dissenting:

Although I agree with the majority’s conclusion that counsel’s limited investigation of Landrigan’s background fell below the standards of professional representation prevailing in 1990,1 would not remand this case for an evidentiary hearing because Landrigan has failed to allege facts that, if proven, would demonstrate Strickland prejudice.

We “review de novo [a] district court’s interpretation of AEDPA standards governing the grant or denial of an evidentia-ry hearing.” Earp, 431 F.3d at 1166. We may reverse the denial of an evidentiary hearing only if the district court abused its discretion. Id.

To establish entitlement to an evidentia-ry hearing, Landrigan must demonstrate an objectively reasonable probability that, had counsel proved the mitigating facts now alleged, the sentencing judge would not have imposed the death penalty. See Summerlin, 427 F.3d at 643. Landrigan alleges that his birth mother exposed him to alcohol and drugs in útero and later abandoned him, that he has a genetic predisposition for violence, that his adoptive mother was an alcoholic, and that, as a result of all these factors, he has antisocial personality disorder. 1 Whether these al *651 leged facts create an objectively reasonable probability that Landrigan would have received a different sentence must be considered in light of the particular circumstances of his case. See Smith v. Stewart, 140 F.3d 1263, 1270 (9th Cir.1998) (explaining that the mitigating effect of antisocial personality disorder is context-dependent and that, in Arizona, the disorder “has often been treated on appeal as insufficient to justify mitigation”). Under these standards, three separate factors convince me there is no reasonable probability that Landrigan would have received a different sentence had his counsel proved the mitigating facts now alleged.

First, the mitigating effect of antisocial personality disorder is greater when there is evidence that the disorder “controlled [the] defendant’s conduct or impaired his mental capacity” than when evidence suggests that, at the time of the crime, the defendant “possessed the ability to restrain himself.” State v. Brewer, 170 Ariz. 486 , 826 P.2d 783, 802-03 (1992). For example, in Brewer , the defendant’s antisocial personality disorder was insufficient to tilt the balance of mitigating and aggravating factors in favor of leniency because, notwithstanding his disorder, he had “resisted whatever impulses he may have experienced long enough to insure his own safety by locking the [victim’s] roommate’s dog in another room before commencing his attack on the victim.” Id. at 803.

Here, Landrigan has submitted an expert declaration suggesting in passing that because of his antisocial personality disorder he has “very limited impulse control.” The evidence, however, is that on the occasion of the murder at issue Landrigan was able to control his impulses long enough to cultivate the victim’s trust and attempt to profit from their encounter. The trial testimony of one of the victim’s friends, “Michael,” was that prior to the murder the victim called him three times. During the first phone call, the victim reported that he and Landrigan were together drinking beer. Approximately fifteen minutes later, the victim called a second time to say he was in the middle of sexual intercourse with Landrigan. Shortly thereafter, the victim called to ask whether Michael could get Landrigan a job. Thus, just as the defendant in Brewer was able to control his impulses long enough to take steps to protect his physical safety, Michael’s testimony suggests Landrigan was able to control his murderous impulses long enough to earn his victim’s trust, engage him in sexual relations, and allow the victim to try to get him a job. This evidence diminishes the mitigating effect of his alleged antisocial personality disorder.

Second, as the three-judge panel that initially heard Landrigan’s appeal correctly recognized, in Landrigan’s case the mitigating value of any proven genetic predisposition for violence would not have outweighed its aggravating tendency to suggest Landrigan was undeterrable and, even from prison, would present a future danger to society. Landrigan v. Stewart, 272 F.3d 1221, 1228-29 (9th Cir.2001). One of the two statutory aggravating factors the sentencing judge found was that Landrigan had prior felony convictions involving the use or threatened use of violence on another person. See Ariz.Rev. Stat. § 13-703(F)(2). Landrigan had murdered his best friend and then, while incarcerated for that crime, had brutally stabbed a fellow inmate. In light of this history, and the fact that Landrigan committed the murder for which he was con *652 victed in the present case after escaping from prison, the potential for future dangerousness — including peril to future fellow inmates — inherent in Landrigan’s alleged genetic pre-disposition for violence would have negated its mitigating capacity for evoking compassion in the sentencing judge. See People v. Franklin, 167 Ill.2d 1 , 212 Ill.Dec. 153 , 656 N.E.2d 750, 761 (1995) (“Although this evidence [regarding defendant’s psychological problems and violent family history] could have evoked compassion in the jurors, it could have also demonstrated defendant’s potential for future dangerousness and the basis for defendant’s past criminal acts. The evidence of defendant’s mental illness may also have shown that defendant was less deterrable or that society needed to be protected from him.”).

Third, but by no means least, Landri-gan’s behavior at trial would have negated any effort by counsel to elicit sympathy based on a genetic or biological predisposition toward violence. The three-judge panel that originally heard this appeal aptly described Landrigan’s behavior as follows:

Landrigan, 272 F.3d at 1226-27 . 2

In light of (1) the evidence that Landri-gan could and did control his murderous impulses, (2) the probability that any mitigating value of evidence of genetic or biological predisposition for violence would have been overwhelmed by the evidence’s aggravating tendency to show that Landri-gan was undeterrable, and (3) Landrigan’s menacing behavior at trial, Landrigan has not alleged facts that if proven would create an objectively reasonable probability that he would have received a different sentence. Accordingly, it was not an abuse of discretion for the district court to deny him an evidentiary hearing. I would affirm the district court’s decision in its entirety.

. Contrary to the majority’s contention, to focus on the potential mitigating effect of Landrigan's purported antisocial personality disorder is not to "seize” on a "small piece” of some larger "puzzle” of mitigating circumstances that counsel failed to investigate or present. Maj. Op. at 649 n. 5. As the majority opinion itself acknowledges, all the mitigating circumstances Landrigan faults counsel for not raising "converge” to support the suggestion that he suffers from antisocial personality disorder and cannot control his actions. See *651 id. at 645. Antisocial personality disorder is the crux of Landrigan's ineffective assistance of counsel claim.

. It appears that Landrigan continues to question his counsel’s approaches to this case, as indicated by his hand-written letter to this court of August 16, 2005, stating:

As of today, I want to withdraw my petition and drop all appeals. I no longer wish to pursue any action from your court or any others, and want my execution date to be set as soon as possible. Thank you.