A jury convicted Appellant-Roger Dale Stafford, Sr., ("Stafford") of three counts of first degree murder and sentenced him to death. Stafford filed a petition for writ of habeas corpus pursuant to 28 U.S.C. Sec. 2254. The district court denied the writ, but issued a certificate of probable cause.

Stafford filed a timely appeal of the court's decision: (1) denying his request to amend his habeas petition to include claims as to the sufficiency of the evidence and actual innocence; (2) holding that he was not denied effective assistance of counsel at the penalty phase; (3) holding that the trial court's denial of his motion for a change of venue did not violate his right to a fair trial at both the guilt and penalty phases; and (4) holding that the Oklahoma Court of Criminal Appeals properly reweighed the aggravating and mitigating factors for sentencing purposes. We affirm.

A jury convicted Stafford of three counts of first degree murder on March 7, 1980, for killing Melvin, Linda, and Richard Lorenz, husband, wife and son, respectively. On March 17, 1980, on the jury's recommendation, the trial court sentenced Stafford to death on all three counts.

Stafford's wife, Verna, provided the prosecution with most of the testimonial evidence used in the trial. She said that Stafford, his brother Harold, and she were traveling along Highway 35, in McClain County, Oklahoma, when the three of them agreed that they would raise the hood of the car so that passersby would think it was disabled. If anyone stopped, they planned to rob them. Verna stood in view of passing cars while Stafford and his brother hid behind the car. The Lorenzes stopped to assist Verna. Melvin got out of his truck to examine Verna's car. At that time, Harold and Stafford, who had been lying in wait, demanded money from Melvin. He told them that he could give them some, but not all, of his money because they were enroute to a funeral. Stafford then shot Melvin twice. When Linda ran from the truck to Verna's side, Verna knocked her to the ground while Stafford shot her twice as she fell.

Stafford then went to the camper on the Lorenz's truck where he had heard a child's voice. Stafford cut a hole in the window screen and shot twice into the camper killing Richard. Melvin and Linda's bodies were found in a field adjacent to the highway. Richard's body was found in a field three-fourths of one mile away.

Stafford denied any involvement in the murders and has stuck to his alibi that he was working approximately 100 miles away in Tulsa when the murders occurred.

Approximately five months prior to his trial, Stafford had been convicted in another trial of killing six people during an armed robbery in what is now known as the "Sirloin Stockade" murders.1 The Sirloin Stockade murders are not before us. However, Stafford argues that they had an impact on his trial in Purcell County for the Lorenz murders because of their extensive publicity and because the Lorenz trial jurors knew he had already been sentenced to death for them.

Stafford's post-conviction relief record for the Lorenz murders is extensive and need not be outlined here. This habeas appeal is the culmination of Stafford's first and only habeas action, although his petition has had a tortuous path as it has travelled between this court and the United States District Court for the Western District of Oklahoma, with a number of stays in the Oklahoma state courts to consider various issues. Stafford began this federal habeas action in August 1985.

Stafford asserts that the district court should have allowed him to amend his habeas complaint to allege that the evidence was insufficient to convict him under Jackson v. Virginia, 443 U.S. 307, 99 S. Ct. 2781, 61 L. Ed. 2d 560 (1979), and to add a claim of actual innocence under Herrera v. Collins, --- U.S. ----, 113 S. Ct. 853, 122 L. Ed. 2d 203 (1993). We review the trial court's decision denying Stafford's motion to amend his petition for abuse of discretion. We have said that "[a] court should freely grant leave to amend when justice so requires. When a court refuses leave to amend, it must state its reasons." Gillette v. Tansy, 17 F.3d 308 (10th Cir.1994) (citing Fed.R.Civ.P. 15(a)) (citations omitted); Foman v. Davis, 371 U.S. 178, 182, 83 S. Ct. 227, 230, 9 L. Ed. 2d 222 (1962).

A district court should apply Fed.R.Civ.P. 15(a) to decide whether to allow an amendment. Rule 15(a) allows the judge to deny a motion to amend because of "undue delay, bad faith or dilatory motive on the part of the movant, repeated failure to cure deficiencies by amendments previously allowed, undue prejudice to the opposing party by virtue of allowance of the amendment, futility of the amendment, etc." Foman, 371 U.S. at 182, 83 S.Ct. at 230.

Insufficiency of the Evidence: The district court denied as untimely Stafford's motion to amend to include an insufficiency of the evidence claim because the issue could have been raised in the initial petition. The court stated that Stafford "conceded [that the claim] could have been raised in the petition as originally filed."

We cannot say that the district court abused its discretion in denying this motion because it was not made until approximately seven years after the federal habeas action was commenced and no credible reasons were advanced as to why this claim was not asserted at that time. Further, it is apparent that such a claim would be futile given the evidence that was adduced against Stafford at trial. Any review of the record on the sufficiency claim would be limited to determining whether the evidence, in the light most favorable to the prosecution, is sufficient so that any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt. Jackson, 443 U.S. at 319, 99 S.Ct. at 2789; Herrera, --- U.S. at ----, 113 S. Ct. at 861.

Here, Verna's eyewitness and participatory testimony, although sometimes inconsistent, was enough to sustain the government's verdict. Verna testified that Stafford, his brother Harold, and she committed the crimes, with Stafford pulling the trigger for all three murders. Additionally, one witness testified that she overheard an argument between Stafford and Verna in which Verna stated that Stafford did the killing. Another witness, Mr. Tackett, provided the police with a description of the man driving the victims' truck that led to a composite drawing that was admitted into evidence. Tackett testified that he observed the person shortly after the murders in Stillwater, Oklahoma with the victims' truck. Tackett ultimately picked Stafford out of a photo line-up.2 Stafford points to alibi evidence, but the proper inquiry would be whether the government's evidence, if believed by a reasonable jury, was sufficient for a guilty verdict. Here, that question would have to be answered affirmatively. Thus, we affirm the district court's decision denying the motion to amend to include an insufficiency of the evidence claim.

Actual Innocence: Stafford's motion to amend his petition to include a claim of actual innocence was also denied. Stafford filed this motion after the Supreme Court granted certiorari on the "actual innocence" issue in Herrera. After Herrera was decided, the district court denied the motion to amend without further comment. Given the Supreme Court's ruling in Herrera, we find this ruling was not an abuse of discretion.

Herrera strongly suggests that "actual innocence" based on newly-discovered evidence is not, by itself, an adequate basis for habeas relief under Sec. 2254. The Court said, "[c]laims of actual innocence based on newly-discovered evidence has never been held to state a ground for federal habeas relief absent an independent constitutional violation occurring in the underlying state criminal proceeding." Herrera, --- U.S. at ----, 113 S. Ct. at 860. Although a strong claim of actual innocence may excuse a habeas petitioner's procedural default, so that a court may address an underlying constitutional claim that would otherwise be barred for procedural reasons, it does not itself provide a basis for habeas relief. Ordinarily, if there is no underlying constitutional error in the trial, an innocent, but convicted person must resort to executive powers of pardon. Herrera, --- U.S. at ----, 113 S. Ct. at 862.

However, the Court left the door open a tiny crack by thereafter assuming "for the sake of argument" that in a capital case a truly persuasive demonstration of "actual innocence" made after trial would render the execution of a defendant unconstitutional, and would warrant federal habeas relief if there were no state avenue open to process such a claim. However, the Court cautioned that "the threshold showing for such an assumed right would necessarily be extraordinarily high." Herrera, --- U.S. at ----, 113 S. Ct. at 869. This high threshold is because of

the very disruptive effect that entertaining claims of actual innocence would have on the need for finality in capital cases, and the enormous burden that having to retry cases based on often stale evidence would place on the States.

Id.

Here, if we similarly assume such a right, Stafford's asserted "newly-discovered evidence" falls considerably short of the "extraordinarily high" showing of actual innocence that would be required. The "evidence" Stafford argues as new includes: (1) two bloody fingerprints on the tailgate of the Lorenz pickup that do not match those of Melvin or Linda Lorenz, or Roger, Verna, or Harold Stafford; (2) the Oklahoma State Bureau of Investigation's unproductive efforts to discredit Stafford's alibi that he was in Tulsa when the crimes were committed; (3) that Verna and the prosecutors had entered into a cooperation agreement despite the state's alleged assertion at trial that no such agreement existed; (4) that a hat found near the pawn shop from which the murder weapons were stolen contained a Negroid hair and that the father of the boy who found the guns is black and has a history of violent criminal behavior; and (5) that the composite allegedly of Stafford did not resemble him closely, whereas the composites of Verna and Harold bore a remarkable likeness of them.

None of this is persuasive evidence of "actual innocence." At most, it is corroborating evidence, impeaching evidence, or evidence merely raising some suspicion or doubt of Stafford's guilt. Furthermore, insofar as we can tell from the record before the district court, some of these allegations are only that--allegations by counsel with no evidentiary support whatsoever in the record. For example, the reference to the unidentified fingerprints appears to have no evidentiary support in the habeas record before the district court. The fingerprints were not introduced, nor was there any effort to introduce testimony identifying to whom they belonged or when they were placed on the truck. We note that after the murders, the testimony at trial was that Stafford and his accomplices took the truck to Stillwater, and then back to Oklahoma City, where it was parked at the airport and thus there were other opportunities for the fingerprints to be placed on the truck. If indeed there were two unidentified bloody fingerprints, they have little evidentiary value absent necessary foundational support.

The Oklahoma State Bureau of Investigation's ("OSBI") unsuccessful effort to discredit Stafford's alibi was only corroborative, at most, of Stafford's alibi. We know no details of the efforts taken by the OSBI to disprove his claims that he was in Tulsa at the time of the murders and there is certainly nothing in the record to suggest that the OSBI efforts proved he was in fact in Tulsa at the time.

None of the other evidence proves actual innocence. The agreement between Verna and the prosecutors is only impeachment evidence, rather than evidence of actual innocence. The import of the hair found in a hat that was discovered near the pawnshop where the murder weapon was stolen is, again, highly speculative and inconclusive. Finally, the composite drawings were used at trial and Stafford had the opportunity to cross-examine the witnesses concerning them. In sum, this evidence, taken singly or together, does not rise to the "extraordinarily high" threshold evidence of innocence postulated by the Supreme Court in Herrera.3

II. Ineffective Assistance of Counsel at Penalty Phase

Stafford asserts that his trial counsel, J. Malone Brewer, ("Brewer") was ineffective at the penalty phase of the trial because he did not adequately prepare for the penalty phase, he did not present any separate mitigating evidence for the jury to consider during that stage of the proceedings, and his closing argument at the penalty stage filled just three pages of the trial transcript. At the penalty stage, both sides merely incorporated the record of the guilt phase into the penalty phase. Thus, no separate mitigating evidence was introduced. Brewer's closing argument at the penalty stage consisted merely of arguing generally against the death penalty and urging a life sentence if the jurors harbored any doubt about Stafford's guilt.

To prove that counsel's performance was deficient, Stafford bears the burden of meeting the two-prong test of Strickland v. Washington, 466 U.S. 668, 104 S. Ct. 2052, 80 L. Ed. 2d 674 (1984). In Strickland, the Court held that a defendant must show: (1) "that counsel's performance was deficient" with reference to prevailing professional norms, and (2) "that the deficient performance prejudiced the defense." Id. at 687, 104 S.Ct. at 2064; United States v. Rivera, 900 F.2d 1462, 1472 (10th Cir.1990). Prejudice is shown by demonstrating that "there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different." Strickland, 466 U.S. at 694, 104 S.Ct. at 2068; Rivera, 900 F.2d at 1472.

We approach this review with considerable restraint. As the Supreme Court in Strickland noted:

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.

Strickland, 466 U.S. at 689, 104 S.Ct. at 2065.

In Oklahoma, the jury is required to weigh the aggravating and mitigating factors to determine whether the death penalty should be imposed.4 In this case, the jury found that the following three aggravating factors applied to all the murders: (1) the defendant knowingly created a great risk of death to more than one person; (2) there exists a probability that the defendant would commit criminal acts of violence that would constitute a continuing threat to society; and (3) the murders were especially heinous, atrocious, or cruel.5 The jury found an additional aggravating factor as to Linda's and Richard's murders because their murders were committed for the purpose of avoiding or preventing a lawful arrest or prosecution.

To determine whether Brewer's performance was below the prevailing standards, we review the evidence presented at the habeas evidentiary hearing.6 Brewer admitted at the hearing that he presented no separate mitigating evidence at the sentencing stage, although some mitigating evidence was available through incorporation of the evidence from the guilt phase into the sentencing phase. Similarly, although Brewer acknowledged that he conducted no specific investigation for mitigation evidence, he and his assistant did explore Stafford's background to some degree. Further, Stafford specifically stated that he did not want an insanity defense or anything concerning his childhood presented and he refused to testify at the penalty phase of the trial because he thought it would be useless. Finally, Brewer maintained that any offer of explicit mitigating evidence would have been inconsistent with the alibi defense of innocence used during the trial's guilt phase.

Mark Donatelli, an attorney who practices in New Mexico and federal courts, testified on behalf of Stafford at the habeas evidentiary hearing as an expert on death penalty trials. Donatelli testified that he had serious concerns with Brewer's handling of the penalty phase. For example, Donatelli said that there was evidence that Stafford had been physically abused as a child, been locked in a closet by his father for days at a time, shot his father in the stomach to stop him from beating his mother, spent time in juvenile facilities, and had a tumor removed from his brain. Donatelli testified that a Bender Gestalt test indicated that something, besides the brain tumor, was missing from Stafford's brain. Additionally, Donatelli testified that records indicated that Stafford had been hospitalized for mental reasons and had been given Lithium and Thorazine. Finally, Stafford told Donatelli that he had served in combat in Vietnam in a capacity that required him to kill.

We conclude that Brewer performed well below the prevailing professional norm in his handling of the penalty phase of this capital case. We do not think that an alibi defense would have been inconsistent at all with the presentation of the mitigating evidence described by Donatelli. The jury obviously had already rejected the alibi defense when they convicted Stafford in the guilt phase in less than two hours, and any good attorney should have pursued alternative mitigating arguments with vigor in the penalty phase of the trial. It was Brewer's responsibility to advise Stafford that the mitigating evidence of his personal background was not inconsistent with his claim that he did not commit the murders. Brewer failed to do so and now attempts to excuse his conduct by telling us that Stafford did not want to take the stand. We realize that an attorney could be hampered when his client refuses to take the stand. However, if Stafford had been properly advised that the two defenses were not inconsistent, he might have made a different decision or changed his mind. Additionally, Brewer could have used other witnesses to ensure that the jury was provided with mitigating evidence and that the penalty phase of the trial had some semblance of an adversary process. At the very least, Brewer should have enumerated for the jury the evidence from the guilt phase that could be considered mitigating and invited a weighing of that evidence in Stafford's favor. He did none of this and we can only conclude that Brewer functioned well below the level of any competent attorney during the penalty phase.

Having concluded that Brewer was ineffective at the penalty phase, we must next determine whether Stafford has met his burden of proving that Brewer's ineffective assistance prejudiced him. We noted in Osborn v. Shillinger that

the Court intended the prejudice standard to be flexible, it emphasized that "a defendant need not show that counsel's deficient conduct more likely than not altered the outcome in the case." Instead, the defendant bears the burden of showing "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 the confidence in the outcome."

861 F.2d 612, 626 (10th Cir.1988) (quoting Strickland, 466 U.S. at 693, 694, 104 S. Ct. at 2067, 2068) (internal citations omitted). When the ineffective assistance claim relates to the sentencing phase of the trial, the standard is whether there is "a reasonable probability that, absent the errors, the sentencer--including an appellate court, to the extent it independently reweighs the evidence--would have concluded that the balance of aggravating and mitigating circumstances did not warrant death." Strickland, 466 U.S. at 695, 104 S.Ct. at 2069. In deciding whether Stafford was prejudiced, we must keep in mind the strength of the government's case and the aggravating factors the jury found as well as the mitigating factors that might have been presented if Stafford had been provided effective assistance of counsel. Strickland, 466 U.S. at 696, 104 S.Ct. at 2069.

The jury deliberated as to guilt for only two hours, indicating that they had little doubt as to Stafford's guilt. Further, the jury found three sustainable aggravating factors for two of the murders and two sustainable aggravating factors for the third. Against this, we must consider the mitigating factors that Stafford now claims should have been introduced and argued at the penalty stage of this trial.

The generalized suggestion that Stafford had a brain tumor removed and that something was missing from his brain is too vague and speculative for us to conclude that it would have had much impact. We are not the first court to reach this conclusion. Indeed, in 1983, Stafford's counsel implied that Stafford's brain tumor may have resulted in psychiatric problems; however, the Oklahoma Court of Criminal Appeals refused to accept counsel's unsubstantiated claim and noted that Stafford's counsel had "likewise failed to offer any authority to show that the appellant had any history of psychiatric disorder." Stafford v. State, 669 P.2d 285, 296, 296 n. 12 (Okl.Cr.1983), vacated and remanded, 467 U.S. 1212, 104 S. Ct. 2652, 81 L. Ed. 2d 359 (1984). Eleven years later Stafford still offers no evidence that the removal of the brain tumor was not entirely successful; there is still no evidence that anything was removed from his brain that related to Stafford's ability to conform his behavior to noncriminal conduct. Stafford offers no medical expert's testimony, nor any medical records describing his brain condition or linking that condition to his behavior. What we have is nothing more than the same unsupported speculation that was offered and rejected eleven years earlier. To the contrary, the state represents that his medical records concluded that his mental capacity and functioning were normal. Similarly, the fact that he has been given Lithium and Thorazine and has been hospitalized for mental reasons is of doubtful mitigation value for several reasons. First, Stafford absolutely refused to allow his attorneys to put on evidence of his mental condition, and, as the client, he has the right to preclude this line of defense. Further, even now Stafford offers no evidence of a mental illness that would have diminished his moral culpability for this crime. On the other hand, the state asserts that his medical records concluded that he had an antisocial, sociopathic personality. Thus this "evidence" is a double-edged sword for Stafford. There is a real risk that the jury would have seen the antisocial, sociopathic personality as corroborative of an aggravating factor7 rather than as a mitigating one. Finally, the allusions to his childhood experiences are similarly conclusory and remote. The only evidence in this regard was from Donatelli, the lawyer who testified as an expert on Stafford's behalf at the habeas hearing. Donatelli said he had information about these facts, but, in the absence of Stafford's willingness to testify, all we have to go on is Donatelli's hearsay conclusions. In any event, Stafford was 27 years old at the time of this crime, and these childhood events were remote in time by then. Although this certainly would be admissible as mitigation evidence, Stafford presented no evidence that these events had any continuing effect on his ability to conform his conduct to noncriminal behavior.

Given the strength of the case against Stafford, the number of aggravating factors found by the jury, and the nature of the crime itself, we have very little doubt that the vague and conclusory mitigating evidence that is now suggested sixteen years after the trial would not have changed the jury's decision to impose the death penalty. Because we conclude that Stafford has failed to satisfy the prejudice prong of Strickland, we affirm the district court's denial of habeas relief on this issue.

Stafford asserts that the district court erroneously found no error in the state trial court's denial of his request for a change of venue. He argues that a change of venue was necessary because the extensive media coverage of the Lorenz and Sirloin Stockade murders and trials warranted a finding of actual prejudice or a presumption of prejudice on the part of the jurors. Stafford argues that the denial of a venue change infected both the guilt and penalty phases of the trial.

In a pretrial motion, Stafford's trial counsel unsuccessfully requested a venue change. When the trial judge denied the venue change, he told counsel that he could raise it again at voir dire. However, counsel did not again raise the issue at that point. Additionally, in post-conviction proceedings, the state court found that, not only did the trial court not abuse its discretion in denying the change of venue motion, but trial counsel failed to present the statutorily required affidavits of three persons from the community to sustain a motion for a change of venue based on jury prejudice.8 Stafford v. State, 669 P.2d 285, 290 (Okl.Cr.1983). We review the trial court's decision denying a transfer of venue for an abuse of discretion. We give great deference to the trial court's exercise of its discretion, United States v. Abello-Silva, 948 F.2d 1168, 1176 (10th Cir.1991), cert. denied, --- U.S. ----, 113 S. Ct. 107, 121 L. Ed. 2d 65 (1992), and its decision is "entitled to a presumption of correctness and will not be overturned unless there is manifest error." United States v. Affleck, 776 F.2d 1451, 1454 (10th Cir.1985).

It is undisputed that there was extensive media coverage of the Sirloin Stockade and Lorenz murders and trials. Stafford asserts that the extensive media coverage resulted in every one of the thirty-six potential jurors in the voir dire pool having had prior knowledge of the defendant, the Lorenz murders, the Sirloin Stockade murders, and his prior death penalty sentence for the Sirloin Stockade murders. Stafford also asserts that the atmosphere in the small community of Purcell was hostile. He argues that we should therefore presume prejudice. Alternatively, he argues, that we should find that there was actual prejudice.

Presumed Prejudice: The defendant bears the burden of establishing that prejudice should be presumed. Abello-Silva, 948 F.2d at 1176. Stafford must establish that "an irrepressibly hostile attitude pervaded the community." Id. This is a difficult standard, even in cases in which there has been extensive media coverage, because

[p]retrial publicity in topical criminal cases is inevitable. The publicity impacts defendant's rights only when it dictates the community's opinion as to guilt or innocence. In rare cases, the community is so predisposed that prejudice can be presumed, and venue must be transferred as a matter of law.

Id. at 1176-77 (internal citations omitted).

We do not have the actual media clippings. However, we do have some testimony from the habeas evidentiary hearing and both parties' statements that there was extensive media coverage. At the habeas hearing and in response to a subpoena for media files, Bob Nelon, on behalf of KWTV, Griffin Television, made the following representations to the court. "[T]he criminal proceedings involving Mr. Stafford and so-called Sirloin Stockade murders and Lorenz Family murders generated a great deal of media coverage.... It extended over a great period of time from the time that the murders were committed through the arrest and trial, and I think there has even been some coverage of some of the appellate proceedings that have gone on in this case." Transcript of Proceedings, Vol. I, at 46. "The best that I can get from our clients are the estimates ... that we are looking at from each station probably anywhere from a hundred to three hundred different stories." Id. at 47. "There is a great deal of video tape, literally tens of hours of tape, ... some of it somewhat duplicative because of the repetitive nature of the news programs." Id.

Mike Arnett, reporter for KWTV, also testified at the evidentiary hearing. The following exchange took place:

Stafford's Attorney: In your experience, have you ever, are you aware of any trial or incidents that occurred, that incurred as much news coverage as the Sirloin Stockade and the Lorenz family killings?

Mike Arnett: No, Sir.

Transcript of Proceedings, Vol. II at 236. Mr. Arnett had nine years of experience in radio and television at the time.

Stafford's trial counsel, Mr. Brewer, is reported to have given more than twenty-five interviews during the two trials. Stafford also points out that the trial was televised and the jury was not sequestered on the eve of trial.9

Although we know there was extensive publicity, we know almost nothing about the nature of the publicity. The law does not require that jurors be ignorant of the controversy; only that they be impartial. Abello-Silva, 948 F.2d at 1178. There is nothing in the record to suggest that this publicity was anything other than factual reporting. There is no suggestion of a circus atmosphere or lynch mob mentality, or that a suppressed confession was broadcasted, or of any other community-wide rush to judgment that infected other trials that have been set aside for lack of an impartial jury. See Sheppard v. Maxwell, 384 U.S. 333, 86 S. Ct. 1507, 16 L. Ed. 2d 600 (1966). There, prejudice was presumed from pretrial publicity, a sample of which included reports that Sheppard refused to take a lie detector test; that he was uncooperative with police; that his family had set up a "protective ring" around him; that there had not yet been an inquest; that the investigation had been inept because of "friendships, relationships, hired lawyers, ..."; and "a front-page charge that somebody is 'getting away with murder.' " Sheppard was also subjected to a televised public inquest at which Sheppard's counsel was forcibly ejected from the room by the Coroner conducting the inquest, causing cheers from the audience. Finally, the press published the names and addresses of the potential jurors, which subjected them to many letters and phone calls regarding the trial. See also, Estes v. Texas, 381 U.S. 532, 550, 85 S. Ct. 1628, 1636-37, 14 L. Ed. 2d 543 (1965) (pretrial and trial media coverage resulted in a disruptive circus atmosphere that deprived the defendant of the solemnity and sobriety to which a defendant is entitled and emphasized the "notorious" character of the defendant); Rideau v. Louisiana, 373 U.S. 723, 726, 83 S. Ct. 1417, 1419, 10 L. Ed. 2d 663 (1963) (prejudice presumed "after the people ... had been exposed repeatedly and in depth to the spectacle of Rideau personally confessing in detail to the crimes ...; [a]ny subsequent court proceedings in a community so pervasively exposed to such a spectacle could be but a hollow formality").

It is Stafford's burden to establish that an irrepressibly hostile attitude pervaded the community. Simply showing that all the potential jurors knew about the case and that there was extensive pretrial publicity will not suffice to demonstrate that an irrepressibly hostile attitude pervaded the community. Abello-Silva, 948 F.2d at 1178; Brinlee v. Crisp, 608 F.2d 839, 847 (10th Cir.1979), cert. denied, 444 U.S. 1047, 100 S. Ct. 737, 62 L. Ed. 2d 733 (1980). We have cautioned that "presumed prejudice is rarely invoked and only in extreme circumstances." Abello-Silva, 948 F.2d at 1177. Here, Stafford has simply not met that burden.

Actual Prejudice: Stafford also asserts that the voir dire showed actual prejudice on the part of a significant number of the potential and actual jurors. We review actual prejudice by examining the totality of the circumstances. Abello-Silva, 948 F.2d at 1177. The trial court "has broad discretion in 'gauging the effects of allegedly prejudicial publicity and in taking measures to insure a fair trial.' " Id. at 1177 (quoting United States v. McDonald, 576 F.2d 1350, 1354 (9th Cir.), cert. denied, 439 U.S. 830, 99 S. Ct. 105, 58 L. Ed. 2d 124 (1978)). The proper time for a trial judge to assess juror partiality is during voir dire. Here, both sides agree that all of the jurors seated said they could be fair. The four potential jurors on the initial panel who expressed doubt about their ability to be fair were dismissed for cause. Stafford does not cite anything in the voir dire that would establish that the jurors who heard his case and sentenced him were not fair and impartial. Instead, Stafford stresses that they all had heard about the case and all knew about the Sirloin Stockade murders. However, jurors need not be ignorant of the case; in fact, jurors who are completely unaware of such a highly publicized trial would probably not be the best jurors. "It is sufficient if the juror can lay aside his impression or opinion and render a verdict based on the evidence presented in court." Abello-Silva, 948 F.2d at 1178 (quoting Irvin v. Dowd, 366 U.S. 717, 723, 81 S. Ct. 1639, 1642-43, 6 L. Ed. 2d 751 (1961)).

Judge West, in denying Stafford's habeas petition on this issue, found

that thirty-six (36) prospective jurors were called. All indicated that they knew of the petitioner and had heard or read reports by the media about this case. They also indicated that they were aware of the prosecution of the petitioner in Oklahoma County and of his convictions for the Sirloin Stockade murders. However, after examination by trial counsel which included inquiry into the effect of the intense media coverage given to the Sirloin Stockade murders and the instant crimes on the jury panel, only four (4) prospective jurors stated that they would be unable to assess the innocence or guilt of the petitioner based upon the evidence presented at trial due to preconceived opinion....

While the petitioner has complained that several jurors stated that they could only "do their best" to forget what they had previously heard or read, these jurors were not identified by the petitioner. The Court has reviewed the voir dire and finds a majority of those who arguably fall within this category were targets of the parties' preemptory challenges.

ROA Vol. II, Doc 75 at 22. The court also noted that Stafford's counsel did not peremptorily challenge Juror Bartley who stated that "[y]eah, I could try. I will do my best on it is all I know" in response to counsel's question of whether he could keep the Sirloin Stockade incidents separate from the Lorenz incidents. Id. at 923. We find that Judge West has fairly characterized the voir dire evidence presented at the habeas evidentiary hearing and that Stafford has failed to prove that his jury was not impartial and fair.

Penalty Phase: Each juror knew of Stafford's prior conviction and death sentence for the Sirloin Stockade murders. Stafford argues that this relieved the jurors of the responsibility for putting him to death and therefore made it easier for them to return the death sentence. This issue was raised in anticipation of the Supreme Court's ruling in Romano v. Oklahoma, cert. granted, --- U.S. ----, 114 S. Ct. 380, 126 L. Ed. 2d 330 (1993). In Romano, the prosecutor admitted into evidence a copy of the judgment and death sentence Romano had received in a prior case. The Oklahoma Court of Criminal Appeals later vacated that judgment and death sentence. Romano, 827 P.2d 1335 (1992).

In Romano, the Supreme Court agreed to hear the following question: "Does admission of evidence that a capital defendant already has been sentenced to death in another case impermissibly undermine the sentencing jury's sense of responsibility for determining the appropriateness of the defendant's death, in violation of the Eighth and Fourteenth Amendments?" Id. On June 13, 1994, the United States Supreme Court answered this question in the negative. Romano v. Oklahoma, --- U.S. ----, 114 S. Ct. 2004, 129 L. Ed. 2d 1 (1994). It concluded that it is impossible to know how knowledge of a defendant's prior death sentence might have affected the jury. The Court thought it "equally plausible that the [knowledge] could have made the jurors more inclined to impose a death sentence, or it could have made them less inclined to do so." Id. --- U.S. at ----, 114 S. Ct. at 2012. The Court refused to engage in "speculation, rather than reasoned judgment," on the actual effect such knowledge had on the jury. Id. --- U.S. at ----, 114 S. Ct. at 2103. We think that the same observation can be made here. In any event, here we do not have to guess whether knowledge of the prior death sentences had an impact on the jury. Unlike in Romano, here these jurors had this knowledge before the voir dire, and during voir dire they gave assurances that they could be fair. Stafford has not presented evidence to overcome our reliance upon those assurances.

There is simply no evidence in this case to suggest that the jurors imposed the death penalty because they felt a lesser sense of responsibility for their decision. Stafford has failed to meet his burden of proof and we affirm the district court on this issue.

IV. Reweighing of Aggravating and Mitigating Factors

The Oklahoma Court of Criminal Appeals, on its second remand as to this issue, 853 P.2d 223, and in an attempt to comply with Clemons v. Mississippi, 494 U.S. 738, 110 S. Ct. 1441, 108 L. Ed. 2d 725 (1990), reweighed the aggravating and mitigating factors after deleting an unconstitutionally vague aggravating factor that had been submitted to the jury. See also, Stringer v. Black, --- U.S. ----, ----, 112 S. Ct. 1130, 1136, 117 L. Ed. 2d 367 (1992) ("We require close appellate scrutiny of the import and effect of invalid aggravating factors to implement the well-established Eighth Amendment requirement of individualized sentencing determinations in death penalty cases."); Richmond v. Lewis, --- U.S. ----, ----, 113 S. Ct. 528, 535, 121 L. Ed. 2d 411 (1992) ("Where the death sentence has been infected by a vague or otherwise constitutionally invalid aggravating factor, the state appellate court or some other state sentencer must actually perform a new sentencing calculus, if the sentence is to stand"). In Clemons, the Court remanded for reweighing of the mitigating and aggravating factors because the reviewing court's opinion was not sufficiently detailed for the Court to conclude that the defendant received individualized treatment. Clemons, 494 U.S. at 752, 110 S.Ct. at 1450. In Stafford's case, this reweighing was necessary because one of the aggravating factors originally found by the jury was subsequently held unconstitutional in Clemons--namely, that the crime was "heinous, atrocious, or cruel."

Stafford argues that the Oklahoma Court of Criminal Appeals did not properly reweigh the relevant factors and that the District Court erred in giving the Oklahoma Court a second chance to conduct the reweighing.10

We review the Oklahoma Court of Criminal Appeals' decision reweighing the aggravating and mitigating factors to determine whether that court's independent reweighing of the aggravating and mitigating factors afforded Stafford "an individualized and reliable sentencing determination based on the defendant's circumstances, his background, and the crime." Clemons, 494 U.S. at 749, 110 S.Ct. at 1449. We review this question under a de novo standard of review. See id. at 750-52, 110 S.Ct. at 1449 (applying a de novo review analysis). However, the state court's factual findings at to the mitigating and aggravating factors should be reviewed under the "rational factfinder" standard. Lewis v. Jeffers, 497 U.S. 764, 783, 110 S. Ct. 3092, 3103-04, 111 L. Ed. 2d 606 (1990).

The Oklahoma Court of Criminal Appeals determined that the jury properly found the following aggravating factors as to Linda and Richard Lorenz' murders: (1) Stafford knowingly created a great risk of death to more than one person; (2) Stafford would probably commit future criminal acts of violence that would constitute a continuing threat to society; and (3) Stafford committed the murders to avoid or prevent lawful arrest or prosecution. As to Melvin Lorenz' murder, two aggravating factors were found: (1) Stafford knowingly created a great risk of death to more than one person; and (2) Stafford would probably commit future criminal acts of violence that would constitute a continuing threat to society.

The court also listed the facts supporting the jury's decision, including that Stafford killed three people after he lured them to their doom and that he killed Linda and Richard after having robbed and killed Melvin. The court also noted that the murders were cold-blooded and committed on innocent people attempting to help a woman with car trouble.

The court also cited the potentially mitigating factors that were presented during the guilt phase of the trial and incorporated into the penalty phase. This evidence included Stafford's alibi and the witnesses who testified to his alibi; that he had financial trouble and lived as a transient; that he had a wife and three young children; that the main witness against him had lied several times about the offenses; that he was a good employee; and that he needed psychiatric help. The court further noted:

the jury was specifically instructed to consider whether the following mitigators applied to the facts of the case: appellant had no significant criminal history; the murder was committed while appellant was under extreme emotional or mental stress; the victim participated in or consented to the homicidal act; appellant believed the circumstances surrounding the murders provided a moral justification for his conduct; appellant was an accomplice to the murder, which was committed by another person, and his participation was relatively minor; appellant acted under duress or under the domination of another person; appellant's capacity at the time of the murders to appreciate the wrongfulness of his conduct, or to conform his conduct to the requirements of the law, was impaired as a result of mental disease or intoxication; and, defendant was only twenty-eight (28) at the time of the murders.

Oklahoma Court of Criminal Appeals, Opinion on Remand, at 225. Brewer did not request this jury instruction, but it was instead given by the judge to inform the jury of some of the mitigating factors that it could consider. The jury was also advised that it could consider any other mitigating circumstances that it found from the evidence presented in the case as well.

In an attempt to ascertain what the jury would have decided absent the invalid aggravating factor, the court then reviewed the aggravating and mitigating factors to determine the role played by the "heinous, atrocious, or cruel" aggravating factor in the jury's decision. The court found that the death sentence should be upheld as factually substantiated and appropriate, and, "[i]ndeed, this case does not present a single close question upon reweighing." Id. at 226. The court found that none of the mitigating factors was compelling, stating:

The mitigating evidence was so insubstantial that, rather than call any second stage witnesses, defense counsel attempted to influence the jury by informing it of his own personal crusade against the death penalty.... Frankly, we are offended by the manner in which defense counsel attempted to mask the lack of mitigating evidence in this case.

Id. at 226. The court also found that the prosecutor did address the "heinous, atrocious, and cruel" factor, but that it was only a small part of her closing argument. More emphasis was placed on the other factors. Id. at 226.

The court said that it engaged in a careful, independent review and consideration of the evidence and that the "jury's improper consideration of the unconstitutional aggravator did not play a significant role in its decision to sentence appellant to death." Id. We conclude that the court thoroughly reviewed the evidence in both the guilt and penalty phases and reweighed the aggravating and mitigating factors consistent with Clemons and its progeny. We also conclude that the court's factual findings as to the mitigating and aggravating factors meet the rational factfinder standard.

We AFFIRM the district court's denial of Stafford's petition for a writ of habeas corpus.