OPINION
MICHAEL, Circuit Judge:Petitioner Michael Satcher was convicted in Arlington County, Virginia, in 1991 of the robbery, assault and battery, and attempted rape of Deborah Abel and of the robbery, rape, and capital murder of Ann Borghesani. Satcher was sentenced to death. After his convictions and sentence were affirmed on direct and habeas review in state court, Satcher filed this habeas case in the United States District Court for the Eastern District of Virginia in 1995. Satcher challenged various state court rulings and produced new DNA evidence, which he maintains proves his innocence or at least allows him to raise certain claims proeedurally defaulted in state court. The district court granted the writ on the grounds that the in-court identification of Satcher by Abel, the attempted-rape victim, violated his right to due process under the Fourteenth Amendment. The court denied his other claims. Virginia appeals the in-court identification issue, arguing that the identification was admissible because it was reliable under all of the circumstances. In the alternative, Virginia argues that admitting the in-court identification testimony was harmless error. We agree with Virginia that admitting the in-eourt identification testimony was, if erroneous, harmless error. We therefore reverse the district court’s grant of the writ. Satcher cross-appeals several of his other claims, arguing that (1) his defaulted claims may now be pressed because new DNA evidence provides proof of actual innocence, (2) his claim for ineffective assistance of trial counsel is not defaulted because the ineffective assistance of his state habeas counsel excuses the default, (3) his claim that the two offenses should not have been tried together is not proeedurally defaulted, and (4) his due process claim that the trial court failed to excuse a juror for cause does not rely on a new rule barred from application on collateral review by Teague v. Lane, 489 U.S. 288, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989). We agree with the district court’s denial of these claims. Accordingly, we remand the case with instructions to dismiss the petition.
I.
Around 7:00 p.m. on March 31,1990, Deborah Abel was riding her bicycle on a path running parallel to Lee Highway in Arlington County, Virginia. As Abel rode along, she entered a section of the bicycle path that is hidden from the view of passing motorists by a sound barrier wall about fifteen to twenty feet high. She noticed an “unthreatening” man walking toward her on the path; they made eye contact as they passed. Two or three seconds later, the man pulled Abel off her bicycle from behind, knocking her eyeglasses off; he then dragged her into a ditch along the path. The man began to beat her *564in the face and head, using one hand to keep her face turned toward the ground. The man also managed to pull Abel’s pants part way down. Meanwhile, as Mark Polemani was riding his bicycle along the same section of the path, he noticed a man kneeling near a bicycle lying just off the path. Looking back over his shoulder, Polemani saw the man “throw a punch to the ground.” Polemani got off his bicycle and walked toward the man to investigate. As Polemani approached, the man stopped beating Abel, grabbed her purse, and ran. Polemani chased the attacker along the path and up a hill to the street, but the man escaped. Polemani returned to the path and helped Abel to a nearby apartment complex, where they called the police. Police arrived at the scene at about 7:30 p.m. and searched the area for about an hour.
Ann Borghesani was expected to arrive at a party around 8:00 p.m. that same night. Borghesani was ironing some clothes for the party when her roommate left their apartment around 7:10 p.m. To get to the party, Borghesani had about a five-minute walk along the same bicycle path to get to the Metro station. When Borghesani failed to arrive at the party, her friends called the police. Her friends then began looking for her along her usual route to the Metro, searching late into the night and resuming the search early the next morning. Shortly after 8:00 a.m. her body was found at the bottom of a stairwell in an Air Force Association building alongside the bicycle path, about 100 yards from the spot where Abel was attacked.1 Borghesani was found nude from the waist down. She had been stabbed twenty-one times with a sharp-tipped object and had been raped. Her purse and some of her jewelry were missing. One of Borghesani’s shoes was found on the bicycle path next to the building. A few days later both victims’ purses were found together in some bushes in a parking lot about two • blocks away from the Air Force Association building.
On August 18, 1990, five months after the attacks on Abel and Borghesani, police arrested Satcher for trying to attack three different women that morning on a different bicycle path in Arlington County. The police said nothing to Satcher about the Borghesani murder before or during the trip to the station. When they arrived, however, an officer asked Satcher, “What’s up?,” and he replied that the police were “trying to frame [him] for a murder or something or rape or something.” Police later found an awl (a pointed metal tool used for boring holes) in the glove compartment of Satcher’s car. Borghesani’s wounds could have been made with the awl, although it was never conclusively identified as the murder weapon.
Satcher voluntarily gave blood, saliva, and hair samples to the police. Tests showed that Satcher’s blood type, carried by seven percent of the population, matched the semen sample taken from Borghesani’s body. Tests performed on pubic hairs found on Borghesani’s clothing were inconclusive: two hairs did not match Satcher’s, and the rest could be neither confirmed nor eliminated as coming from Satcher. Virginia’s DNA tests, however, showed that Satcher’s DNA matched the DNA from swabs taken from Borghesani’s body and clothing.
Abel gave the police a description of her attacker just after the attack in March 1990. She described the attacker as a stocky black male between twenty-five and thirty years old, about 5'9" or 5'10" and 190 to 200 pounds. She did not remember any facial scars and said the attacker had a short “Afro” haircut. A police artist made a sketch from Abel’s description. Polemani reported similar physical characteristics, and the police artist also drew a sketch from Polemani’s description. Polemani was startled to see how similar the sketches were, noting that they were “almost identical,” though he thought the sketch based on Abel’s description was better. When Satcher was arrested in August, he was twenty-one years old, 5'6", 152 pounds, with short hair and a facial scar. These details differed *565somewhat from those reported by Abel and Polemani, and nothing in the record suggests that Satcher’s appearance had changed significantly between the attack in March and his arrest in August.
A grand jury indicted Satcher on November 19, 1990, for the murder, rape, and robbery of Borghesani. On April 15,1991, three months before trial, Satcher was also indicted for the attack and robbery of Abel. Police brought in Abel and Polemani to view a lineup fifteen days before trial. Before looking at the lineup, they reviewed the sketch the police artist had made from Abel’s description of the attacker the year before. At the lineup Abel narrowed the choices down to number two and number four (Satcher). She decided that number four looked “almost identical” to the sketch. She picked out number two, however, because he looked “unthreatening,” remembering that the man who attacked her had looked unthreatening as she passed him on the bicycle path. Polemani was unable to positively identify anyone at the lineup, although he testified at trial that he was “pretty sure” that the attacker was number four (Satcher).
Abel was in the courtroom during the two days of jury selection. She observed Satcher as he sat at the defense table and was led in and out of the courtroom. After the first few hours Abel approached the prosecutor and told her, “that’s the guy.” At trial, over Satcher’s objection, the judge allowed Abel to identify Satcher for the jury as her attacker. Abel explained (on redirect examination) that after watching Satcher in the courtroom, she believed he was the attacker because of “the way he walked, the way he shrugged his shoulders when he would come back from the bench, it would put the picture in my mind or remind me of exactly that night.” The sketch based on Abel’s description, the sketch based on Polemani’s description, both witnesses’ testimony about the lineup, and the lineup picture itself were also introduced as identification evidence.
Virginia sought the death penalty. In the first phase of the bifurcated trial, the jury found Satcher guilty of the robbery, assault and battery, and attempted rape of Abel and of the robbery, rape, and capital murder of Borghesani. -In the sentencing phase the jury recommended the death penalty for the killing of Borghesani based on the statutory predicates of “future dangerousness” and “vileness.” See Va.Code Ann. § 19.2-264.2. The Circuit Court of Arlington County sentenced Satcher to death. The Virginia Supreme Court affirmed Satcher’s conviction and sentence on direct appeal. See Satcher v. Virginia, 244 Va. 220, 421 S.E.2d 821 (1992). On February 22, 1993, the United States Supreme Court denied Satcher’s petition for writ of certiorari on direct review. See Satcher v. Virginia, 507 U.S. 933, 113 S.Ct. 1319, 122 L.Ed.2d 705 (1993). Satcher’s motion for rehearing on the petition for certiorari was denied on April 19, 1993. See Satcher v. Virginia, 507 U.S. 1046, 113 S.Ct. 1888, 123 L.Ed.2d 504 (1993).
In November 1993 Satcher filed a habeas petition for collateral review in the Circuit Court of Arlington County, asserting the same arguments he had made on direct appeal. The Circuit Court dismissed the petition. Because of an error in the Circuit Court Clerk’s office, the Clerk failed to notify Satcher’s state habeas counsel that the petition was dismissed until after the deadline for filing a notice of appeal. The Circuit Court entered a new order dismissing the petition to preserve Satcher’s right to appeal. The Supreme Court of Virginia dismissed the appeal anyway, holding that the Circuit Court did not have jurisdiction to enter the new order and that Sateher’s appeal was untimely. The United States Supreme Court denied Satcher’s petition for a writ of certiorari (from state habeas review) on February 27, 1995. See Satcher v. Netherland, 513 U.S. 1193, 115 S.Ct. 1259, 131 L.Ed.2d 139 (1995).
On July 21,1995, Satcher filed a petition in federal court for a writ of habeas corpus pursuant to 28 U.S.C. § 2254. As noted in the introduction, the district court granted the petition on the grounds that Abel’s in-court identification of Satcher violated due process and that the admission of her identification testimony was not harmless error. The district court denied Satcher’s other claims. Virginia appeals the grant of the *566writ, and Satcher cross-appeals the denial of the writ on several of his other claims.
II.
This case requires us to compare on harmless error review the relative weight of one piece of identification testimony (Abel’s in-court identification of Satcher as the man who attacked her) against the weight of the other identification evidence. The Supreme Court has recognized “that the confrontation compelled by the State between the accused and the victim or witnesses to a crime to elicit identification evidence is particularly riddled with innumerable dangers and variable factors which might seriously, even crucially, derogate from a fair trial.” United States v. Wade, 388 U.S. 218, 228, 87 S.Ct. 1926, 1933, 18 L.Ed.2d 1149 (1967). The district court in this case concluded that the in-court confrontation was suggestive, concluded that Abel’s in-court identification testimony was unreliable, and held that the admission of the testimony violated due process. The district court then focused on the“profound impact” that such testimony usually has and found that its admission was not harmless error.
We disagree with the district court’s harmless error analysis. This case did not turn on the high drama of an in court identification. Regardless of whether it was reliable, Abel’s in-court identification was just one part of the relevant identification evidence. We believe the in-court identification did not have a “substantial and injurious effect or influence” on the verdict, and therefore the error, if any, must be considered harmless on collateral review. Brecht v. Abrahamson, 507 U.S. 619, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993). We therefore reverse the district court on this issue.
A.
To determine whether identification testimony is admissible, a court must engage in a two-step process. First, the court must consider whether the identification procedure is unnecessarily suggestive. See Manson v. Brathwaite, 432 U.S. 98, 110, 97 S.Ct. 2243, 2251, 53 L.Ed.2d 140 (1977); Neil v. Biggers, 409 U.S. 188, 198-99, 93 S.Ct. 375, 381-82, 34 L.Ed.2d 401 (1972). A procedure is unnecessarily suggestive if a positive identification is likely to result from factors other than the witness’s own recollection of the crime. See, e.g., United States v. Peoples, 748 F.2d 934, 935-36 (4th Cir.1984) (noting that in-court identification of “the only young black male in the courtroom” might have been unnecessarily suggestive). Second, if the procedure was unnecessarily suggestive, a court must look at several factors to determine if the identification testimony is nevertheless reliable under the totality of the circumstances. See Manson, 432 U.S. at 100, 97 S.Ct. at 2245-46; Biggers, 409 U.S. at 198, 93 S.Ct. at 381-82. These factors include the opportunity of the witness to view the accused at the scene of the crime, the witness’s degree of attention, the accuracy of the witness’s prior description of the accused, the level of certainty demonstrated by the witness at the confrontation, and the length of time between the crime and the confrontation. Biggers, 409 U.S. at 199-200, 93 S.Ct. at 382-83.
Satcher argues" that Abel’s identification testimony was unreliable and rendered his trial unfair. When Abel tried to identify the attacker in a fair pretrial procedure, she had difficulty. Her indecision and ultimate selection of someone else at the lineup, Satcher argues, made it unfair to allow her to identify Satcher in what amounts to an unnecessarily suggestive in-court procedure. What happened, Satcher says, was so suggestive that it throws the reliability of the in-court identification into doubt. The argument merits serious consideration. Abel positively identified Satcher only after she knew that Satcher was the man the Commonwealth was putting on trial and only after she watched deputies lead him in and out of the courtroom during jury selection. After Abel identified Satcher in her trial testimony, Satcher’s counsel objected, calling the identifieation“basically an in-court show-up of one person” that was “highly suggestive, overly suggestive,” and likely to “lead to a misidentification.”
The trial court allowed Abel’s in-court identification of Satcher to stand. The Virginia Supreme Court, declining to comment on whether the in-court identification was unnecessarily suggestive, concluded that Abel’s *567testimony was reliable under the totality of the circumstances. Satcher v. Virginia, 244 Va. 220, 421 S.E.2d 821, 838-39 (1992). But see United States v. Emanuele, 51 F.3d 1123, 1131 (3d Cir.1995) (finding identification impermissibly suggestive and unreliable because government inadvertently walked defendant in front of witness “in shackles with a U.S. Marshal at each side” prior to in-court identification). The district court disagreed with the Virginia Supreme Court, holding that Abel’s in-court identification was unreliable and could not be excused as harmless error.
We find it unnecessary to resolve the question of whether Abel’s in-court identification was unreliable under all of the circumstances and thus inadmissible under the due process clause. Unlike the district court, we believe the admission of Abel’s in-court identification testimony, if erroneous, was harmless error.
B.
The harmless error standard on federal habeas review differs from the standard used on direct review, and this difference reflects “the state’s interest in the finality of convictions that have survived direct review within the state court system.” Brecht v. Ahrahamson, 507 U.S. 619, 635, 113 S.Ct. 1710, 1720, 123 L.Ed.2d 353 (1993). On habeas review a trial error can be reversed only if the error “had substantial and injurious effect or influence in determining the jury’s verdict.” Id. at 637, 113 S.Ct. at 1722 (quoting Kotteakos v. United States, 328 U.S. 750, 776, 66 S.Ct. 1239, 1253, 90 L.Ed. 1557 (1946)). The standard for harmless error on direct review, on the other hand, requires that constitutional error be “harmless beyond a reasonable doubt.” Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 828, 17 L.Ed.2d 705 (1967).2 In addition, on habeas review, “if the federal court is ‘in grave doubt’ about whether the trial error had a ‘substantial and injurious effect or influence’ on the verdict and therefore finds itself ‘in virtual equipoise’ about the issue, the error is not harmless.” Cooper v. Taylor, 103 F.3d 366, 370 (4th Cir.1996) (en banc) (quoting O’Neal v. McAninch, 513 U.S. 432, 435, 115 S.Ct. 992, 994, 130 L.Ed.2d 947 (1995)).
In this case the in-court identification that was objected to (Abel’s pointing at Satcher) was only one part of the evidence identifying Satcher as Abel’s attacker. Therefore, assuming it was error, we believe it was harmless error under Brecht.
In making the harmless error inquiry under Brecht, we must conduct a de novo examination of the trial record as a whole. See Correll v. Thompson, 63 F.3d 1279, 1291 (4th Cir.1995). As will quickly become clear, however, our primary focus must be on how Abel’s in-court identification testimony relates to other identification evidence. This approach places the error in proper context and focuses the inquiry on the actual effect the error may have had on the jury. In Kotteakos v. United States, 328 U.S. 750, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946), the case that provided the “substantial and injurious effect” language adopted in Brecht, the Court provided a good explanation of how the question of harmless error is best approached. The Court explained that the inquiry turns on the effect the error may have had on the jurors’ minds. It is more than a simple balancing test.
[T]he question is, not were [the jurors] right in their judgment, regardless of the error or its effect upon the verdict. It is rather what effect the error had or reasonably may be taken to have had upon the jury’s decision. The crucial thing is the impact of the thing done wrong on the minds of other men, not one’s own, in the total setting. This must take account of what the error meant to them, not singled out and standing alone, but in relation to all else that happened. And one must judge others’ reactions not by his own, but with allowance for how others might react *568and not be regarded generally as acting without reason. This is the important difference, but one easy to ignore when the sense of guilt comes strongly from the record.
Kotteakos, 328 U.S. at 764, 66 S.Ct. at 1247-48 (citations omitted). Here, we must recognize that the Commonwealth’s case depended on solid and persuasive identification evidence. Therefore, the best way to understand the impact the alleged error may have had on the jury is to focus on the context of the identification evidence, rather than to make a laundry list of all the evidence that supports a guilty verdict and then weigh it against a list of exculpatory evidence.
We now turn to the trial record in order to put the in-court identification in context. A police artist made a sketch of the attacker based on Abel’s description the day after the attack while she still “had a pretty clear picture in [her] mind.” At trial this sketch was admitted into evidence and shown to the jury. Abel testified that she “felt that it was an extremely accurate picture” and that she “would rate it an eight or nine on a scale of ten.” (Polemani, the next witness, confirmed that the sketch was “almost identical” to a sketch based on his recollection.) Abel was then asked to explain what happened at the lineup in 1991. She said that the police “told me to choose one [of the lineup participants] only if I felt completely, absolutely sure that I could match that person with who attacked me.” Abel picked number two and number four (Satcher) out of the lineup and had the detective ask them to come forward and turn around. She then picked out number two. She explained her thinking to the jury as follows,
Well, number four, I had decided in my mind, looked extremely close and accurate to the picture that I had seen since fifteen months previous to that, since the attack and since I had helped the artist draw it. And I had decided that four looked very, very — almost identical to the picture.
And I also had a weird feeling about number two. To me, he looked very unthreatening. And that night that I had ridden my bike, I felt that the man that passed me looked unthreatening.
A photograph of the lineup was admitted into evidence.
At this point in the trial the jury had a fairly complete picture of the identification evidence. The jury had in hand a sketch of the attacker made the day after the attack. The jury knew the extent to which Abel had observed her attacker before she described his features to the sketch artist: she had looked at her attacker a few times before they passed face-to-face on the path; she had made eye contact with him as they passed; and she had seen “him run towards [her] bicycle” when he abruptly abandoned the assault. The accuracy of the sketch, which Abel rated as an eight or nine on a scale of ten, would later be confirmed by Polemani’s testimony and by the separate sketch from his description. The jury had a photograph of the lineup and heard Abel’s testimony about the lineup, so the jurors could weigh for themselves whether she was credible in saying that number four looked “almost identical” to the sketch and whether number two looked “unthreatening” enough to confuse her. Of course, the jury could also compare the person depicted in the sketch with Satcher, who was sitting at the defense table.
Abel was not asked whether she could point to her attacker until after the jury had all of the following evidence: (1) Abel’s testimony about her encounter with the attacker and her participation in the making of the sketch, (2) her account of what happened at the lineup, (3) the sketch, and (4) the lineup photograph. When she was asked if her attacker was in the courtroom, she pointed to Satcher. Satcher’s counsel objected, the objection was overruled, and cross-examination began.
On cross-examination Satcher’s counsel chose not to focus on the circumstances and reliability of the in-court identification. Instead, he first challenged the accuracy of Abel’s memory of the attack. Counsel questioned Abel about the amount of light in the area of the attack and about the length of the time she saw the attacker before her glasses were knocked off. Abel explained that she was forced to lie face down while the man beat her in the head, that she did not try to look behind her at the attacker, and that she *569saw only his profile as he ran away. Next, Satcher’s counsel questioned Abel about the disparities between Satcher’s height and weight and the description she gave to the police. Turning to the lineup, Satcher’s counsel had Abel again explain that she picked out number two, not Satcher. Counsel finally turned to the in-court identification, briefly challenging its reliability by having Abel confirm that she knew Satcher was on trial, that she had watched him sitting up front at the defense table, and that he was “the only black man in this area [of the courtroom].”
The prosecutor asked Abel to elaborate on the circumstances of her in-court identification on redirect examination. Abel explained that during jury selection,
I just would observe him and just look at him a lot ... to kind of bring myself back to that night.
And the way he walked, the way he shrugged his shoulders when he would come back from the bench, it would put the picture in my mind or remind me of exactly that night.
Satcher, of course, questions the truthfulness of this explanation, since the positive identification could have come from a subconscious change in Abel’s recollection of her attacker’s appearance to fit that of the person accused. See United States v. Emanuele, 51 F.3d 1123, 1131 (3d Cir.1995) (discussing how victim’s reaction of “it has to be him” following suggestive confrontation greatly diminishes the reliability of subsequent in-court identification).
After reviewing the trial record of this sequence of events in context, we believe that the in-court identification evidence objected to (Abel’s pointing at Satcher) was not significant enough to the case for it to have a substantial and injurious effect on the verdict. The in-court identification neatly tied together the more important identification evidence of the Abel sketch, the Polemani sketch, the testimony of both eyewitnesses about the lineup, and the lineup picture itself, but it was not the cornerstone of the identification testimony. Both the prosecutor and Satcher’s counsel treated the in-court identification for what it was — a final brush stroke rather than the essential outline of the picture identifying the attacker. The fact that the jury was able to compare the person depicted in the sketches with Satcher is quite important. The sketches came from the recollection of the two eyewitnesses (Abel and Polemani) and, unlike the in-eourt identification, could not have been tainted in any way by a suggestive confrontation. We believe, therefore, that Abel’s in-court identification testimony did not have the “substantial and injurious” effect required to reverse a trial error on federal habeas review.3
III.
Satcher cross-appeals several issues: he argues that (1) new DNA evidence provides proof of actual innocence, which should allow him to argue his defaulted claims, (2) his *570claim for ineffective assistance of trial counsel is not defaulted because the ineffective assistance of his state habeas counsel excuses the default, (3) his claim that the two offenses should not have been tried together is not defaulted, and (4) his claim that the trial court’s failure to excuse a juror for cause violates his right to due process is not a new rule barred from application on collateral review by Teague v. Lane, 489 U.S. 288, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989). We will address each claim in turn.
A.
The district court held that many of Satcher’s claims were procedurally defaulted because he did not raise them on appeal in state court. Relying on Schlup v. Delo, 513 U.S. 298, 115 S.Ct. 851, 130 L.Ed.2d 808 (1995), Sateher contends that he is actually innocent of the crimes and that his innocence provides a “gateway” through which he may pass to argue the merits of his defaulted claims. To pass through this gateway, Sateher must show “that it is more likely than not that no reasonable juror would have convicted him in light of the new evidence.” Schlup, 513 U.S. at 327, 115 S.Ct. at 867. Satcher’s new evidence is based on DNA testing performed in 1995, which he claims proves his innocence. The district court held that Satcher’s new evidence did not meet the Schlup standard and dismissed the “gateway” claims. We affirm.
In 1995 a sample of Satcher’s blood was sent to Lifeeodes Corporation, a DNA testing laboratory, to compare with the results of the Commonwealth’s lab (the Tidewater lab) test of swabs taken from Borghesani’s body and clothing after the murder. DNA test results are compared by looking at a sheet of photographic film called an autoradiograph. An autoradiograph has bands of differing widths on it, each band representing a certain identifiable characteristic of the DNA sample. Satcher’s expert, Dr. Aimee Bakken, concluded that on one of four “probes” comparing Satcher’s DNA to the DNA taken from the crime scene, the width of the relevant bands on the Lifeeodes autoradiograph (from Satcher’s blood) differed by 3.66% and 3.06% from the width of the bands on the Tidewater autoradiograph (from the sample taken from Borghesani’s clothing). Sateher therefore argues that the results fell outside the Tidewater lab’s 2.5% match criterion. Virginia argues, however, that the greater difference is attributable to the fact that the procedures used to create the Lifeeodes autoradiograph differ from the procedures used to create the Tidewater autoradiograph against which it was compared. Virginia also argues that even if the difference in procedures did not affect the results, the new evidence does not show that Sateher is innocent.
Under the Schlup standard Sateher may argue his defaulted claims only if he can show that “it is more likely than not that no reasonable juror would have convicted him in light of the new evidence.” Schlup, 513 U.S. at 327, 115 S.Ct. at 867. Sateher argues that his new DNA evidence shows that the Tidewater test was “wrong because [the new evidence] shows that Mr. Satcher’s DNA does not match the DNA in the semen on the vaginal swabs.” Brief for Petitioner at 46. We agree with Virginia that Sateher has overstated the significance of the Lifeeodes test. Satcher’s DNA evidence does not show that the Commonwealth’s test was wrong. Specifically, it does not show that someone else was the source of the DNA sample taken from the crime scene. Sateher does not contest the fact that three out of the four probes still match.4 The slight difference on the fourth probe only shows, at best, that Sateher is not conclusively a four-probe “match” within the scientific meaning of that term as defined by the Tidewater lab’s procedural criteria. None of the experts conclude that Sateher did not contribute the DNA from the sample taken from Borghesani In light of the Commonwealth’s affidavits explaining the likely cause of the small discrepancy between the Lifeeodes Test and the Tidewater test, a reasonable juror would still be likely to give weight to the DNA results showing that Sateher was the contributor of the DNA at the crime scene.
*571Moreover, even discounting Virginia’s DNA test altogether, we cannot say that “no reasonable juror would have found petitioner guilty beyond a reasonable doubt.” Schlup, 513 U.S. at 327, 115 S.Ct. at 867. It would be a different matter if, as Sateher argues, his new DNA evidence showed that he was definitely not the contributor of the DNA on the swab taken from the crime scene. But his new evidence only suggests, at best, that Virginia’s test was inconclusive. See O’Dell v. Netherland, 95 F.3d 1214, 1246-1253 (4th Cir.1996) (rejecting defendant’s actual innocence claim based on new DNA evidence arguably showing state’s test to be inconclusive), cert. granted and aff'd. on other grounds, — U.S.-, 117 S.Ct. 1969, 138 L.Ed.2d 351, (1997).5 We therefore affirm the district court on the denial of Satcher’s actual innocence claims.
B.
The district court held that Satcher’s ineffective assistance of counsel claim was defaulted because state habeas counsel did not raise it. Sateher argues, however, that default is excused because counsel’s failure to raise the ineffective assistance of counsel claim in the state habeas proceeding was itself ineffective assistance of counsel. In other words, Sateher argues that the ineffectiveness of state habeas counsel excuses the procedural default on the ineffective trial counsel claim.
The Supreme Court has not specifically addressed an ineffective assistance claim like Satcher’s. As a general rule, there is no constitutional right to effective assistance of counsel when collaterally attacking a conviction. Coleman v. Thompson, 501 U.S. 722, 750, 111 S.Ct. 2546, 2565, 115 L.Ed.2d 640 (1991). An indigent defendant does, however, have a right to effective assistance of counsel on the first appeal as of right. Douglas v. California, 372 U.S. 353, 83 S.Ct. 814, 9 L.Ed.2d 811 (1963). Sateher reasons that because ineffective assistance of counsel cannot be raised on direct appeal in Virginia, state habeas is the first opportunity in which he could raise the issue. Therefore, he argues that this case falls within the holding of Douglas. Specifically, he says that an indigent defendant must have the effective assistance of counsel when raising an ineffective assistance claim in state habeas because it is his first appeal as of right on that claim. Virginia argues, however, that allowing Sateher to claim ineffective assistance at the federal level despite having failed to raise it on state collateral review “would condemn the state and federal judicial systems to endless rounds .of habeas litigation, each challenging the ‘effectiveness’ of the prior counsel’s performance.” Answering Brief for Respondent at 20. Compare Smith v. Angelone, 111 F.3d 1126, 1133 (4th Cir.1997) (holding that counsel’s errors on state habeas cannot be “cause” to excuse procedural default because there is no right to effective assistance of counsel on state habeas) with Mackall v. Murray, 109 F.3d 957, 962-63 (4th Cir.) (holding that right to effective assistance of counsel extends to representation in the state habeas trial court with respect to any issues not directly appealable), vacated and reh’g en banc granted (May 21, 1997). We need not resolve this issue to review Satcher’s claim, however.
To excuse a procedural default in state court, a federal habeas petitioner must show both cause for the default and actual prejudice.
In all eases in which a state prisoner has defaulted his federal claims in state court pursuant to an independent and adequate state procedural rule, federal habeas re*572view of the claims is barred unless the prisoner can demonstrate cause for the default and actual prejudice as a result of the alleged violation of federal law, or demonstrate that failure to consider the claims will result in a fundamental miscarriage of justice.
Coleman, 501 U.S. at 750, 111 S.Ct. at 2565. The thrust of Satcher’s argument on the Coleman/Douglas issue discussed above is that the ineffectiveness of state habeas counsel was cause for the procedural default. Virginia argues that even if Satcher is right about the import of Douglas (that ineffective assistance on state habeas is cause), Satcher cannot show prejudice. Brief for Respondent at 22 n. 14. We agree that Satcher was not actually prejudiced by state habeas counsel’s failure to raise an ineffective assistance claim, and affirm the district court on that ground.
To show actual prejudice, Satcher must demonstrate that the error worked to his “actual and substantial disadvantage,” not merely that the error created a “possibility of prejudice.” Murray v. Carrier, 477 U.S. 478, 494, 106 S.Ct. 2639, 2648 (1986); Poyner v. Murray, 964 F.2d 1404, 1425 (4th Cir.1992). In this case, actual prejudice depends on whether state habeas counsel would have succeeded had he raised the ineffective assistance of trial counsel claim, which in turn would have led to a new trial for Satcher.
To succeed on the ineffective assistance of trial counsel claim, Satcher would have to meet the two-prong test under Strickland, establishing both that his trial attorney’s performance was deficient and that it prejudiced the defense. Strickland v. Washington, 466 U.S. 668, 687, 104 S.Ct. 2052, 2064, 80 L.Ed.2d 674 (1984); Barnes v. Thompson, 58 F.3d 971, 978 (4th Cir.1995). The first prong is met only if trial counsel’s performance “fell below an objective standard of reasonableness.” Strickland, 466 U.S. at 688, 104 S.Ct. at 2064. The second prong is met if there is a “reasonable probability” that, absent counsel’s errors, the sentencer would have concluded that the balance of aggravating and mitigating circumstances did not warrant imposition of the death penalty. Strickland, 466 U.S. at 695, 104 S.Ct. at 2068-69; Barnes, 58 F.3d at 979 n. 11.
Assuming only for the sake of argument that Satcher is correct about the right to effective assistance on state habeas and that we must therefore reach this question, we believe that Satcher cannot meet the Strickland test. Sateher’s two primary claims of ineffective assistance of trial counsel raised in district court were (1) that trial counsel was ineffective “with respect to the investigation and development of mitigation evidence based on [Satcher’s] mental status” and (2) that trial counsel was ineffective for failing to have an independent DNA test of Satcher’s blood. See Petitioner’s Memorandum in Opposition of Summary Judgment 52-71.
The claim that counsel was ineffective with respect to the development of mitigating evidence is not supported by the record. Trial counsel retained a psychologist and a psychiatrist to examine Satcher for the purpose of developing mitigating evidence. The record indicates that these experts found no psychiatric or neurological disorders but found that Satcher had an antisocial personality disorder that might make him a “future danger.” See Respondent’s Motion to Dismiss 53-54. In the face of this potentially damaging psychiatric evidence, counsel decided not to investigate further and instead presented a case for mitigation by having family members and friends testify on Satcher’s behalf. Satcher v. Virginia, 244 Va. 220, 421 S.E.2d 821, 844 (1992). This testimony about the “history and background of the defendant” was properly admissible as evidence in mitigation, see Va. Code Ann. § 19.2-264.4, and could have been relied on by the jury to decline to recommend the death penalty. Trial counsel’s decision not to continue the investigation into psychiatric evidence thus did not fall below an objective standard of reasonableness.'Tn any ineffectiveness case, a particular decision not to investigate must be directly assessed for reasonableness in all the circumstances, applying a heavy measure of deference to counsel’s judgments.” Strickland, 466 U.S. at 690-91, 104 S.Ct. at 2066. Compare Clabourne v. Lewis, 64 F.3d 1373, 1384-87 (9th Cir.1995) (holding that trial *573counsel “failed to adequately prepare and present a case for mitigation” by calling only one “wholly unprepared” expert witness to testify as to defendant’s mental state and failing to discover available evidence that could have proven a psychological defect). We therefore conclude that Satcher did not have an ineffectiveness claim based on any failure by counsel to investigate and develop mitigating evidence about his mental state.
Satcher’s second claim of ineffective assistance is based on trial counsel’s failure to obtain independent DNA testing. This claim also fails because trial counsel’s performance did not fall below an objective standard of reasonableness. Counsel did not allow the testimony of Virginia’s DNA experts to go unchallenged, conducting what Satcher himself characterizes as “extensive and effective cross-examination.” Brief for Petitioner at 33. Satcher later called two DNA experts to contradict the state’s experts and to criticize the reliability of DNA testing. These experts also called into question the procedures of Virginia’s DNA lab and challenged the reasonableness of the conclusion of Virginia’s expert that Satcher was the source of the DNA. Id. at 33-34. Although independent DNA testing might have helped Satcher, it also could have hurt him. Introducing evidence of an independent DNA test might have signalled implicit approval of the usefulness of DNA testing, thereby undermining the argument made at trial that DNA testing is inherently unreliable. Applying a “heavy measure of deference to counsel’s judgments,” Strickland, 466 U.S. at 691, 104 S.Ct. at 2066, this claim also fails to meet the Strickland standard. Therefore, without reaching the question of whether there exists a right to effective assistance of counsel to collaterally attack the effectiveness of counsel at the trial level, we affirm the district court because Satcher cannot meet the Strickland standard for showing ineffective assistance at the trial level.
C.
Next, Satcher argues that trying him for both the Abel and Borghesani offenses together in the same proceeding violated his right to due process under the Fourteenth Amendment. The district court held that this claim was defaulted because it was not raised in state court. On direct appeal Satcher argued that the joinder of the two offenses violated state law because they were not “parts of a common scheme or plan.” See Satcher v. Virginia, 244 Va. 220, 421 S.E.2d 821, 829 (1992).
In order to preserve the right to collateral review in federal court, Satcher must have fairly presented the claim to the state court. Duncan v. Henry, 513 U.S. 364, 365, 115 S.Ct. 887, 888, 130 L.Ed.2d 865 (1995) (per curiam). Satcher failed to mention the federal constitution or cite any cases examining the right to be tried separately under the due process clause; instead, he mentioned only the right to be tried separately under Virginia law.
We agree with the district court that this fails to meet the requirement that a claim be fairly presented to the state court.6
D.
Lastly, Satcher claims that his right to due process under the Fourteenth Amendment was violated when the trial court failed to remove a prospective juror for cause. One of the prospective jurors, Mr. Middle, admitted that he had a very close relationship with some law enforcement officers in Arlington County and that the relationship might affect his ability to weigh the facts. The trial judge refused to excuse the juror for cause, and Satcher used a peremptory *574challenge to strike him. The district court noted that the refusal to excuse the juror was “manifest error” and therefore should have been reversed under Virginia law. See Calhoun v. Commonwealth, 226 Va. 256, 307 S.E.2d 896, 898 (1983). The court held, however, that granting relief under the due process clause would require the use of a new rule barred from application on collateral review by Teague v. Lane.7 As we will explain below, under the Teague framework the issue is whether the right (under the Due Process Clause) to exercise peremptory challenges on a panel free from jurors who should have been excused for cause was dictated by existing precedent when his conviction became final in state court in February of 1993. See O’Dell v. Netherland, — U.S. -, 117 S.Ct. 1969, 1973, 138 L.Ed.2d 351 (1997). We conclude that the rule Satcher seeks was not dictated by existing precedent at that time, and we affirm the district court.
In Ross v. Oklahoma, 487 U.S. 81, 108 S.Ct. 2273, 101 L.Ed.2d 80 (1988), the Supreme Court held that when determining whether a defendant was deprived of the Sixth Amendment right to an impartial jury, a reviewing court may look only at the jury that actually sat, not the bias of the jurors excluded for cause or by peremptory challenges. Thus, the trial court’s refusal to strike a juror for cause does not affect the right to an impartial jury if the defense in fact strikes the juror with a peremptory challenge. Id. at 88, 108 S.Ct. at 2278. There is no Sixth Amendment right to peremptory challenges, so losing one peremptory is not a constitutional violation in and of itself. Satcher, like the defendant in Ross, struck from the panel the juror that he argues should have been excused for cause. Because the prospective juror did not hear the case, his right to a fair trial and an unbiased jury under the Sixth Amendment was not violated. See id. at 88, 108 S.Ct. at 2278.
In Ross the defendant also argued that the error was a due process violation. Under Oklahoma law, however, the defense was obligated to use peremptories to challenge jurors who should have been excused for cause in order to preserve appeal on the impartial jury issue. The Ross Court therefore held that there was no due process violation because the defendant “received all that Oklahoma law allowed him, and therefore his due process challenge fails.” Id. at 91, 108 S.Ct. at 2279-80. In footnote four of the opinion, the Court explicitly noted that it was not deciding whether losing a peremptory would be a due process violation in the absence of a restriction like Oklahoma’s. “We need not decide the broader question whether, in, the absence of Oklahoma’s limitation on the ‘right’ to exercise peremptory challenges,’a denial or impairment’ of the exercise of peremptory challenges occurs if the defendant uses one or more challenges to remove jurors who should have been excused for cause.” Id. at 91 n. 4, 108 S.Ct. at 2280 n. 4.
Satcher raises the issue that was left open in footnote four of the opinion in Ross. Thus, Satcher argues that (1) the juror should have been excused for cause, (2) a peremptory challenge was used to strike that juror, and (3) Virginia law gives defendants the right to exercise peremptories on a panel free from jurors who should have been excused for cause. See Va.Code Ann. § 8.01-357; Breeden v. Virginia, 217 Va. 297, 227 S.E.2d 734, 737 (1976).
Determining whether Teague bars the application of a rule on collateral review is a three-step process:
First,' the court must ascertain the date on which the defendant’s conviction became final for Teague purposes. Second the court must survey the legal landscape as it then existed, and determine whether a state court considering the defendant’s claim at the time his conviction became final would have felt compelled by existing precedent to conclude that the rale he seeks was required by the Constitution. Finally, even if the court determines that the defendant seeks the benefit of a new rule, the court must decide whether that *575rule falls within one of the two narrow exceptions to the nonretroactivity principle.
Caspari v. Bohlen, 510 U.S. 383, 389, 114 S.Ct. 948, 953, 127 L.Ed.2d 236 (1994).
Satcher’s conviction became final on February 19, 1993, when the United States Supreme Court denied his petition for certiorari on direct appeal. See Satcher v. Virginia, 507 U.S. 933, 113 S.Ct. 1319, 122 L.Ed.2d 705 (1993); Teague v. Lane, 489 U.S. 288, 301, 109 S.Ct. 1060, 1070 (1989). Therefore, we must examine the legal landscape in February 1993 and determine whether the rule Satcher seeks “was dictated by precedent existing” at that time. O’Dell v. Netherland, 95 F.3d 1214, 1220 (4th Cir.1996), aff'd, — U.S.-, 117 S.Ct. 1969, 138 L.Ed.2d 351, (1997).
The rule Satcher seeks to invoke in this case is that a court’s failure to remove a juror for cause violates due process if (1) the juror should have been excused for cause and (2) state law gives the defense the right to exercise peremptory challenges on a panel free from jurors who should have been excused for cause. This question was explicitly left undecided by the Supreme Court in Ross, which was decided in 1988. The Court has not revisited the issue since then. This suggests that the rule was not dictated by existing precedent in 1993. See Gray v. Thompson, 58 F.3d 59, 66 (4th Cir.1995).
Satcher tries to overcome this common sense conclusion by identifying several post-Ross, pre-1993 state court decisions and one pre-1993 federal court decision suggesting that Ross implicitly decided in Satcher’s favor the very question it explicitly left open. See Bonin v. Vasquez, 794 F.Supp. 957, 975 (C.D.Cal.1992); Trotter v. Florida, 576 So.2d 691, 692-93 & n. 4 (Fla.1990); People v. Webster, 54 Cal.3d 411, 285 Cal.Rptr. 31, 814 P.2d 1273, 1287-88 (1991); People v. Gordon, 50 Cal.3d 1223, 270 Cal.Rptr. 451, 792 P.2d 251, 265 n. 4 (1990); People v. Coleman, 46 Cal.3d 749, 251 Cal.Rptr. 83, 759 P.2d 1260, 1273 (1988); Commonwealth v. Freiberg, 405 Mass. 282, 540 N.E.2d 1289, 1297 (1989).8 Continuing confusion about the meaning of footnote four in Ross, however, indicates to us that the rule Satcher seeks was not dictated by existing precedent in 1993. None of the cases cited by Satcher holds that such a due process right exists; the cases merely suggest in dicta that Ross supports the existence of such a right. Accordingly, we affirm the district court on this issue.9
IV.
The judgment of the district court is reversed on the in-court identification issue and affirmed on all other claims. Accordingly, we reverse in part, affirm in part, and remand with instructions to dismiss the petition.
REVERSED IN PART, AFFIRMED IN PART, AND REMANDED WITH INSTRUCTIONS
. It is not clear from the record whether the attack on Borghesani happened before or after the police responded to the attack on Abel.
. Virginia argues that our review of the state court decision is even more limited under the Antiterrorism and Effective Death Penalty Act (AEDPA). See Pub.L. No. 104-132, 110 Stat. 1214 (April 24, 1996), codified in 28 U.S.C. §§ 2244 et seq. Because we conclude that Satcher must be denied relief under the old standards of review, we need not consider the potential applicability of the AEDPA in this case. See Arnold v. Evatt, 113 F.3d 1352, 1362 (4th Cir.1997); Matthews v. Evatt, 105 F.3d 907, 922 n. 12 (4th Cir.1997).
. Satcher does not seem to contest the admissibility of Abel’s other identification testimony, such as her testimony about the sketch and the lineup. Even if we were to discount all of Abel’s identification testimony as improperly admitted, it is not clear that her testimony had a substantial and injurious effect on the verdict. We say this because on balance the trial record supports our conclusion that the error in this case (assuming it was error) was harmless. Leaving Abel’s identification testimony aside, Polemani also confirmed the accuracy of the sketch as "almost identical” to the attacker. The physical descriptions given to the police made Satcher out to be bigger than he was, but the circumstances make an overestimate understandable.
The physical evidence linking Satcher to the Borghesani murder, which occurred nearby at about the same time, was very strong. The DNA evidence matched, and Virginia’s experts testified that there was a 99.999998% chance that they had the right man. The awl found in Satcher’s car was consistent with the murder weapon. The evidence from the hairs found was largely inconclusive. The semen sample was also a match; Satcher belonged to the seven percent of the population that could have contributed the semen.
Other circumstantial evidence linked the two crimes: they occurred in the same area at about the same time; both victims were beaten repeatedly in the face; Borghesani was raped, and before Abel’s attacker was chased off, he was pulling down her pants in an apparent rape attempt; one of Borghesani's shoes was found along the bike path, which indicates that she, too, may have been abducted from the path; and the purses of both victims were found together in a nearby parking lot.
. One of Satcher's experts cited a study estimating the chance of a false three-probe match among African-Americans, using a 2.5% match criterion, at 1 in 119,000.
. Because Sateher cannot satisfy the Schlup standard, we are also convinced that he cannot prove actual innocence, and we need not consider whether he would be able to make a freestanding actual innocence claim. See O'Dell, 95 F.3d at 1246 n. 25.
Sateher also argues that the district court erred by not granting an evidentiary hearing on the actual innocence claims. We review this determination for abuse of discretion. Pruett v. Thompson, 996 F.2d 1560, 1577 (4th Cir.1993). The district court refused to have an evidentiary hearing, but it did so only after reviewing the affidavits of Satcher’s experts and concluding that his evidence was not sufficient to show actual innocence. It was within the court’s discretion to conclude that the new evidence created only a "disagreement between experts” that would not satisfy the Schlup standard.
. Satcher also argues that because he mentioned the federal constitution in his state habeas petition, he preserved the claim for appeal. We agree with the district court, however, that “it is not enough to make a general appeal to a constitutional guarantee as broad as due process to present the 'substance' of such a claim to the state court.” Gray v. Netherland, -U.S.-, -, 116 S.Ct. 2074, 2081, 135 L.Ed.2d 457 (1996). In his state habeas petition Satcher made the reference to due process in a conclusory statement after reiterating the same Virginia law misjoinder claims presented on direct appeal. We agree with the district court that it is "[n]ol surprising” that the state habeas court interpreted the claim as being based on Virginia law alone and analyzed it accordingly.
. The Virginia Supreme Court had ruled that the juror was not biased. The district court decided that Teague barred collateral review, but it noted that if its Teague analysis was wrong and it had to reach the question, it would find that the failure to excuse the juror was clear error. Because we affirm the district court’s decision that Teague applies, we need not review its decision that failing to excuse the juror was "manifest error.”
. The Supreme Court has, of course, noted that "[c]onstitutional law is not the exclusive province of the federal courts, and in the Teague analysis the reasonable views of state courts are entitled to consideration along with those of federal courts.” Caspari v. Bohlen, 510 U.S. at 395, 114 S.Ct. at 956.
. The district court found that the rule Satcher seeks to invoke does not fit within either of the two exceptions to Teague. Satcher does not challenge this ruling on appeal.