Act ("AEDPA") ensured federal deference to state fact findings. Notwithstanding its decision to grant habeas relief, the lower state habeas court made findings that significantly narrowed the parameters of federal review. This Court parsed through the lower court's opinion and afforded AEDPA deference to those findings supporting a denial of habeas relief. In particular, the state habeas findings relied on Guevara's testimony suggesting that Collier was alive before and after the shotgun blast. See FFCL, at 10, ¶ 45 (stating that Collier "was hollering, moving, and saying `I'm shot' after the first shot (EB-1); ... [Collier] was alive when [Williams] shot [him] with the shotgun."). The state court findings significantly constricted this Court's review of forensic evidence by holding that EB-1 never penetrated Collier's cranial cavity and thus Williams "fail[ed] to demonstrate that EB-1 ... [caused] a potentially fatal wound[.]" FFCL, at 7-8, ¶¶ 35, 38. The AEDPA foreclosed any conclusion other than that Collier "was alive when [Williams] shot him." FFCL, at 10, ¶ 45; see also FFCL, at 8, ¶ 38 (stating that Williams failed to show that Collier "was dead when [he] shot him with the shotgun."). Based on those factfindings, the Court denied relief.

On appeal, the Fifth Circuit found that "no factual findings survive to which we can give deference" and ordered this Court to "conduct a de novo hearing of Williams's ineffective assistance of counsel claims premised on trial counsel's failings with respect to the ballistics and pathological evidence, as well as Williams's cumulative error point." Williams v. Quarterman, 551 F.3d 352, 359 (5th Cir.2008). Now disentangled from the specific state court findings that previously bound habeas review, the Court considers Williams' habeas claims in a much different light. Pursuant to this Court's observations at the evidentiary hearing, the Court can fully consider how the new evidence would impact the prosecution's case against Williams. Freed from the AEDPA's restrictive review, the Court finds various significant facts and reaches the conclusion that Williams is entitled to federal habeas corpus relief.4

II. LEGAL STANDARDS

This Court's inquiry begins with defining both the scope and nature of federal review at this juncture. Even when liberated from the AEDPA standards, the Fifth Circuit's order on remand and the parties' agreements still limit this Court's review. The Fifth Circuit has confined this Court's review Williams' allegation that trial counsel provided ineffective assistance. See Williams, 551 F.3d at 359. "The essence of an ineffective-assistance claim is that counsel's unprofessional errors so upset the adversarial balance between defense and prosecution that the trial was rendered unfair and the verdict rendered suspect." Kimmelman v. Morrison, 477 U.S. 365, 374, 106 S.Ct. 2574, 91 L.Ed.2d 305 (1986); see also United States v. Cronic, 466 U.S. 648, 658, 104 S.Ct. 2039, 80 L.Ed.2d 657 (1984) ("[T]he right to the effective assistance of counsel is recognized not for its own sake, but because of the effect it has on the ability of the accused to receive a fair trial."). Under Strickland v. Washington, 466 U.S. 668, 686, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), a reviewing court must decide whether (1) an attorney's performance fell below an objective standard of reasonableness and (2) actual prejudice ensued.

The parties have further narrowed this Court's Strickland inquiry through stipulation. The parties agree that trial counsel provided constitutionally deficient performance by failing to: (1) secure independent testing of Guevara's handgun; (2) obtain independent testing of EB-1, the fired bullet found during autopsy, thereby linking the slug to Guevara; and (3) secure the assistance of a pathologist. Accordingly, the parties stipulate that the only question remaining is whether trial counsel's deficiencies prejudiced the defense, focusing on the cumulative effect of trial counsel's errors.

Given the fact-dependant nature of the inquiry, Strickland's prejudice prong entails the application of broad principles, not mechanical tests. Strickland established that a "defendant must show that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome." Strickland, 466 U.S. at 694, 104 S.Ct. 2052; see also Rompilla v. Beard, 545 U.S. 374, 387, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005); Yarborough v. Gentry, 540 U.S. 1, 3, 124 S.Ct. 1, 157 L.Ed.2d 1 (2003). Beyond that bare outline definition, a court must consider the possibility of prejudice against all the discrete circumstances that a criminal defendant faced at trial. See Wong v. Belmontes, ___ U.S. ___, 130 S.Ct. 383, 386, 175 L.Ed.2d 328 (2009) ("In evaluating [Strickland prejudice], it is necessary to consider all the relevant evidence that the jury would have had before it."); Strickland, 466 U.S. at 695, 104 S.Ct. 2052 ("In making this determination, a court hearing an ineffectiveness claim must consider the totality of the evidence before the judge or jury."). The prejudice inquiry operates in the context of the reasonable-doubt standard; in other words, a defendant must show "reasonable probability that ... the factfinder would have had a reasonable doubt respecting guilt." Strickland, 466 U.S. at 695, 104 S.Ct. 2052. This requires showing "at least one juror would have" entertained reasonable doubt. Wiggins v. Smith, 539 U.S. 510, 513, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003); see also Loyd v. Smith, 899 F.2d 1416, 1426 (5th Cir.1990) ("Paraphrased, [to deny relief] the reviewing court must be confident that at least one juror's verdict would not have been different had the new evidence been presented.").

In assessing prejudice, however, a court cannot focus solely on whether the outcome of trial may have been different. In Lockhart v. Fretwell, 506 U.S. 364, 369-70, 113 S.Ct. 838, 122 L.Ed.2d 180 (1993), the Supreme Court observed that "an analysis focusing solely on mere outcome determination, without attention to whether the result of the proceeding was fundamentally unfair or unreliable, is defective." See also Goodwin v. Johnson, 132 F.3d 162, 174 (5th Cir.1997) ("[T]he presence or absence of prejudice, both with respect to claims of ineffective assistance of counsel at the trial and appellate levels, hinges upon the fairness of the trial and the reliability of the judgment of conviction resulting therefrom."); Cf. Williams v. Taylor, 529 U.S. 362, 393, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000) (reaffirming that Strickland "focuses on the question whether counsel's deficient performance renders the result of the trial unreliable or the proceeding fundamentally unfair"). Thus, "Strickland's prejudice inquiry is process-based: Given counsel's deficient performance, do we have confidence in the process afforded the criminally accused?" Virgil v. Dretke, 446 F.3d 598, 612 (5th Cir.2006).

III. ANALYSIS

Pursuant to the parties' stipulation, the Court assumes that trial counsel should have retained a forensic expert before trial who would have correctly identified EB-1 as a .22 caliber slug. In addition, trial counsel should have sought a medical expert to evaluate the autopsy results. The Court, therefore, starts with the presumption that trial counsel failed to defend Williams in a manner consistent with constitutional expectations. The issue is how those deficiencies impacted the trial. The prejudice flowing from trial counsel's deficient performance becomes most clear when considering how the evidence developed after trial would drastically alter: (1) the prosecution's theory of the case; (2) Guevara's culpability for the crime; and (3) testimony about whether EB-1 caused a fatal injury.

A. The Prosecution's Case

Without question, Williams' trial would have been entirely different had defense counsel acted with zeal and alacrity. Identifying Guevara as the one who fired EB-1 would not only have changed how the defense approached the case, but would have altered the fabric of the entire trial proceedings. From inception, the prosecution assumed that Williams had fired two weapons at Collier's head. The prosecution's opening arguments outlined the heart of their case against Williams:

Tr. at 26:15 (italics added). Post-trial development has shattered three underlying assumptions in the State's case: (1) the question of who killed Collier is not easily answered; (2) the full forensic evidence conclusively proves that Guevara fired the.22; and (3) Williams did not hold the shotgun to Collier's head when he fired. The first two assumptions in particular undermine the prosecution's whole theory of who committed capital murder and how he did it.

The jury heard that Guevara played some role in the robbery/murder, but clearly one ancillary and subsidiary to Williams. The prosecution's opening statements acknowledged that Guevara would not come before the jury with a pristine history:

Tr. at 26:20. Nevertheless, the prosecution confidently assured the jury that, no matter what other crimes he had committed, Guevara was not a killer. The trial prosecutor's opening statement described how Guevara was only a witness, not a party, to murder:

Tr. at 26:24-26.

Against that version of the events, trial counsel recognized that Williams' culpability hinged on Guevara's credibility. Trial counsel's opening argument focused on making Guevara a scapegoat. Trial counsel challenged Guevara's credibility by arguing that he initially told the police that he "did not actually participate in the shooting," Tr. at 26:34, but later "admit[ed] to firing his weapon at Adonius Collier" once he worked out a plea deal, Tr. at 26:40. Trial counsel, however, apparently realized that Williams' best defense would be proving that Guevara was the killer. She argued:

Tr. at 26: 43. The defense's evident plan was to blame Guevara for firing EB-1. In essence, trial counsel promised to present the case that has been developed in post-trial proceedings. However, at trial, without correct forensic testing, this strategy was toothless.

Because the defense did not follow through with the case it promised to put before the jury, the trial focused on the prosecution's theory that Williams fired two shots into the victim's head. The prosecution, therefore, did not have to emphasize the effect each gunshot wound had on Collier. The prosecution could label EB-1 a "fail-safe" to proving Williams' culpability:

Tr. at 31:104-05. The question of whether EB-1 inflicted a fatal injury was not a major focus of trial. Causation was a minor, and somewhat irrelevant, theme. Had trial counsel secured forensic assistance, the focus of trial would have changed from confirming Williams' guilt to allocating the blame between the two shooters.

As it was, the prosecution insisted that Mr. Baldwin was "absolutely, positively certain [that EB-1] couldn't be fired out of [Guevara's] Derringer." Tr. at 31:110. The prosecution mocked the suggestion that causation could be an issue in this case:

Tr. at 31:64-65. Testing has shown that the theory once labeled ridiculous should have been the focal point of trial. With the correct identification of EB-1, the jury's attention would have to have been on which injury caused Collier's death and would center on observations from the autopsy. The parties would have had to emphasize expert testimony about Collier's physiological condition before the shotgun blast. This change would foreclose the prosecutor's dismissive closing argument:

Tr. at 31:101-02. Without EB-1 as the "fail-safe" to conviction, the autopsy would have been the central focus of trial. The Strickland prejudice inquiry in this case hones in on the original jury's lack of consideration of causation.

B. Guevara's Credibility and Testimony

Guevara's participation in the crime made his credibility an important issue for the jury. Guevara claimed that he was duped into participating in a shoot-out and only began firing in self-defense. Guevara's self-serving account disturbed even the prosecution. The prosecutor hinted that Guevara's culpability was greater than the crime to which he pleaded guilty:

Tr. at 31:57. The prosecutor lamented: "We don't have enough evidence to prove [Guevara] guilty of capital murder." Tr. at 31:57. In reality, the State had the evidence, they had only misidentified it.

Defense counsel made some effort to pin the murder on Guevara, but relied on accusations, not evidence. The defense's unsupported claims allowed the prosecution's closing arguments to lay all the blame on Williams: "I don't want to bore you. I don't want to waste your time. I gave you about 10 or 15 reasons why in this case it just ain't Vaal [Guevara]. Who else could it be? Nanon Williams." Tr. at 31:99.5

A correct understanding of who fired EB-1 would undercut the prosecution's reliance on Guevara. Had the defense known who fired the slug into Collier's head, trial counsel would have had a strong argument that Guevara was either a party to, or a cause of, the murder himself. At a minimum, trial counsel could have gutted Guevara's testimony through cross-examination. Had trial counsel dismantled Guevara's account bit-by-bit, the jury likely would have put no credibility on his self-serving testimony.

Respondent seeks to limit the effect that EB-1's true caliber would have on Guevara's usefulness to the prosecution. To some extent, Respondent's position on Guevara's role has evolved throughout these proceedings. Initially, Respondent argued that forensic testing would not alter much of Guevara's testimony because "EB-1 was found outside the cranial cavity." [Doc. # 20 at 37]. Now, Respondent labels Guevara a "killer." [Doc. # 128 at 12]. Even so, Respondent still confines the impeachment of Guevara's testimony to only two points: "[e]ven if Williams had presented evidence that Guevara shot Collier first, it would merely discredit Guevara's belief that he missed Collier, or that Collier was outside the effective range of Guevara's .22 magnum derringer." [Doc. # 128 at 11]. The Court finds that, with the inevitable shift in the trial's focus because of the new evidence, Respondent seriously underestimates how post-trial events would affect the jury's perception of Guevara. With the knowledge that Guevara could have been the "killer," the jury likely would disregard much of, if not the whole of, his testimony. The trial prosecutor himself acknowledged that post-trial testimony had gutted Guevara's credibility. In a letter to the Texas Department of Criminal Justice, Pardons and Paroles Division, the trial prosecutor "protest[ed] any possible early parole" for Guevara because "[a]t trial [he] was very evasive and apparently not at all truthful." F.E.H., PX 16.

Moreover, correct forensic testing very likely would have prevented the jury from even hearing Guevara's testimony. Almost unquestionably, competent forensic testing would have (1) prompted the prosecution to reconsider its plea agreement with Guevara or (2) cause him to assert his Fifth Amendment rights.6 Guevara testified at a time he thought that science wholly laid the blame on his co-participant Williams. Once post-trial testing showed that the bullet in Collier's head came from Guevara's gun, he stopped talking. Guevara did not testify in state habeas court. When subpoenaed as a witness in federal court, Guevara, through counsel, moved to prevent his testimony claiming his right against self-incrimination. Had Williams' defense counsel investigated the true caliber of EB-1, Guevara would have been a codefendant, not a star witness. Even the trial prosecutor later bemoaned that "rather than being a witness, [Guevara] likely participated in Collier's murder." F.E.H., PX 16. Accordingly, this Court's inquiry into Strickland prejudice must ignore Guevara's trial testimony, either because he would not have testified had counsel performed adequately or because the jury likely would have deemed him altogether lacking in credibility.

Although the prosecution relied heavily on Guevara to prove Williams' guilt, uneasiness with his credibility prompted the prosecution to bolster its case through other witnesses. The prosecutor confidently stated in closing: "Vaal Guevara ... he's a criminal, he has prior convictions. If you don't want to believe anything that he told you, that's also fine. If I was you, perhaps I wouldn't believe anything he told me. That's up to you. I got plenty of evidence without him." Tr. at 31:52.7

The prosecution's confidence made sense when Guevara could testify that he did not shoot Collier. The prosecution claimed that testimony from the other witnesses inculpated Williams:

Tr. at 31:100. But with a correct understanding of EB-1's caliber and without Guevara's testimony, the entire landscape of trial would shift. With it, the value of eyewitness testimony would change. Now, their testimony would need to do more than to place the guns in Williams' hands; witnesses would have to provide information about Collier's death.8

Aside from Guevara, the prosecution assured the jury that there was "a bunch of other evidence that tends to connect the defendant to the offense[.]" Tr. at 31:55. Without Guevara, the prosecution had no supporting lay testimony about when Collier died. Raul, Anderson, and Winn all provided valuable testimony.9 They identified the four men who went across the street and excluded each other (and Smith) as the killer. They put the shotgun in Williams' hands. But they could not answer the most important remaining questions in this case. Without Guevara, the witnesses could not even sequence the shots fired. All of the other lay witnesses testified that they heard three shots, and opined that the shotgun was last. The forensic evidence, however, showed that the men fired at least four shots, two at Rasul and two at Collier. Witnesses could not account for when Guevara shot Collier with the .22. Nor could they see or hear Collier either before or after the shotgun blast. Their testimony would not have materially added to the question of causation.

With a correct identification of EB-1, and inevitable removal of Guevara's testimony from the jury's deliberations, the prosecution's focus would have had to be on expert testimony. Forensic and medical testimony would have become the emphasis of the prosecution, rather than substantial reliance on Guevara's account. The federal evidentiary hearing proved that the inevitable conflicting expert testimony would have provided difficult issues for the jury.

C. The Fatal Injuries

The prosecution's insistence that Williams fired twice into Collier's head framed the trial discussion of which was the killing gunshot. Dr. Brown's detailed and extensive testimony concerning his autopsy findings unflinchingly designated the shotgun as the cause of Collier's death. Dr. Brown not only discounted the possibility that EB-1 had killed Collier, he confessed that the mutilated slug had no influence at all on his opinion.

Trial counsel provided a substantively weak and emotionally tepid response. Trial counsel asked Dr. Brown only three questions: (1) whether the red ring he observed indicated blood pressure at the time of death; (2) "really you can't really account for ... that .25 caliber bullet?"; and (3) "Is a .25 caliber pistol a deadly weapon, also, sir?" Tr. at 29:133-38.

The defense provided no response to Dr. Brown's testimony. Trial counsel's closing briefly mentioned that "blood pressure could still be there" leaving the "red wound" because the shotgun blast occurred right at "the time of death." Tr. at 29:87-88. The trial evidence gave the jury nothing to consider but the prosecution's mocking insistence that EB-1 was inconsequential. Significantly, the evidence in the federal hearing demonstrated that EB-1 should have been a core issue at trial.

1. Federal Evidentiary Hearing Testimony

Two witnesses testified at the federal evidentiary hearing on August 10, 2010: Williams called Dr. Marc Andrew Krouse and Respondent called Dr. Vincent J.M. Di Maio. Both experts are highly regarded in their field and, while providing technical testimony that differed in some respects, both were credible witnesses. A review of their testimony confirms that the jury would have had to confront hard questions with a revised understanding of Collier's death.

Dr. Krouse, a board certified expert in anatomic and forensic pathology, testified previously in both state habeas evidentiary hearings. In preparation for his federal hearing testimony, he reviewed the autopsy report, crime scene photographs, his earlier testimony, and portions of the record. Williams called Dr. Krouse to testify concerning two matters: (1) the condition of EB-1 and (2) the effect of that projectile on the victim. Dr. Krouse explained that EB-1 was a jacketed small caliber projectile that, given its deformity, must have hit something hard. F.E.H. at 56. Dr. Krouse assumed that EB-1 became distorted when it struck Collier's head. F.E.H. at 56-57.

Dr. Krouse could not forensically sequence the gunshots, but thought that "it's a good assumption that [Collier] was struck with the handgun first and was on the ground ... when the shotgun was fired." F.E.H. at 61. Even though the autopsy did not mention an exit wound, Dr. Krouse "would love to see that skull" because he thought he could "find a separate entry wound for the .22 in the bone." F.E.H. at 61-62. He explained: "A .22 entry wound [in a] hair covered scalp is not real easy to see. And I think it was just missed." F.E.H. at 68. Dr. Krouse's testimony assumed that EB-1 struck Collier's head and entered his cranial cavity.10

Dr. Krouse identified several factors which indicated that Collier's body did not have blood pressure when hit by the shotgun blast. The crime scene photographs only showed a limited amount of blood around Collier's head—between one liter and a liter-and-a-half. F.E.H. at 57. If Collier had blood pressure, the head wound would have bled much more. F.E.H. at 57-58. The blood oozed from Collier's head in a waterfall pattern. Blood pressure would have caused blood to spurt from Collier's head in an arterial pattern, rather than to leak passively to the ground. The blood would have flowed, rather than seeped out, of his ear. F.E.H. at 57-58. The pattern of blood expelled from the wound by the shot would have been different if Collier had blood pressure, with a distinct back splatter. F.E.H. at 78.

With the assumption that EB-1 penetrated Collier's skull, Dr. Krouse would consider it a fatal injury. He opined: "Even with rapid medical response and intervention, that penetrating handgun wound to the head has a high potential for being lethal. And if treatment is delayed, then I think death becomes almost inevitable, certainly beyond a 80 to 90 percent probability." F.E.H. at 67. Dr. Krouse explained that as a projectile such as EB-1 "plows through the brain" it would cause "massive internal bleeding. Elevation of intercranial pressure, the pressure will build up, and blood flow to the brain will eventually cease." F.E.H. at 63. In that case, "the heart continues to beat, but it's not—it's not got anything to push." F.E.H. at 63. "He could have been effectively brain dead, moribund, mortally wounded, and still have a bit of a pulse and blood pressure, but effectively dead before he was shot with a shotgun[.]" F.E.H. at 66. Even with a brain-stem injury, "[t]he heart will beat for several minutes." F.E.H. at 64.11

He disputed Dr. Brown's trial testimony that the red ring observed around the shotgun wound meant that Collier was alive when shot. Dr. Krouse instead opined that the red-ringed entry only meant that it was a "peri-mortem injury. It's occurring right around the time of death, either immediately before, at the point of, or even slightly after physiologic death." F.E.H. at 87. Nevertheless, Dr. Krouse did not exclude the shotgun as a possible cause of Collier's death. He explained, "this man died of multiple firearm injuries without any—without further information, particularly regarding the putative entry wound of the handgun bullet, the .22 ... I would have to put both the handgun wound and the shotgun wound in a cause of death." F.E.H. at 60. Specifically, EB-1 could have been a fatal wound. F.E.H. at 60. In sum, Dr. Krouse would list both the shotgun and EB-1 as a cause of death. F.E.H. at 67 ("[G]iven the limited data that is available to us at this point in time without having an established entry wound for the .22, without having any potential of looking at fracturing eruptions or something else to sequence these wounds, objectively and certainly, that's why I'm saying I have to put both of the— both things as a cause of death.").

Respondent's cross-examination of Dr. Krouse attempted to poke holes in his opinion that EB-1 was a cause of Collier's death. Respondent asked Dr. Krouse questions tending to show that only speculation could account for what EB-1 did to Collier's brain without evidence of whether and where it entered his cranial cavity. F.E.H. at 96-97.12 On cross, Dr. Krouse admitted that Collier likely lived for a short period of time after being shot with EB-1. Dr. Krouse conceded that Collier's heart could have been beating when shot the second time, though he refused to speculate on whether Collier was "legally dead" or not. F.E.H. at 102-04. In the end, Dr. Krouse could not "exclude that [the shotgun] is a significant contributing factor" to Collier's death. F.E.H. at 105.

Respondent relied on testimony from Dr. Vincent J.M. Di Maio to show that Collier still had blood pressure, and was thus alive, when hit with the shotgun. Dr. Di Maio, who also had extensive experience with forensic pathology, largely based his opinions on the autopsy report and crime scene photographs. F.E.H. at 149. Dr. Di Maio testified that, at the time of the shotgun blast, "he was alive. His heart was beating. And he had blood pressure." F.E.H. at 150, 170.13 Dr. Di Maio pointed to several facts which indicated that Collier was alive when shot: (1) the wound had blood in the soft tissue, which would have had a yellow, parchment-like look if he were dead, F.E.H. at 151; (2) satellite injuries around the main wound from pellets that had separated from the shot column were red, suggesting that Collier had blood pressure, F.E.H. at 153-54; and (3) gravity and leakage could not account for the amount of blood in the wound, F.E.H. at 156-57.

The most prominent difference in the testimony of the two experts is how they would evaluate the existence of EB-1. Dr. Di Maio would not have listed the .22 bullet as a cause of death; rather, he "would list it as a finding." F.E.H. at 159. Dr. Di Maio premised that conclusion on the fact that the record did not show that EB-1 entered Collier's cranial cavity:

F.E.H. at 159. Recognizing the other errors in this case, Dr. Di Maio admitted:

F.E.H. at 162. Cross-examination clarified that EB-1 did not play any part in Dr. Di Maio's conclusions. Dr. Di Maio thought that he bullet entered Collier's head, but Dr. Di Maio did not observe anything in the record showing that the bullet had entered the cranial cavity. F.E.H. at 192-93. Nonetheless, Dr. Di Maio agreed: "A .22 magnum has high potential for fatality if you're struck in the head with it." F.E.H. at 208.

2. Effect of Expert Testimony on the Question of Causation

Two competent, credible, and knowledgeable experts provided useful testimony relating to what EB-1 and the shotgun did to Collier. To summarize, some points of agreement exist with the testimony from the experts. Both Dr. Di Maio and Dr. Krouse thought that Collier lived for some period of time after being shot with EB-1, although they disagreed on the length of that period. Both doctors agreed that Collier's heart may have been beating when shot the second time. The two experts would have listed the shotgun as "a" cause of death, but Dr. Di Maio would have listed it as the only one.

The two experts disagreed on many points concerning Collier's death, but most critically on how to interpret the existence of EB-1. Dr. Krouse assumed that EB-1 struck Collier in the head and entered his cranial cavity. Thus, Dr. Krouse would have labeled EB-1 "a" cause of death. Dr. Di Maio essentially refused to factor the EB-1 slug into his findings, despite the fact that the trial prosecutor conceded that the evidence "inferr[ed]" that EB-1 struck Collier in the head. Tr. at 26:25. As a subsidiary point of disagreement, their testimony differed on whether Collier's body had blood pressure when the shotgun blast hit him, using that as an indicator of life. Dr. Krouse opined that the victim was "moribund" or "effectively dead" when Williams shot him, but Dr. Di Maio definitely thought that he was alive. Both experts gave credible testimony and appeared to base their conclusions on scientific principles, as they perceived them to apply in this case. Nevertheless, they disagreed significantly in how to interpret evidence about the murder of Collier.

This Court's role is not to decide which expert was correct, but to evaluate how the testimony likely would have influenced the jury. Regardless of Dr. Di Maio's reluctance to opine definitively on what EB-1 did to Collier, had the doctor testified at trial, the jury would inevitably have had to factor EB-1 into the evidentiary landscape. There is no meaningful debate that the velocity of Guevara's pistol was enough to penetrate Collier's cranial cavity. The Court finds on de novo review that, with adequate defense investigation, Dr. Krouse would have testified at trial and the trial jury reasonably could have found that EB-1 likely penetrated Collier's skull and knocked him to the ground. With correct information about the potential effect of EB-1, the jury would have had to decide if reasonable doubt existed as to whether the shotgun was the source of one of two killing shots or the only killing shot. A reasonable juror could have believed Dr. Krouse's highly credible testimony and found that EB-1 caused a fatal wound. With the complex question of causation fully before the jury, a reasonable juror could easily have entertained reasonable doubts about Williams' guilt for capital murder.

D. Reasonable Probability of a Different Result

1. The Effect of Trial Counsel's Deficient Performance

Had trial counsel provided effective assistance, the jury would have had before it a more fulsome understanding of Guevara's role and his possible culpability. The jury would have had to entertain a complete alternate explanation for Collier's death. Independent testing of EB-1 would have, at a bare minimum, provided crucial tools for cross-examination of Guevara, the assistant medical examiner, and the ballistic forensic specialist who botched his investigation. With the assistance of forensic and medical experts, the defense could have poked serious and persuasive holes in the prosecution's case.

Respondent, nonetheless, argues that the evidence as developed after trial still allows a reasonable basis for a capital murder conviction. This argument misperceives the Strickland standard of proof. Strickland's prejudice prong is not concerned with whether, notwithstanding trial counsel's errors, sufficient evidence still exists to convict. Although a petitioner's burden under Strickland "is substantial, he is not required to establish his innocence or even demonstrate that counsel's deficient conduct more likely than not altered the outcome in the case." Soffar v. Dretke, 368 F.3d 441, 479 (5th Cir.2004) (quotation omitted); accord Porter v. McCollum, ___ U.S. ___, 130 S.Ct. 447, 455-56, 175 L.Ed.2d 398 (2009) ("We do not require a defendant to show that counsel's deficient conduct more likely than not altered the outcome of his penalty proceeding, but rather that he establish a probability sufficient to undermine confidence in that outcome.") (quotation omitted); Strickland, 466 U.S. at 693, 104 S.Ct. 2052 ("[W]e believe that a defendant need not show that counsel's deficient conduct more likely than not altered the outcome in the case."). The question is not whether enough salvageable threads of the prosecution's case remain to convict Williams. The "overriding interest," Williams v. Taylor, 529 U.S. 362, 392, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000), is whether counsel's errors "`were so serious that they rendered the proceedings fundamentally unfair or the result unreliable.'" Soffar, 368 F.3d at 471 (quoting Fretwell, 506 U.S. at 372, 113 S.Ct. 838). Trial counsel's deficient performance leaves this Court with no confidence in the process afforded to Williams.

Had the defense's trial preparation included the facts developed post-trial, the case before the jury would have differed fundamentally from what occurred. The focus would have shifted to allocating blame between co-defendants. With the correct factual scenario available, plea negotiations and jury instructions also would have taken different courses.

Respondent contends that Williams' conviction would have been an inevitable result of the prosecution reviewing the new totality of evidence and then requesting a law-of-the-parties jury instruction. The Court is unpersuaded. Strickland does not sanction such speculation.14 Moreover, it is unavailing to Respondent. The same hypothetical trial reconstruction suggests that the trial court likewise would have been requested to deliver—and likely would have delivered—a lesser-included-offense instruction, allowing for Williams' conviction of a crime less severe than capital murder. Respondent's argument reinforces the Court's conclusion that had the key forensic information been known to the parties before trial, the jury would have had to entertain materially different issues. In a more fulsome trial, Williams would still have appeared before the jury as a man who used violence while trying to steal drugs. But the jury would have had before it a much more complex scene where his culpability and Guevara's were intertwined and unclear. The vast majority of the prosecutor's arguments likely would not have been made, and certainly not made derisively as occurred in 1995. Given the opportunity to evaluate the evidence as developed on federal review, a reasonable juror easily could have had reasonable doubt about Williams' guilt of a capital offense. The Court finds that trial counsel's deficient performance prejudiced the defense.

2. State Forensic Errors

Trial counsel alone cannot be faulted for the circumstances that downplayed and mischaracterized Guevara's role in the murder. It is inescapable that the Houston Police Department's Firearms Laboratory set the stage for this drama to unfold. A duty to investigate crime competently cabins the State's enormous criminal law enforcement power. The State of Texas in this case, however, fumbled through its investigation of Collier's murder. The prosecution developed the evidentiary support for its case sloppily. These failings distorted trial counsel's understanding of the evidence before the jury. Had the State performed a competent forensic inquiry, the parties would have been able to debate before the jury Guevara's role as a potential killer, rather than a mere witness. Whether by inadvertence or by incompetence, the State's untrustworthy investigation hampered trial counsel's performance. Baldwin misidentified the type of pistol shot and the source of EB-1, and thus the shooter. The forensic investigation in this case was "at best extreme carelessness[.]" F.E.H., PX3(5) (Affidavit of Ronald Lopez Singer, dated April 15, 1998). The administration of criminal justice cannot countenance such fundamental errors.

These failings extended into the examination of Collier's corpse. Respondent now argues Guevara's lack of culpability, contending nothing shows how EB-1 came to rest in the victim's head among the shotgun pellets. Respondent, relying on the incomplete autopsy and forensic analysis, ignores the fact that EB-1 (by the State's own admission) was found among the shotgun pellets removed from Collier's head. Respondent instead suggests that EB-1 never penetrated Collier's cranial cavity. This flawed logic does not inure to the State's benefit.15 To the extent the location of EB-1 in Collier's head was unknown, the State investigators bear much of the responsibility. Indeed, Respondent's own expert lamented the careless manner in which the medical examiner's office handled the autopsy. Dr. Di Maio struggled to create scenarios in which the mutilated bullet could have entered the victim's head. Dr. Di Maio likewise expressed dismay because the taking of the x-rays should have disclosed the bullet's location, had the autopsy been performed competently. F.E.H. at 187-89. These failings by the medical examiner highlight the fact that the jury had a grossly incomplete, if not tainted, view of the circumstances surrounding Collier's death. Poor police investigation and inept autopsy procedures veiled the true events of the crime from appointed counsel.

This lack of professionalism extended into the courtroom. The prosecution approached trial with an unnecessarily mocking, derisive, and sneering tone. As this Court previously observed:

Memorandum and Order filed March 29, 2005, [Doc. #41] at 48 n. 50. Unfairly, irony dripped from the prosecution's burlesque of a defensive theory that correct forensic testing has proven to be true. This Court therefore has no confidence that the circumstances of the whole case represent a fair inquiry into Collier's murder and what crime Williams committed.

3. Cumulative Prejudice

Given the errors by the prosecution team and, most importantly, trial counsel, a reasonable probability exists that a juror would have harbored reasonable doubt about Williams' guilt. The murder occurred in a flurry of gunshots, with Guevara and Williams firing multiple rounds. Given the unrebutted testimony that put both the .25 caliber pistol and the shotgun in Williams' hands, Williams had to depend at trial on convincing the jury that he shot someone who had already been killed or who already had been shot by a fatal bullet. But this Court's concern is not whether Williams was guilty of murder as the shooter or under another theory, or of some lesser offense. This Court's duty is to decide whether trial counsel's deficiencies, as conceded by Respondent, prejudiced the defense.

Given the distorted factual scenario presented to the jury by the State, and without deference the state court's factfinding, this Court concludes that Williams has met Strickland's prejudice prong. This Court does not have confidence in the outcome of Williams' 1995 trial. It is very likely that, had the jury known the truth about the source of EB-1, the parties' arguments would have been materially different from what was presented, and at least one juror would have entertained reasonable doubt about Williams' guilt of capital murder. Williams has met Strickland's prejudice prong and shown that counsel's deficient performance undermines confidence in the outcome of his trial.

It is noted also that, in any event, if the Court were required to give AEDPA deference to the state court's legal conclusion, the Court would conclude based on the evidentiary record that the state court's decision was an unreasonable application of Strickland jurisprudence. The Court accordingly conditionally grants William's federal petition for a writ of habeas corpus.

IV. CONCLUSION

For the foregoing reasons, the Court concludes Williams' petition for a writ of habeas corpus has merit and ORDERS the following: