Bryant v. Commissioner of Correction

Opinion

The sole issue in this certified appeal is whether trial counsel for the petitioner, Bemale Bryant, rendered ineffective assistance by failing to present four independent witnesses whose testimony, the petitioner claims, would have supported a third party culpability defense and substantially impeached the evidence presented against the petitioner. The petitioner appeals from the judgment of the Appellate Court reversing the judgment of the habeas court, which had granted his petition for a writ of habeas corpus. On appeal, the petitioner claims that the Appellate Court improperly concluded that the performance of his trial counsel was *504not deficient. The petitioner contends that the defense counsel’s failure to present the witnesses constituted ineffective assistance of counsel, and that the habeas court properly rejected the argument of the respondent, the commissioner of correction, that it was a matter of trial strategy. We agree with the petitioner and reverse the judgment of the Appellate Court.

The Appellate Court’s decision on direct appeal in State v. Bryant, 71 Conn. App. 488, 802 A.2d 224 (2002), sets forth the facts, as found by the jury, that led to the petitioner’s conviction. “The jury reasonably could have found that on April 14,1996, the [petitioner] repeatedly and seriously injured [the victim] Edward Jones, and those injuries caused Jones’ death. . . .

“After drinking beer and using narcotics together, in the early morning horns . . . Gary Fournier and [Jones], drove to Irving Street and Albany Avenue in Hartford to get more narcotics. The pair planned to obtain the narcotics and drive off without paying for them. Fournier stopped his car and was approached by Terry ‘T-Time’ Davis, a known drug seller. Shortly thereafter, the [petitioner] approached the car and handed Fournier some cocaine.

“As soon as Fournier had the cocaine in hand, he drove off without paying for the contraband. The [petitioner] continued to hang onto the car as Fournier drove off. The [petitioner] released his hold of the car just as it ran through a stop sign and was struck by another automobile. The [petitioner] then went to Fournier’s car, dragged Fournier from the car, pushed him to the ground and kicked him several times before running to the passenger’s side of the car. The [petitioner] dragged Jones through the passenger window, and hit and kicked him repeatedly while he lay on the street. The [petitioner] then ran from the scene.” Id., 490. Jones *505subsequently was taken to the hospital, where he was pronounced dead.1 Id., 491.

The petitioner was charged with murder in violation of General Statutes § 53a-54a. After a jury trial, the petitioner was acquitted of the murder charge, but was found guilty of the lesser included offense of manslaughter in the first degree in violation of General Statutes § 53a-55 (a) (1). The petitioner appealed from the judgment of conviction to the Appellate Court, which affirmed his conviction.2 See id., 488. Upon the Appellate Court’s affirmance of his conviction, this court denied the petitioner’s petition for certification to appeal. State v. Bryant, 261 Conn. 939, 808 A.2d 1133 (2002).

Subsequently, the petitioner filed a petition for a writ of habeas corpus alleging that his confinement was illegal because he had been denied the effective assistance of counsel. The petitioner based his claim principally on his trial counsel’s failure to present four *506independent witnesses at trial whose testimony would have supported an alternative theory of the cause of Jones’ death. That theory did not implicate the petitioner, but instead implicated a small group of unidentified Hispanic men with a gun traveling in a white automobile. Trial counsel did not present this third party culpability defense to the jury. More specifically, the petitioner alleged that counsel should have called the following witnesses to testify to support that theory: Thomas Davis, the driver of the automobile that was struck by Fournier’s vehicle; Melissa Young-Duncan, an emergency medical technician who responded to the scene and assisted Jones; John Gartley, a second emergency medical technician who responded to the scene and assisted Jones; and Rene Fleury, Fournier’s girlfriend, who spoke with Fournier shortly after the incident. The habeas court found that the petitioner had been denied effective assistance of counsel under Strickland v. Washington, 466 U.S. 668, 686, 104 S. Ct. 2052, 80 L. Ed. 2d 674 (1984), because “the missing testimony could easily have led a jury to harbor a reasonable doubt as to the guilt of the petitioner.” (Emphasis in original.) Accordingly, the habeas court set aside the petitioner’s conviction, and ordered a new trial.

The respondent filed a petition for certification to appeal, which the habeas court denied. Thereafter, the respondent appealed to the Appellate Court, claiming, inter alia, that the habeas court: (1) abused its discretion by denying the petition for certification to appeal; and (2) improperly concluded that the petitioner’s trial counsel was ineffective by failing to present a theory of defense that was not supported by forensic evidence or the petitioner’s testimony. The Appellate Court agreed and reversed the judgment of the habeas court. Bryant v. Commissioner of Correction, 99 Conn. App. 434, 441-44, 914 A.2d 585 (2007). We granted the petitioner’s petition for certification to appeal limited to *507the following issue: “Did the Appellate Court properly reverse the habeas court’s decision to grant the petitioner’s writ of habeas corpus?” Bryant v. Commissioner of Correction, 282 Conn. 910, 922 A.2d 1098 (2007).

The following additional relevant facts found by the habeas court are necessary for our resolution of the petitioner’s claim that he was rendered ineffective assistance of counsel. The habeas court heard testimony from the four witnesses whom the petitioner claims should have testified at trial: Davis, Young-Duncan, Gar-tley and Fleury. After hearing their testimony, the habeas court found that “[h]ad [these witnesses] testified at the petitioner’s trial, it is reasonably probable that the jury could have found the following scenario to be true”: Davis, a Marine Corps veteran and gunnery specialist, employed in the Hartford area as a security guard, was driving in a company vehicle the evening of April 14, 1996. As he approached the intersection of Albany Avenue and Irving Street, Davis heard gunshots and identified them as originating from a small caliber weapon, probably a .22 caliber pistol. Immediately thereafter, Davis observed a blue Ford Escort (Ford),3 traveling in the wrong direction, exit from Irving Street, a one-way street, onto Albany Avenue. The driver of the Ford appeared to be slumped over the wheel. A white Cadillac or Lincoln, also traveling in the wrong direction on Irving Street, pursued the Ford. Davis’ vehicle collided with the Ford. Immediately after the accident, he observed the white vehicle stop and a light skinned Hispanic man exit the rear passenger seat of the vehicle.4 The man approached Davis’ vehicle carrying an object in his hand, which Davis could not see. Having heard the gunshots and believing the neighborhood to be dangerous, Davis displayed his own firearm, upon which the man went back to the white vehicle and it *508departed. At no time did Davis observe anyone remove the occupants from the Ford; nor did he observe anyone beat or kick them. Prior to trial, the investigator who worked for the petitioner’s trial counsel had contacted and interviewed Davis, who related these events to the investigator. Davis also reported the gunshot he heard to the Hartford police department, but the police did not include his statement in their incident report. The habeas court further found that Davis was available to be subpoenaed by either party at the original trial.

The court also found that a jury could have determined that Young-Duncan and Gaxtley are both experienced emergency medical technicians who were serving as an ambulance crew on the night of the incident. After receiving a call to assist at an automobile accident on the comer of Albany Avenue and Irving Street, they promptly arrived at the scene and provided the initial medical treatment to Jones. Both Young-Duncan and Gartley noticed what appeared to them, based on their training and experience, to be a gunshot wound to Jones’ left temple. Specifically, they identified the wound as a gunshot wound based on its small size and round shape as well as what appeared to be a powder bum. Furthermore, on the other side of Jones’ head, there appeared to be a powder bum that was consistent with an exit wound. Young-Duncan brought the gunshot wound to the attention of an unidentified Hartford police officer at the scene, who agreed that it appeared to be a gunshot wound.

The habeas court additionally found that a jury could have determined that Fleury, who was not at the scene but who was the owner of the Ford involved in the accident, spoke with Fournier after he had arrived home from the hospital. When she inquired about what had happened, Fournier told her that “there had been an incident with three Hispanic males and a gun.”

*509The petitioner’s trial counsel, David Smith, testified at the habeas hearing that he was aware of these four witnesses, and had reviewed all of their statements prior to the petitioner’s trial. Smith testified that he decided not to “[create] a specter of manslaughter with a firearm” by presenting these witnesses for two reasons: (1) he determined that there was no evidence to support the theory of the involvement of a gunman because there was no forensic evidence of a gunshot wound; and (2) he did not want to introduce the possibility that a gun had been involved in the incident for fear that it might “increase the likelihood of a penalty range” for the petitioner. Smith offered this justification, however, to explain only his failure to call Young-Duncan and Gartley as witnesses.

As to Davis, Smith testified that he did not call him because Davis could not be located.5 Smith also appeared to believe that Davis’ testimony could not assist in the defense of the petitioner. Specifically, Smith testified that “Davis, although he was there [at the scene], his existence or nonexistence either way did not . . . hurt [the petitioner] because he didn’t say he saw someone else do something. He didn’t say he saw [the petitioner] do anything.” As to Fleury, Smith testified that he did not present her testimony because “[s]he was not a witness to the case. . . . [S]he didn’t have any impeachment knowledge or evidence that we could have used.”

We begin with the applicable standard of review and the law governing ineffective assistance of counsel claims. “The habeas court is afforded broad discretion in making its factual findings, and those findings will not be disturbed unless they are clearly erroneous. . . . *510The application of the habeas court’s factual findings to the pertinent legal standard, however, presents a mixed question of law and fact, which is subject to plenary review.” (Internal quotation marks omitted.) Sastrom v. Mullaney, 286 Conn. 655, 661, 945 A.2d 442 (2008).

“A criminal defendant is constitutionally entitled to adequate and effective assistance of counsel at all critical stages of criminal proceedings. Strickland v. Washington, supra, 466 U.S. 686. This right arises under the sixth and fourteenth amendments to the United States constitution and article first, § 8, of the Connecticut constitution.” Copas v. Commissioner of Correction, 234 Conn. 139, 153, 662 A.2d 718 (1995). “As enunciated in Strickland v. Washington, supra, 687, this court has stated: It is axiomatic that the right to counsel is the right to the effective assistance of counsel. ... A claim of ineffective assistance of counsel consists of two components: a performance prong and a prejudice prong. To satisfy the performance prong . . . the petitioner must demonstrate that his attorney’s representation was not reasonably competent or within the range of competence displayed by lawyers with ordinary training and skill in the criminal law. ... To satisfy the prejudice prong, a claimant must demonstrate that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. . . . The claim will succeed only if both prongs are satisfied.” (Citations omitted; internal quotation marks omitted.) Sastrom v. Mullaney, supra, 286 Conn. 662.

The habeas court found the credibility of Davis, Young-Duncan, Gartley and Fleury to be “considerable and compelling” because all four were neutral witnesses who were not meaningfully impeached at the *511habeas hearing.6 As a result, the court found that “a jury likewise would have found their testimony to be credible and highly persuasive.” On the basis of this testimony, and pursuant to Strickland v. Washington, supra, 466 U.S. 686, the habeas court concluded that “it was deficient performance on the part of trial defense counsel not to present this testimony at the petitioner’s original trial,” and that “it was harmful to the petitioner and constituted inadequate representation to avoid introducing available and credible evidence of a clearly exculpatory nature in an ill-advised effort to avoid any mention of a firearm.” The court also concluded that, in the absence of the deficient performance by trial counsel, the exculpatory testimony of these four witnesses could have “induce [d] reasonable doubt in the minds of the jury” and, therefore, “the reliability of the petitioner’s conviction has . . . been undermined.”

On appeal, the Appellate Court concluded that the petitioner had failed to satisfy the performance prong under Strickland v. Washington, supra, 466 U.S. 687, because he could not demonstrate that his trial counsel’s representation fell below an objective standard of reasonableness. Instead, the Appellate Court determined that trial counsel decided not to call the four witnesses as a matter of trial strategy and held that the habeas court had failed to accord any deference to Smith’s tactical decision or his perspective at the time *512of trial. Bryant v. Commissioner of Correction, supra, 99 Conn. App. 443-44. We disagree and conclude that the decision not to call these four witnesses was not a matter of sound trial strategy within the wide range of reasonable professional assistance.

We first address Smith’s actions in light of the performance prong. “To satisfy the performance prong, a claimant must demonstrate that counsel made errors so serious that counsel was not functioning as the counsel guaranteed ... by the [s]ixth [a]mendment.” (Internal quotation marks omitted.) State v. Brown, 279 Conn. 493, 525, 903 A.2d 169 (2006). “In any case presenting an ineffectiveness claim, the performance inquiry must be whether counsel’s assistance was reasonable considering all the circumstances. Prevailing norms of practice . . . are guides to determining what is reasonable, but they are only guides. No particular set of detailed rules for counsel’s conduct can satisfactorily take account of the variety of circumstances faced by defense counsel or the range of legitimate decisions regarding how best to represent a criminal defendant. . . .

“Judicial scrutiny of counsel’s performance must be highly deferential. It is all too tempting for a defendant to second-guess counsel’s assistance after conviction or adverse sentence, and it is all too easy for a court, examining counsel’s defense after it has proved unsuccessful, to conclude that a particular act or omission of counsel was unreasonable. ... A fair assessment of attorney performance requires that every effort be made to eliminate the distorting effects of hindsight, to reconstruct the circumstances of counsel’s challenged conduct, and to evaluate the conduct from counsel’s perspective at the time. Because of the difficulties inherent in making the evaluation, a court must indulge a strong presumption that counsel’s conduct falls within the wide range of reasonable professional assistance; *513that is, the defendant must overcome the presumption that, under the circumstances, the challenged action might be considered sound trial strategy. . . .

“Thus, a court deciding an actual ineffectiveness claim must judge the reasonableness of counsel’s challenged conduct on the facts of the particular case, viewed as of the time of counsel’s conduct. A convicted defendant making a claim of ineffective assistance must identify the acts or omissions of counsel that are alleged not to have been the result of reasonable professional judgment. The court must then determine whether, in light of all the circumstances, the identified acts or omissions were outside the wide range of professionally competent assistance.” (Citations omitted; internal quotation marks omitted.)Stricklands. Washington, supra, 466 U.S. 688-90.

Accordingly, the question of whether Smith’s actions fell below an objective standard of reasonableness turns on whether his decision not to solicit the testimony of the four witnesses to support the third party culpability defense can be considered sound trial strategy, or whether it constitutes a serious deviation from the actions of an attorney of ordinary training and skill in criminal law. We need not decide whether Smith’s failure to elicit the testimony of any particular witness would suffice because Strickland directs us to look at the “totality of the evidence before the judge or jury”; id., 695; keeping in mind that “[s]ome errors . . . have ... a pervasive effect on the inferences to be drawn from the evidence, altering the entire evidentiary picture . . . .” (Internal quotation marks omitted.) Lindstadt v. Keane, 239 F.3d 191, 199 (2d Cir. 2001), quoting Strickland s. Washington, supra, 466 U.S. 695-96; see also Pavel v. Hollins, 261 F.3d 210, 228 (2d Cir. 2001) (cumulative weight of flaws deprived defendant of effective assistance of counsel). We therefore consider Smith’s failure to present the mosaic of the alternative *514theory, a third party culpability defense, as supported by the four witnesses.

We first review the standards governing the admissibility of third party culpability evidence, mindful that the petitioner cannot prevail unless he demonstrates that his alternative theory would not be excluded. “It is well established that a defendant has a right to introduce evidence that indicates that someone other than the defendant committed the crime with which the defendant has been charged. . . . The defendant must, however, present evidence that directly connects a third party to the crime. ... It is not enough to show that another had the motive to commit the crime . . . nor is it enough to raise a bare suspicion that some other person may have committed the crime of which the defendant is accused.” (Internal quotation marks omitted.) State v. Arroyo, 284 Conn. 597, 609, 935 A.2d 975 (2007).

“The admissibility of evidence of third party culpability is governed by the rules relating to relevancy. ” (Internal quotation marks omitted.) State v. Ortiz, 252 Conn. 533, 564, 747 A.2d 487 (2000). “Relevant evidence is evidence having any tendency to make the existence of any fact that is material to the determination of the proceeding more probable or less probable than it would be without the evidence.” (Internal quotation marks omitted.) State v. West, 274 Conn. 605, 625, 877 A.2d 787, cert. denied, 546 U.S. 1049, 126 S. Ct. 775, 163 L. Ed. 2d 601 (2005); Conn. Code Evid. § 4-1. Accordingly, in explaining the requirement that the proffered evidence establish a direct connection to a third party, rather than raise merely a bare suspicion regarding a third party, we have stated: “Such evidence is relevant, exculpatory evidence, rather than merely tenuous evidence of third party culpability [introduced by a defendant] in an attempt to divert from himself the evidence of guilt.” (Internal quotation marks omitted.) State v. *515Smith, 280 Conn. 285, 304, 907 A.2d 73 (2006). “In other words, evidence that establishes a direct connection between a third party and the charged offense is relevant to the central question before the jury, namely, whether a reasonable doubt exists as to whether the defendant committed the offense. Evidence that would raise only a bare suspicion that a third party, rather than the defendant, committed the charged offense would not be relevant to the jury’s determination. Atrial court’s decision, therefore, that third party culpability evidence proffered by the defendant is admissible, necessarily entails a determination that the proffered evidence is relevant to the jury’s determination of whether a reasonable doubt exists as to the defendant’s guilt.” State v. Arroyo, supra, 284 Conn. 609-10.

It is not ineffective assistance of counsel, however, to decline to pursue a third party culpability defense when there is insufficient evidence to support that defense. See Dunkley v. Commissioner of Correction, 73 Conn. App. 819, 827, 810 A.2d 281 (2002) (no evidence to support third party claim, in part, because no one at scene implicated alleged third party), cert. denied, 262 Conn. 953, 818 A.2d 780 (2003); see also Floyd v. Commissioner of Correction, 99 Conn. App. 526, 531-32, 914 A.2d 1049 (insufficient evidence to substantiate third party claim when predicated on alleged testimony of unlocated drug dealers who were also gang members), cert. denied, 282 Conn. 905, 920 A.2d 308 (2007); Santiago v. Commissioner of Correction, 87 Conn. App. 568, 591-92, 867 A.2d 70 (third party statements did not contain sufficient substance to support viable third party claim), cert. denied, 273 Conn. 930, 873 A.2d 997 (2005); Alvarez v. Commissioner of Correction, 79 Conn. App. 847, 851, 832 A.2d 102 (insufficient evidence to support third party culpability defense when petitioner called only one witness at habeas hearing who did not even observe shooting), cert. denied, 266 Conn. *516933, 837 A.2d 804 (2003); Daniel v. Commissioner of Correction, 57 Conn. App. 651, 684, 751 A.2d 398 (testimony not sufficient to raise third party culpability defense because supporting witnesses’ statements were inconsistent), cert. denied, 254 Conn. 918, 759 A.2d 1024 (2000).

In contrast, in the present case, the habeas court found that the testimony of Davis, Young-Duncan, Gar-tley and Fleury would have worked in concert to create a credible scenario in which the cause of Jones’ death was a gunshot wound to the head perpetrated by a small group of unidentified Hispanic males driving a white Cadillac or Lincoln, not the actions of the petitioner. First, Davis’ testimony supports three assertions: (1) that prior to the collision, Davis heard a gunshot, possibly from a .22 caliber weapon; (2) that Jones’ car was being pursued by a white Cadillac or Lincoln; and (3) that after the collision, an unidentified Hispanic man, with an unidentified object in his hand, exited the white vehicle and briefly approached the accident scene. Moreover, his testimony also undermines the state’s theory of the case. Davis was present at the crime scene from the point of impact until the time the police arrived. Yet during that period, Davis did not witness the supposedly savage and deadly beating, which, according to the state’s theory, occurred at the same accident scene.7 Second, Young-Duncan and Gar-*517tley were two trained emergency medical technicians. On the basis of their testimony, a jury reasonably could have concluded that Jones had sustained a gunshot wound to the left temple area of the head, and that the autopsy performed by the state’s medical examiner, Arkady Katsnelson, was potentially incomplete or inaccurate. Finally, Fleury’s testimony explicitly links the testimony of Davis, with that of Young-Duncan and Hartley. Fleury testified that when Fournier arrived home that night, he stated that he had been involved in “an incident with three Hispanic males and a gun.”8 Moreover, there was no evidence in the record to suggest that any of these witnesses’ statements, regarding either the gunshot or the Hispanic males, were influenced by the statements made by the other witnesses regarding those similar observations. When reviewed in its totality, the testimony of these neutral witnesses, each of whom the habeas court found to be credible and highly persuasive, creates a plausible, well supported third party culpability defense. That is, the testimony of these four disinterested witnesses, when considered as a whole, raises more than a “bare suspicion” of third party culpability. State v. Arroyo, supra, 284 Conn. 610. Instead, the testimony directly connects a third party to the crime—the small group of unidentified Hispanic males.9 We therefore conclude that under our third party culpability rules, the testimony tending to show that the unidentified Hispanic males with a gun were responsible for Jones’ death would have been relevant to the jury’s determination of whether there existed a reasonable doubt as to the petitioner’s guilt. Therefore, the failure to present this relevant, plausible *518third party culpability defense constituted deficient performance on the part of defense counsel under, Strickland.

In addition, we conclude that in circumstances that largely involve a credibility contest, as did the petitioner’s trial, “the testimony of neutral, disinterested witnesses is exceedingly important.” Williams v. Washington, 59 F.3d 673, 682 (7th Cir. 1995). While each habeas petition for ineffective assistance of counsel relies on its own unique set of facts, various state and federal decisions have addressed analogous circumstances. For example, in Lindstadt v. Keane, supra, 239 F.3d 203, the United States Court of Appeals for the Second Circuit concluded that the performance of the defendant’s counsel was deficient, namely, because he failed to enter the testimony of two neutral probation officers who would have testified that the defendant’s wife, who was the state’s chief witness, had made numerous, unsubstantiated allegations of abuse against the defendant in the hopes of having him reincarcerated.10 See also Pavel v. Hollins, supra, 261 F.3d 222-23 (in “credibility contest,” performance deficient, in part, because of failure to call neutral court-appointed mediator to support theory that mother manipulated children’s testimony in effort to gain full custody); Siano v. Warden, 31 Conn. App. 94, 100-105, 623 A.2d 1035 (in conviction based on credibility of testimony of coconspirator, deficient performance for failure to call physician as lone neutral witness to testify that defen*519dant was incapable of crime due to extensive injuries from recent motor vehicle collision), cert. denied, 226 Conn. 910, 628 A.2d 984 (1993).

At trial in the present case, the state’s only evidence that specifically connected the petitioner to the assault was predicated entirely on the credibility of the eyewitness testimony of Fournier and Ewan Sharp.11 Fournier admitted that he had lied to the police during the investigation and that he had consumed approximately eighteen beers and an unknown quantity of cocaine on the night of the incident. Detective Steven Grabowski of the Hartford police department testified that, at the time, Fournier was “very uncooperative [and] intoxicated . . . .” A jury could have found that Sharp’s testimony was equally suspicious. Sharp did not come forward until four years after the fact, and only while undergoing police interrogation for an unrelated felony. Moreover, the testimony of Fournier and Sharp was inconsistent as to critical facts: Fournier testified that the petitioner assaulted him in addition to attacking Jones, whereas Sharp testified that the petitioner did not assault Fournier. In the absence of the third party culpability evidence presented by neutral, credible witnesses, however, the jury was not presented with a plausible alternative to Sharp and Fournier’s descriptions of the events.

In short, Smith’s performance was deficient. The alternative theory raising a third party culpability *520defense was well supported by neutral, credible witnesses,12 whose statements were made contemporaneous to the events.13 This was not an alternative theory concocted years later by a defendant in a jailhouse cell, unsupported by credible evidence. Under our third party culpability rules, we conclude that the evidence was relevant, and could have raised a reasonable doubt in the minds of the jury as to the petitioner’s guilt. The testimony would have called into question the most basic elements of the state’s case: (1) that the petitioner was the individual who killed Jones; and (2) that Jones died as a result of a beating. In the light of all the circumstances, including the dubious credibility of the state’s two eyewitnesses, the decision not to present the third party culpability defense fell below an objective standard of reasonableness, and, therefore, constituted deficient performance under the principles enunciated in Strickland.14

*521We are not persuaded by the state’s argument that deference to trial strategy saves Smith’s actions. While we acknowledge that the decision whether to call a particular witness falls into the realm of trial strategy, which is typically left to the discretion of trial counsel; see State v. Gore, 288 Conn. 770, 779 n.9, 955 A.2d 1 (2008); it does not follow necessarily that, in every instance, trial counsel’s strategy concerning these decisions is sound. As we have demonstrated previously in this opinion, in light of an objective standard, Smith’s decision not to present these witnesses was unreasonable.15

*522“An error by counsel, even if professionally unreasonable, does not warrant setting aside the judgment of a criminal proceeding if the error had no effect on the judgment.” Strickland v. Washington, supra, 466 U.S. 691. To satisfy the second prong of Strickland, that his counsel’s deficient performance prejudiced his defense, the petitioner must establish that, as a result of his trial counsel’s deficient performance, “there remains a probability sufficient to undermine confidence in the verdict that resulted in his appeal.” Bunkley v. Commissioner of Correction, 222 Conn. 444, 454, 610 A.2d 598 (1992), overruled in part on other grounds by Small v. Commissioner of Correction, 286 Conn. 707, 946 A.2d 1203 (2008). The second prong is thus satisfied if the petitioner can demonstrate that there is “a reasonable probability that, but for that ineffectiveness, the outcome would have been different.” Summerville v. Warden, 229 Conn. 397, 430, 641 A.2d 1356 (1994); see also Siano v. Warden, supra, 31 Conn. App. 98.

Because the Appellate Court concluded that the petitioner had failed to satisfy the performance prong of Strickland, it did not reach the question of prejudice. The habeas court, for its part, concluded that, based on its findings, “there is clearly sufficient reason to doubt the reliability of [the jury’s] verdict, based as it is upon incomplete evidence.” We agree.

*523Foremost, while it is not necessary to repeat our analysis addressing the performance prong, suffice it to say that the third party culpability defense, as supported by the testimony of Davis, Young-Duncan, Gar-tley and Fleury, likely would have permeated to some degree every aspect of the trial and raised a reasonable doubt in the minds of the jury as to the petitioner’s guilt.16 See Strickland v. Washington, supra, 466 U.S. 695-96 (“[s]ome errors will have had a pervasive effect on the inferences to be drawn from the evidence, altering the entire evidentiary picture”). That theory, when juxtaposed with the testimony of Fournier and Sharp, creates great doubt in the validity of the jury’s verdict. See id., 696 (“a verdict or conclusion only weakly supported by the record is more likely to have been affected by errors than one with overwhelming record support”).

As we previously have noted, in the present case, the only evidence directly connecting the petitioner to the attack was the eyewitness testimony of Fournier and Sharp. The record reveals, however, that both witnesses were subject to substantial impeachment evidence. First, Sharp never made a statement to police on the night of the incident, but rather made his statement four years later and only during an interrogation by the Hartford police department subsequent to an unrelated felony arrest. Only at that point did Sharp tell the police that he had witnessed the petitioner beating Jones.17 *524Sharp then testified at the petitioner’s probable cause hearing in a manner largely consistent with the facts, as noted previously, that the Appellate Court relied on to reverse the habeas court. When it was Sharp’s turn to testify at trial, however, Sharp refused. After the prosecutor offered Sharp transactional immunity, Sharp continued to refuse to testify and subsequently was held in contempt of court. Thereafter, the state read Sharp’s probable cause testimony into the record. Second, at the time of the probable cause hearing, Sharp testified that he had four prior felony convictions, and had charges pending against him. Finally, Sharp testified that, at the time he allegedly witnessed the events, he was fourteen years old and was under the influence of marijuana.

Fournier’s testimony was also vulnerable to impeachment evidence. To begin, on the night in question, Four-nier also did not make any statements to police asserting that the petitioner had attacked Fournier or Jones. When police recontacted Fournier on or about February 21, 1997, he again failed to make any statements alleging an attack by the petitioner.18 While on the witness stand, Fournier testified that he had at least one prior felony conviction and had various charges pending against him. In addition, Fournier admitted that at the time of the incident it was his and Jones’ intention to solicit the sale of cocaine and depart without paying for the contraband. Even more remarkable, Fournier testified on cross-examination that he had consumed as many as eighteen beers and an unknown quantity of cocaine on the night of the incident. Fournier then testified that he had lied to the police on multiple occasions during their investigation, specifically, that he had been the passenger and not the driver of the Ford,19 *525and that he had given a false name to the police to avoid arrest on an outstanding warrant. Most importantly, as noted previously, the testimony of Sharp and Fournier diverged in one significant respect: Fournier testified that the petitioner had kicked him several times prior to the petitioner’s alleged beating of Jones, whereas Sharp testified that he witnessed the entire incident and that Fournier “didn’t get beat up. He was down by the package store.”

Accordingly, not only was the testimony that linked the petitioner to the attack of dubious credibility, it also was internally inconsistent with respect to significant facts. While Katsnelson’s testimony and report admittedly were strong evidence, they were not conclusive of the petitioner’s guilt. Without the jury crediting the testimony of Sharp and Fournier, Katsnelson’s conclusion that Jones died of blunt trauma, assuming it is correct, only supports the conclusion that Jones died of blunt trauma—it does not prove that the petitioner committed the attack. Nothing in the autopsy report permitted Katsnelson to point to the petitioner as the assailant, nor did Katsnelson do so at trial. As a result, the failure to challenge the testimony of Fournier and Sharp with neutral, disinterested testimony was exceedingly damaging to the petitioner’s defense.

Moreover, Katsnelson’s testimony was not irrefutable. Though Katsnelson observed various injuries suffered by Jones that he concluded were inconsistent with a motor vehicle accident,20 and testified that he did not find evidence of a gunshot wound,21 the failure *526to call Yomg-Dmcan and Gartley left Katsnelson’s conclusion largely uncontested.22 Without that testimony, the jury never had to reconcile how two trained emergency medical technicians observed a pencil sized entry and exit gunshot womd to the head moments after their arrival at the scene. Perhaps most importantly, while it is difficult to understand how the autopsy did not reveal a gunshot womd, it is equally difficult to explain how Davis, a former Marine and disinterested witness, who saw the entire incident, and testified that at no time did he lose consciousness, never saw a supposedly savage and deadly beating taking place in his presence.

On the basis of the foregoing, we conclude that but for the deficient performance of the petitioner’s trial counsel, there was a reasonable probability sufficient to undermine confidence in the jury’s verdict; Strickland v. Washington, supra, 466 U.S. 694; and, as a result, the petitioner received ineffective assistance of comsel in violation of his rights mder the sixth and fourteenth amendments to the United States constitution and article first, § 8, of the Comecticut constitution.23

*527The judgment of the Appellate Court is reversed and the case is remanded to that court with direction to affirm the judgment of the habeas court.

In this opinion NORCOTT and VERTEFEUILLE, Js., concurred.