We disagree with the district court's conclusion that the defense did not exercise due diligence to discover the statement regarding Cookie's involvement in the crimes. Graves' counsel had specifically requested the information disclosed in the statement. We view Sebesta's statement regarding the polygraph, his discovery responses and questioning of Carter as misleading and a deliberate attempt to avoid disclosure of evidence of Cookie's direct involvement. At a minimum, Sebesta's minimal disclosure was insufficient to put the defense on notice to inquire further, particularly in light of the state's discovery disclosure.
 
Exculpatory?

Graves next challenges the district court's conclusion that the statement regarding Cookie's involvement is not exculpatory  [*22]  because the statement implicated Graves as well. n8 The district court found that the statement is not exculpatory because it implicated Graves based on the government's three person theory. It also found that the statement would have contradicted the testimony of one of Graves' witnesses, Tametra Ray, who testified that Cookie was home at the time of the murders. Again, we disagree.

The statement regarding Cookie's direct involvement in the crime is exculpatory for several reasons. First,  [*23]  each party's theory about how many people were actively involved in the crime is just a theory based on the number of people killed and the number of weapons used. The defense had submitted that two people were probably involved and had specifically requested any information related to any party, other than Graves and Carter, who the state alleged was involved in the crime. Although Cookie had been indicted, the defense viewed the indictment as a tool to pressure Carter into testifying. As we noted in our prior opinion, "if Graves had been furnished with Carter's statement, it could have provided him with an argument that those two persons were Carter and his wife rather than Carter and Graves." Graves II, 351 F.3d at 159. Also, Carter's statement, placing Cookie directly at the scene and actively involved in the murders, puts his deal with the state to testify only on the condition that he not be questioned about Cookie's involvement in a different light. It provides a stronger argument to Graves that Carter was lying about Graves involvement to save Cookie.

The district court did not reach the issue of materiality of the statement. That issue will be discussed in  [*24]  the following section regarding the effect of the two statements considered together.

c. The statements considered together?

The sole remaining issue under Graves' Brady claim is whether, considered together, the two statements - Carter's claim that he did it himself and Carter's statement directly implicating his wife Cookie in the murders - are material. We conclude that they are. If both statements had been timely furnished to Graves, he could have persuasively argued that (1) the murders were committed by Carter alone or by Carter and Cookie; and (2) Carter's plan from the beginning was to exonerate Cookie, but a story that he acted alone was not believable, so he implicated Graves so the prosecution would accept his story and decline to prosecute Cookie.

The state argues that the combined statements are not material because they are inconsistent and could have been damaging to Graves if the jury believed that the most credible account of the murders involved three killers, Carter, Cookie and Graves. The problem with the state's argument is that it analyzes the significance of the suppressed evidence against a backdrop of how the defense presented its case at trial  [*25]  without the suppressed statements. If the two statements had been revealed, the defense's approach could have been much different (as set forth above) and probably highly effective.

Case law from the Supreme Court is supportive of a finding of materiality on these facts - particularly because the case against Graves rests almost entirely on Carter's testimony and because the state presented testimony inconsistent with the two suppressed statements. In Giglio v. United States, 405 U.S. 150, 92 S. Ct. 763, 31 L. Ed. 2d 104 (1972), the Supreme Court reversed the defendant's judgment of conviction and remanded for a new trial because the prosecutor failed to disclose a promise of leniency to a key witness. The court concluded that the suppression affected the co-conspirator's credibility which was an important issue in the case and therefore material.

In Banks v. Dretke, 540 U.S. 668, 124 S. Ct. 1256, 157 L. Ed. 2d 1166 (2004), the Supreme Court reversed this court's denial of COA to the defendant on his Brady claim. The state withheld evidence that would have allowed defendant to show that two essential prosecution witnesses had been coached by police and prosecutors before they  [*26]  testified and also that they were paid informants. In addition, prosecutors allowed testimony that they were not coached to stand uncorrected at trial. In Kyles v. Whitley, 514 U.S. 419, 115 S. Ct. 1555, 131 L. Ed. 2d 490 (1995), the defendant's conviction was reversed and remanded for a new trial. The prosecution had suppressed statements of key witnesses and an informant who were not called to testify resulting in a Brady violation because their statements had significant impeachment value. Graves' case presents a cumulation of the elements found violative of a defendant's right to exculpatory evidence in the above cases.

IV.

Because the state suppressed two statements of Carter, its most important witness that were inconsistent with Carter's trial testimony, and then presented false, misleading testimony at trial that was inconsistent with the suppressed facts, we have no trouble concluding that the suppressed statements are material. Carter made several inconsistent statements throughout the investigation and pre-trial period. In some he denied all involvement, in some he implicated himself and Graves, and then, just before he testified against Graves, he gave the statements  [*27]  at issue in this appeal accepting full responsibility as the sole murderer and another statement placing his wife Cookie as an active participant in the murders. If the defense had known about the statement placing Cookie at the scene and given Carter's continuing condition that he would only testify if he were not asked about Cookie's involvement, the defense could have explained every statement implicating Graves as a means of protecting Cookie. As indicated above, these statements are particularly important in this case because Graves' conviction rests almost entirely on Carter's testimony and there is no direct evidence linking him with Carter or with the murder scene other than Carter's testimony. In addition, Carter's statement that he committed the crimes alone is important as the only statement he made exculpating Graves while implicating himself. The combination of these facts leads us to conclude "that the favorable evidence could reasonably be taken to put the whole case in such a different light as to undermine confidence in the verdict." Kyles, 314 U.S. at 435. Stated differently, disclosure of the statements "would have resulted in a markedly weaker case  [*28]  for the prosecution and a markedly stronger one for the defense." Id. at 441.

Under Delaware Rule of Evidence 401, the proffered evidence is relevant if it tends to make the existence of the defendant's guilt more or less probable. n21 In ruling the fact that Brown had pleaded guilty and the facts supporting Brown's proffer logically irrelevant, the trial judge relied on Potts, a case where the police raided the defendant's house pursuant to a search warrant. During the course of the search, the police arrested Potts and four of his companions, who were later indicted for the same drug charges: Possession with Intent to Deliver Heroin; Possession with Intent to Deliver Cocaine; Possession with Intent to Deliver Marijuana; Possession of Hypodermic Needles and Syringes; Maintaining a Dwelling for the Keeping of Controlled Substances; and Conspiracy Second Degree. Before Potts's trial, two of his companions pleaded guilty to simple possession of drugs under a plea agreement with the State. Potts sought  [*22]  to admit his codefendants' pleas at his later trial on the basis that their statements were exculpatory because they corroborated his defense that the drugs seized did not belong to him. In other words, Potts sought to admit his codefendants' guilty pleas to possession of drugs to demonstrate that they - and not he - possessed the drugs. We agreed with the trial judge's conclusion that the relevance of the codefendants' pleas was tenuous at best. We stated:

    The two codefendants who entered guilty pleas did so only as to the charges against themselves. Defendant did not establish that his companions' pleas constituted confessions to exclusive possession of the drugs. Hence, their pleas were not shown to exculpate defendant.

 In this case, the trial judge relied on Potts, stating:


The trial judge's reading of Potts and his application of Potts to this case was misplaced. The trial judge concluded that Potts announced a categorical rule that an accomplice plea and proffered statements in support of the plea are always irrelevant because they do not exonerate a codefendant. We did not announce any such rule in Potts. In Potts, the defendant was attempting to establish, through the fact that his codefendants pleaded guilty, that he did not possess drugs. We held that the codefendants' pleas were irrelevant because they did not independently establish that the codefendants' exclusively possessed the drugs. That is, in Potts, the codefendants' guilty pleas did not make it more or less probable that the defendant also possessed drugs.

Potts is distinguishable from this case. Linda did not seek the admission of Brown's plea and proffer to  [*24]  exonerate herself by seeking to establish that Brown, and not she, committed the murders. Rather, Linda sought to introduce Brown's plea and proffer to "test Ms. Rucinski's credibility and to support an argument that the State does not have confidence in the strength of its case." n22

We agree that Brown's plea to two counts of First Degree Murder for a deal on his penalty did not, alone and without more, impeach Mellisa. The mere fact that Brown pleaded guilty did not give Mellisa motive to lie. But, Brown's proffered statements in support of his plea were relevant to test Mellisa's credibility on the degree of Linda's involvement in the crimes with which Mellisa, Brown and Linda were charged.

The trial judge recognized that "the State had good reason to believe Brown put too much of a finger of blame on Linda Charbonneau to get himself out of trouble and was less than truthful." n23 What the trial judge apparently failed to apprehend, however, is that the jury, after hearing the inconsistencies  [*25]  in the State's witnesses' versions of the events, might conclude that Mellisa had a similar motive to implicate Linda falsely. Brown proffered that Mellisa was the primary actor in Sproates's death. Brown's proffered statements gave Mellisa a motive to lie and to implicate falsely her codefendants. Defense counsel wanted to use Brown's proffer to demonstrate to the jury that Mellisa had a motive to lie both in her proffer and in her live testimony.

The trial judge's refusal to credit the relevance of Brown's proffered statements to impeach Mellisa is best understood by focusing on the following question the trial judge posed to the defense:


That question illustrates that the trial judge did not understand how Brown's proffered statements could impeach Mellisa. The reason, in our view, is that the trial judge had unqualifiedly endorsed the State's contention that it was Brown - and not Mellisa -who was lying. We agree that if Brown lied, his statements would provide a motive for Mellisa to falsely implicate him but not necessarily Linda. But, if Brown's proffered statements created a reasonable doubt in the jurors' minds about Mellisa's credibility generally, that doubt might also have affected the jury's analysis of the believability of Mellisa's version of Linda's involvement in the murders.

It was error for the trial judge to accept the State's contention (and essentially find as fact) that Brown (not Mellisa) was lying and then to remove that issue from the jury. The prosecutors argued two reasons why they believed Brown was lying: (1) his statements were substantially inconsistent with Mellisa's; and, (2) no DNA was found on the knife that Brown claimed Mellisa used in killing Sproates. These reasons alone could not justify the trial judge taking from the jury the issue of the credibility of Mellisa's testimony  [*27]  about Linda's involvement in the murder.

First, the fact that Brown's statements were inconsistent with Mellisa's could not, without more, justify concluding that Brown was lying. The State itself acknowledged that "there are always inconsistencies in people's statements." n25 Further, the State and the trial judge both knew Mellisa was an admitted liar. At trial she admitted lying at every opportunity she had to speak about the case. n26 Certainly one could not reasonably conclude that Brown's statements were false solely because they were inconsistent with statements from someone who openly admitted she herself was a liar.

Therefore, the trial judge was left with the argument that there was no DNA on a knife that Brown alleged Mellisa used to stab Sproates. But, to conclude that Brown must have been lying simply because there was no DNA on the knife is misguided. The absence of DNA on the knife can be explained by any number of circumstances - including the possibility that someone cleaned the knife after the murder. n27 Thus, we hold that the trial judge abused his discretion by accepting, as fact, the State's contention that Brown was lying but that Mellisa was truthful and by removing from the jury the issue of who spoke truthfully and whether the inconsistencies resulting raised a reasonable doubt about Linda's guilt.

Mellisa provided all of the testimony necessary to convict Linda. Because Mellisa's testimony was the linchpin of the State's case, we cannot be confident that any evidence that could impeach Mellisa's credibility would not create a reasonable doubt about Linda's guilt.

Because the jury might possibly believe Brown, Brown's proffered statements would be relevant to test Mellisa's credibility. In his proffer, Brown heavily implicated Mellisa. Brown claimed that Mellisa assisted in John's burial, and actively participated in Sproates's murder. The State itself, armed with Brown's statements that were posited as truthful, leveraged a deal with Mellisa. Surely the State suggested to Mellisa that she was facing First Degree Murder charges and a possible death sentence if a jury believed Brown's proffered statements. Motivated to ensure that she received a favorable plea bargain and to avoid a possible death sentence based on Brown's statements, Melissa had an interest in falsely implicating either or both of her codefendants and in exculpating herself. Brown's proffered statements were clearly relevant to challenge Mellisa's truthfulness when she gave testimony implicating Linda. The  [*30]  trial judge erred in holding otherwise.
 
2. Whether the Probative Value of Brown's Proffer was Substantially Outweighed by the Danger of Unfair Prejudice.

The logical relevance of Brown's guilty plea and proffer does not conclude the analysis. A trial judge may exclude otherwise relevant evidence if "its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues or misleading the jury, or by considerations of undue delay, waste of time or needless presentation of cumulative evidence." n28 As discussed above, we review for an abuse of discretion a trial judge's ruling under D.R.E. 403 for an abuse of discretion. n29

Here, the trial judge found that any probative value of Brown's proffer was substantially outweighed by the danger of unfair prejudice, confusion of the issues, misleading the jury, or wasting time. n30 The trial judge stated:


In so holding, the trial judge abused his discretion. Brown's proffered statements were highly probative. Mellisa was the State's primary witness who provided the testimony that enabled the jury to find her mother guilty of two intentional murders. Any evidence that would operate to impeach Mellisa's credibility had probative value. If the jury found Mellisa's testimony generally incredible, then Linda would have had an opportunity to create a reasonable doubt in the jurors' minds about the extent of her own involvement in the crimes charged.

The trial judge held, however, that the jury would have viewed the "case as closed" if Brown's plea and proffered statements were admitted. We interpret that view to mean that if the jury was told that Brown  [*32]  had pleaded guilty to two counts of First Degree Murder and if the jury had accepted the facts underlying Brown's proffer, the jurors would have concluded that Linda was guilty of two intentional murders. We do not agree. The State's theory, which the jury understood, was that Linda, Mellisa, and Brown all committed the murders. Both Brown and Mellisa's statements implicated themselves, each other, and Linda to some degree. Admitting Brown's proffered statements, certainly would not give the jury any indication that the "case (against Linda) was closed." The trial judge should have admitted the proffered statements.

Thus, we hold that the trial judge abused his discretion by holding that the probative value of the Brown statement was outweighed by the prejudice that evidence would occasion to Linda (on the theory it would "close" the case against her) or that it would be a waste of time to explore the inconsistency in the testimony of two eyewitnesses to two intentional murders. There certainly, on the other hand, was no cognizable prejudice to the State in admitting statements that the State itself had procured as part of its plea bargain and that, by not rescinding its plea  [*33]  bargain with Brown, continued impliedly to endorse.

The record and transcripts in this appeal support the habeas court's findings describing the careless and unreasonable manner in which defense preparations were undertaken and juxtapose the evidence presented at trial with the evidence that trial counsel failed to discover. As determined by the habeas court, "this is not a case where, after investigation, counsel for the defendant decided to pursue one strategy rather than another" but a case where counsel's "investigation into their own theory of the case was entirely inadequate." (2) In light of the record evidence, we agree with the habeas court's conclusion that trial counsel rendered deficient performance by failing to investigate the factual defense to the  [*16]  crime and failing to obtain available testimony confirming that defense and their client's own statements to them. Furthermore, counsel's decision to end the investigation into Woods's involvement when they did was neither consistent with professional standards nor reasonable in light of the evidence obtained by habeas counsel, evidence that would have caused reasonably competent counsel to investigate further.

Having determined that trial counsel rendered constitutionally deficient performance, we must consider the prejudice Jenkins suffered as a result of counsel's deficient performance. As the above discussion indicates, and as our review of the remaining evidence presented at trial confirms, there was reason for very strong suspicion to rest on Jenkins in the minds of the jurors. Had trial counsel performed adequately and presented the available evidence that habeas counsel has brought to light, reasonable jurors still may have concluded beyond a reasonable doubt that Jenkins was not only present during the murders but was either the actual perpetrator of the crimes or an accomplice thereto. As recently emphasized by this Court, however, the burden in an ineffective assistance  [*17]  claim "is to show only 'a reasonable probability' of a different outcome, not that a different outcome would have been certain or even 'more likely than not.' [Cit.]" Schofield v. Gulley, 279 Ga. 413, 416 (I) (A) (614 SE2d 740) (2005). Whatever our own opinions may be about the verdict in this case, we conclude, in agreement with the habeas court, that there is a reasonable probability that the jury would have reached a different verdict in the guilt/innocence phase of trial if confronted with the evidence presented by habeas counsel that trial counsel failed to obtain.

    The Penry Court's conclusion that Lockett and Eddings dictated the rule sought by Penry must be understood in terms of the Court's ruling in Jurek, and its application in later cases. We did not view Lockett and Eddings as creating a rule different from that relied upon in Jurek; rather, we indicated that Lockett and Eddings reaffirmed the reasoning in Jurek and confirmed the necessity of its application to Penry's claim.

    For ten years, this court has [asked] . . .: Was the criminal act "due to the uniquely severe permanent handicaps with which the defendant was burdened through no fault of his own"? Graham v. Collins, 950 F.2d 1009, 1029 (5th Cir. 1992) (en banc), aff'd, 506 U.S 461, 113 S. Ct. 892, 122 L. Ed. 2d 260, (1993). This formulation encompasses four principles found in Penry I: voluntariness, permanence, severity, and attribution. Did the defendant acquire his disability voluntarily or involuntarily? Is the disability transient or permanent? Is the disability trivial or severe? Were the criminal acts a consequence of this disability?

    Typically, evidence of good character, or of transitory conditions such as youth or  [*27]  being under some particular emotional burden at the time, will tend to indicate that the crime in question is not truly representative of what the defendant's normal behavior is or may become over time, and that the defendant may be rehabilitable so as not to be a continuing threat to society. The core of Jurek--which we cannot conclude has been abandoned--is that the mitigating force of this kind of evidence is adequately accounted for by the second special issue. But in Penry the Court was faced for the first time with a wholly different type of mitigating evidence. Not evidence of good character, but of bad character; not evidence of potential for rehabilitation, but of its absence; not evidence of a transitory condition, but of a permanent one; but nonetheless evidence which was strongly mitigating because these characteristics were due to the uniquely severe permanent handicaps with which the defendant was burdened through no fault of his own, mental retardation, organic brain damage and an abused childhood. There was no way this type of evidence could be given any mitigating force under the second special issue. To recognize that, as Penry did, is not necessarily  [*28]  to deny the validity of Jurek as it applies to the more typical case.