Timothy Wilson Spencer attacks a Virginia state court judgment sentencing him to death for the murder of Dr. Susan Hellams. The district court denied Spencer's petition for a writ of habeas corpus. We affirm.

I

* Dr. Susan Hellams was a resident in neurosurgery at the Medical College of Virginia in Richmond.1 She was murdered in her home on the night of October 2, 1987 or the early morning of October 3, 1987.

The police were notified by her husband after he returned home and discovered her partially-clothed body on the floor of the couple's bedroom closet. Dr. Hellams's attacker apparently gained access to the house by cutting out a large portion of a second-story bedroom window screen.

The medical examiner testified at trial that the cause of Dr. Hellams's death was ligature strangulation, apparently caused by two belts found around her neck. The medical examiner also testified that Dr. Hellams sustained other injuries, including a fractured nose, a blunt force injury to the lower lip, various bruises and scrapes, and an injury consistent with one made by a shoe on the back of her right leg.

In addition, the medical examiner had found fluid consistent with seminal fluid on her back and in the gluteal fold. The medical examiner also observed small mucosal tears of the anal ring, which were "consistent with the anus having been penetrated 'by a hard object, such as a penis.' " 238 Va. 563, 385 S.E.2d at 852.

The presence of spermatozoa was found on swabs taken from the vagina, rectum, and perianal area. Seminal fluid and spermatozoa also were found on Dr. Hellams's skirt and slip. The swab from the perianal area, as well as the stains on Dr. Hellams's skirt and slip, were examined by the Commonwealth's expert serologist and compared to Spencer's blood. Based on her analysis of identifiable secretions, the serologist testified at trial that the source of the secretions was a third party, because neither Dr. Hellams nor her husband could have produced the secretions.

The serologist further stated that the secretions in the seminal fluid found on the skirt and slip were consistent with Spencer's secretion type and inconsistent with Dr. Hellams's husband's type.2 The secretions in the seminal fluid found on the perianal swabs were consistent with a combination of Spencer's and Dr. Hellams's blood types and inconsistent with a combination of the blood types of Dr. Hellams and her husband.

A sample of Spencer's blood and a sample of the seminal fluid found on Dr. Hellams's slip were subjected to DNA analysis. The two samples matched. This evidence was admitted at trial.

The trial commenced in the Circuit Court of the City of Richmond, Manchester Courthouse, on January 17, 1989. The jury convicted Spencer of capital murder, rape, sodomy, and burglary. 238 Va. 563, 385 S.E.2d at 351. At the penalty phase of the trial, the jury fixed Spencer's punishment for the capital murder at death. Spencer appealed his convictions and death sentence to the Supreme Court of Virginia, which affirmed. Spencer v. Commonwealth, 238 Va. 563, 385 S.E.2d 850 (1989). The United States Supreme Court denied Spencer's petition for a writ of certiorari. Spencer v. Virginia, 493 U.S. 1093, 110 S.Ct. 1171, 107 L.Ed.2d 1073 (1990).

Spencer next filed a petition for a writ of habeas corpus with the state trial court on September 10, 1990. The petition was dismissed on November 15, 1990. Spencer v. Murray, No. ML2232 (Cir.Ct. for the City of Richmond, Manchester Courthouse, Nov. 15, 1990). The Supreme Court of Virginia affirmed. Spencer v. Murray, No. 910252 (Va. June 4, 1991). Spencer then turned to the United States District Court for the Eastern District of Virginia. The district court denied his habeas petition. Spencer v. Murray, No. 3:92CV160 (E.D.Va. Jan. 21, 1993). Spencer then asked the district court, on February 11, 1993, for a Certificate of Probable Cause to appeal to this court. That request was denied. Spencer v. Murray, No. 3:92CV160 (E.D.Va. March 30, 1993).

Spencer filed his Notice of Appeal in the district court on April 29, 1993. Spencer then applied to this court for a Certificate of Probable Cause on May 25, 1993. Appellee Murray responded with a motion to dismiss the appeal on May 11, 1993. By order filed June 21, 1993, we denied Murray's motion to dismiss and, as individual judges, granted Spencer's application for a Certificate of Probable Cause. Spencer v. Murray, No. 93-4002 (4th Cir. June 21, 1993).

On the same day that we entered our order, the Commonwealth sought and received from the state trial court an execution date of August 26, 1993, in this case. Commonwealth v. Spencer, Nos. 88-181-F to 88-184-F (Cir.Ct. for the City of Richmond, Manchester Courthouse, June 21, 1993). On July 23, 1993, Spencer applied to this court for a stay of execution, which we granted on July 27, 1993, for the pendency of this appeal or until further order of this court. Spencer v. Murray, No. 93-4002 (4th Cir. July 27, 1993).

On appeal, Spencer raises seven issues: (1) his trial counsel were ineffective because they failed to secure a DNA expert for the defense; (2) he is "actually innocent" of the crime for which he was sentenced to death, and he would not have been convicted if he had been able to challenge the DNA evidence and if the "prejudicial injection of astronomical probability ratios" into the trial had not occurred; (3) his trial counsel were ineffective because they did not conduct voir dire on the issue of racial prejudice; (4) Virginia's proportionality review is unconstitutional, and her application of procedural default rules was unconstitutional because it did not make "rational exceptions"; (5) the jury instructions at trial on mitigating evidence were constitutionally inadequate; (6) his trial counsel were ineffective because they did not explore or present certain mitigating evidence; (7) the DNA analysis used on the evidence in this case was subject to error and produced unreliable results, the results should not have been admitted, and his trial counsel were ineffective in handling this evidence.

We do not consider Spencer's Issues 4 (proportionality review and default rules), 5 (jury instructions on mitigating evidence), and 7 (DNA analysis claims), except to the extent that in Issue 7 he alleges his counsel were ineffective in their handling of the DNA evidence and to the extent he raised a challenge to the admissibility of the DNA evidence on direct review. The Supreme Court of Virginia held that these issues were procedurally defaulted under the rule of Slayton v. Parrigan, 215 Va. 27, 205 S.E.2d 680 (1974), cert. denied, 419 U.S. 1108, 95 S.Ct. 780, 42 L.Ed.2d 804 (1975). Spencer v. Murray, No. 910252 (Va. June 4, 1991).

When a habeas petitioner has defaulted in a federal claim in state court pursuant to an independent and adequate state procedural rule, federal review of the defaulted claim is barred, absent two exceptions not urged upon us by Spencer. Coleman v. Thompson, 501 U.S. 722, 111 S.Ct. 2546, 115 L.Ed.2d 640, (U.S.1991).3 We address each of Spencer's remaining issues, turning first to his allegations of ineffective assistance of counsel, and then to his actual innocence claim.4

Claims of ineffective assistance of counsel are governed by the now-familiar standard of Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). To prevail on an ineffective assistance of counsel claim, the petitioner must show that counsel's performance was deficient and that the petitioner suffered prejudice as a result. 466 U.S. at 687, 104 S.Ct. at 2064. The petitioner must show both deficient performance and prejudice; the two are separate and distinct elements of an ineffective assistance claim. See 466 U.S. at 687, 104 S.Ct. at 2064.

In examining a claim that counsel's performance was deficient, we examine whether counsel's performance was reasonable under prevailing professional norms. 466 U.S. at 688, 104 S.Ct. at 2064. We conduct this review not by asking whether we would have defended the petitioner in the same way, but instead with a deferential eye, and we presume that challenged acts are likely the result of a sound trial strategy. 466 U.S. at 689, 104 S.Ct. at 2065.

Just as the petitioner carries the burden of proving that counsel's performance was deficient, the petitioner also carries the burden of affirmatively proving that prejudice resulted from counsel's deficient performance. 466 U.S. at 693, 104 S.Ct. at 2067. The petitioner must affirmatively prove "that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome." 466 U.S. at 694, 104 S.Ct. at 2068.

Further, when it is a conviction the petitioner is challenging, "the question is whether there is a reasonable probability that, absent the errors, the factfinder would have had a reasonable doubt respecting guilt." 466 U.S. at 695, 104 S.Ct. at 2068-69. When it is a death sentence that the petitioner is challenging, "the question 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." 466 U.S. at 695, 104 S.Ct. at 2069. Keeping the proper standard in mind, we turn to each of Spencer's allegations of ineffective assistance of counsel.

Spencer's Issue 1 on appeal is that his trial counsel were ineffective because they failed to procure a defense DNA expert. At this point, we feel it necessary to point out that Spencer was tried twice for capital murder in the Circuit Court for the City of Richmond by the same trial judge and defended by the same attorneys, Jeffrey L. Everhart and David J. Johnson. The first trial commenced on September 19, 1988, and Spencer was found guilty of the capital murder of Miss Debbie Davis. That trial has been before us for review and is the subject of our recent opinion Spencer v. Murray, 5 F.3d 758 (4th Cir.1993), cert. denied, --- U.S. ----, 114 S.Ct. 1208, 127 L.Ed.2d 555 (1994). We mention the earlier trial because the state trial judge, confronted with two very similar trials just four months apart, allowed the parties to file consolidated motions for both cases.

On June 15, 1988, Spencer's attorneys filed a motion for funds for experts with the trial court to put the court on notice that they intended to seek such funds. The court had an extended discussion with Spencer's counsel about the motion and procuring an expert. Spencer, 5 F.3d at 760 n. 2. Spencer now argues that trial counsel's performance was deficient because

[n]othing in the record reveals that counsel did anything to follow through with this motion. No mention is made anywhere in the record of any additional requests for hearings or experts.

As counsel recognized the need for specific experts, some affirmative steps should have been taken to secure them.... If nothing else, counsel should have read the current literature dealing with forensic DNA.

Spencer's attorneys filed an affidavit with the state trial court concerning the allegations made in the state habeas petition. The affidavit details the research Spencer's attorneys conducted into DNA evidence. Spencer's attorneys also questioned at least four experts and attempted to find one who would be willing to serve as a defense witness, but they "were unable to find an expert who was willing to accept such an appointment." So Spencer's attorneys did take affirmative steps to secure a DNA expert. The fact that they could not find one cannot be charged to them as deficient performance. See also Spencer v. Murray, No. 3:92CV160, slip op. at 4-5, (E.D.Va. Jan. 21, 1993) (supplying additional reasoning for rejecting claim).

Further, we doubt very much from the record we have in front of us that Spencer is correct when he charges in his brief that his attorneys did not read the current literature. Spencer has the burden of proof on the issue of deficiency, and he has not carried it. We note that the attorneys even attended Spencer's Arlington trial in these serial killings and had a blind DNA test run by an independent laboratory, which only corroborated the Commonwealth's evidence.5

Spencer charges that his counsel's performance was deficient because they did not conduct voir dire on the issue of racial bias. The affidavit submitted by Spencer's attorneys shows that they made a strategy decision not to conduct voir dire on this issue:

Prior to trial, because of the publicity from the first Richmond trial, we asked for and obtained a change of venire and the jury selection actually took place in Norfolk. We also asked for and obtained individual voir dire. During jury selection, the questions we asked any given juror were based upon our combined professional judgment as to how best to determine whether the juror was impartial or would be favorable or unfavorable to the defense. If a prospective juror's answers gave us any doubt about his or her impartiality, we either challenged the juror for cause or followed up with additional questions until, in our judgment, the doubt was removed or we believed the juror's answers warranted a challenge for cause.

In our view, particularly because of the change of venire, race was simply not an issue in the case. We had no reason to believe that any prospective juror harbored any racial bias against Spencer, and our decision not to ask any questions on voir dire that might have injected race into the case was a matter of trial tactics.

Under Strickland, we defer to counsel's sound trial strategy decisions. 466 U.S. at 689, 104 S.Ct. at 2065.6

Spencer also argues that his counsel were ineffective because if they had adequately investigated his case, they would have discovered that his presentence report, school history, and Department of Corrections reports show that he was a troubled child, that he was emotionally damaged by being told his father was dead when he in fact was alive, that he used a dangerous drug, PCP, and that he may suffer from organic brain damage. He also argues that they failed to seek the appointment of a psychologist to evaluate his mental state. Once again, we turn to counsel's affidavit.

The record shows that they conducted a thorough investigation of Spencer's background, both personally and through the use of a private investigator. Counsel or their investigator interviewed family members, neighbors, teachers, employers, and halfway house personnel.7

We therefore reject Spencer's claim of deficient investigation. In addition, counsel had observed the mitigation witnesses in the Arlington trial and in the first Richmond trial. To the extent that Spencer argues that there were witnesses other than mental health experts who should have been presented, Spencer's attorneys formulated a trial strategy based on their "combined professional judgment to determine which persons would make good witnesses and which ones would be poor witnesses."

As for Spencer's claims about his mental state, Spencer's attorneys assert that they never had any reason from their own experiences in talking with him to doubt his sanity or his ability to recall his whereabouts at the times of the crimes. They also had never encountered any evidence of drug use, other than two urinalyses in the halfway house that had shown marijuana use.

Spencer's attorneys knew that the attorneys who represented Spencer in the Arlington trial had hired both a psychiatrist and a psychologist who had found a complete lack of any mitigating circumstances. The Arlington attorneys did not want to inquire further because they feared they would find information that would harm the defense. Rather than request a court-appointed expert in the case, Spencer's Richmond counsel first asked members of the Richmond criminal defense bar to recommend a psychiatrist.

Spencer's attorneys then hired a psychiatrist, Dr. Mullaney, who evaluated Spencer before his first Richmond trial. Dr. Mullaney found nothing of any real help to the defense. However, the report did contain an opinion that Spencer's crimes were "victim specific," and his imprisonment would minimize his future dangerousness to society. However, Spencer's attorneys decided against using Dr. Mullaney. Their reasoning is spelled out in their affidavit:

We decided ... not to use Dr. Mullaney as a witness for several reasons. The sole "plus" of his testimony would have been an opinion that Spencer's future dangerousness would be minimized if he were kept in prison. In our judgment, this "plus" was negligible and we were able to argue that same theory to the jury even without Dr. Mullaney's testimony. Moreover, the minimal "plus" was outweighed by the fact that, if Mullaney were to testify at the penalty stage, the jury would have already found that Spencer committed the offense, and Mullaney would have to admit that Spencer continued to deny his guilt and had shown absolutely no remorse.

Further, we knew that if we wanted to use Dr. Mullaney, then pursuant to Virginia Code Sec. 19.2-264.3:1F the prosecution would be entitled to have Spencer evaluated by its own expert. Based upon what we knew about Spencer and his offenses, we had no doubt that the state's experts would render an opinion that Spencer was, in fact, "future dangerous." We affirmatively wanted to avoid an expert opinion to that effect. Based upon all of this information we made a strategic decision not to request a court-appointed expert and not to present any mental health evidence at the penalty stage.

In light of the reasonable strategy decision made by Spencer's counsel, we do not find their performance deficient.

Claim 7--Deficient Handling of DNA Evidence

Spencer's final claim is that his counsel were ineffective because they "were unable to cope with the tremendous volume of DNA evidence presented by the Commonwealth through its witnesses."8 The affidavit submitted by counsel once again belies the claim.

Counsel conducted a thorough investigation of DNA evidence. They contacted several experts, and even had expert help throughout the trial from experts unwilling to testify. A claim of ineffective assistance of counsel requires us to look at counsel's conduct, not at the experts who aided counsel. See Pruett v. Thompson, 996 F.2d 1560, 1573 (4th Cir.), cert. denied, --- U.S. ----, 114 S.Ct. 487, 126 L.Ed.2d 437 (U.S.1993). Spencer would have us find fault with his counsel's conduct because they could not bring to light arguable flaws in DNA testing that the experts in the field did not themselves know about at the time and that are still far from certain today. We cannot fault Spencer's lawyers for an inability to find something that did not then exist. Their conduct was not deficient, and Spencer therefore cannot prevail under Strickland.

Spencer claims that he is "actually innocent" of the crime for which he was sentenced to death, and he would not have been convicted if the "prejudicial injection of astronomical probability ratios" into the trial had not occurred. Spencer's specific fault-finding with the probability statistics is a claim that is procedurally defaulted because not presented to the Virginia Supreme Court on direct appeal. See Part II, supra. If we construe Spencer's claim as an assertion that he is innocent of the crime for which he was convicted, we cannot entertain his claim because "a claim of 'actual innocence' is not itself a constitutional claim, but instead a gateway through which a habeas petitioner must pass to have his otherwise barred constitutional claim heard on the merits." Herrera v. Collins, --- U.S. ----, ----, 113 S.Ct. 853, 862, 122 L.Ed.2d 203 (U.S.1993).

A claim of factual innocence, in the I-didn't-do-it sense, and actual innocence, which excuses procedural default of a constitutional claim, are two different things. See Spencer, 5 F.3d at 765. We have very limited discretion, if any, to consider claims of factual innocence, and Spencer has not produced any evidence that would cause us to undertake such an inquiry. See id. at 765-66. Therefore, we must assume that Spencer is asserting actual innocence in an attempt to have the merits of his defaulted probability claim heard.

At the outset, we note that the district court was concerned that we may hold that the actual innocence test for defaulted claims under Sawyer v. Whitley, --- U.S. ----, 112 S.Ct. 2514, 120 L.Ed.2d 269 (U.S.1992), applies only to penalty-phase errors and not to guilt-phase errors. See Spencer v. Murray, No. 3:92CV160, slip op. at 8 n. 13 (E.D.Va. Jan. 21, 1993).9 We have implicitly held that Sawyer is not so limited by applying it to guilt-phase errors in Pruett v. Thompson, 996 F.2d 1560, 1568 (4th Cir.), cert. denied, --- U.S. ----, 114 S.Ct. 487, 126 L.Ed.2d 437 (U.S.1993), and we now explicitly hold that Sawyer applies to such errors.

Spencer's claim is that he would not have been convicted if the probability statistics had not been admitted. To have a defaulted claim reviewed, a petitioner must first prove, under the actual innocence test of Sawyer, "by clear and convincing evidence that but for a constitutional error, no reasonable juror would have found the petitioner eligible for the death penalty under the applicable state law." Sawyer, --- U.S. ----, ----, 112 S.Ct. 2514, 2515, 120 L.Ed.2d 269 (U.S.1992). Spencer's claim fails at the outset because he cannot show a constitutional error that could have infected the jury's verdict.

The question of whether the probability statistics should have been admitted is a question of state law that does not involve a federal constitutional issue. It is only in extraordinary circumstances that federal review of these questions is warranted. Grundler v. North Carolina, 283 F.2d 798, 802 (4th Cir.1960); see also Spencer, 5 F.3d at 762. Here, we find that the state trial court carefully considered the DNA evidence, including the population statistics. Spencer's counsel cross-examined the Commonwealth's experts about the Hardy-Weinberg equilibrium,10 ] and they had obtained and reviewed the transcripts from cases in Florida and New York where some of the Commonwealth's DNA experts had previously testified.

The state trial court heard the information about the limitations of the statistics along with the rest of the information about DNA evidence and decided to admit the statistics. We simply cannot find any constitutional error in the admission of the statistics, and we therefore must decline Spencer's invitation to reach the underlying merits of the probability statistics either through the Sawyer actual innocence inquiry or otherwise. Even if his claim were not defaulted, the result would not be any different, for under our holding in Grundler we leave these questions to the state courts.

We hold that Spencer's trial counsel were not ineffective within the meaning of Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). We further hold that Spencer's claim of actual innocence does not state a constitutional claim, and to the extent that his argument is an attempt to have the merits of his defaulted probability statistics claim heard, we find that he is not entitled to have that claim reviewed.

The judgment of the district court is accordingly

AFFIRMED.

*****