Stephen Anthony Mobley was convicted of malice murder, armed robbery and other crimes in 1994 and sentenced to death for the murder. This Court affirmed Mobley’s convictions and sentence in 1995,1 and the United States Supreme Court subsequently denied Mobley’s petition for certiorari.2 In 1996, Mobley filed this habeas action, raising several claims, including ineffective assistance of counsel. The habeas court ruled that all of Mobley’s claims other than ineffectiveness of trial counsel were either barred or defaulted. The habeas court then vacated Mobley’s death sentence because Mobley’s trial counsel had been ineffective in the preparation for and conduct of the sentencing phase. The state appeals, S98A0410, and Mobley cross-appeals, S98X0411. We reverse and reinstate Mobley’s death sentence.

The evidence adduced at trial showed that Mobley stole a .380 *636pistol from an acquaintance. On February 17, 1991, Mobley robbed a Domino’s Pizza restaurant just after midnight. The only person in the restaurant was the manager, John Collins. After taking all the money in the cash register and back office, Mobley shot Collins in the back of the head as he was kneeling and facing away from Mobley. Over the next three weeks, Mobley committed six additional armed robberies of restaurants and dry-cleaning shops. The police arrested him after a high-speed chase as he was fleeing the last armed robbery. Mobley confessed to the armed robberies and the murder of John Collins, and a gun he tossed from his car during the chase, the .380 pistol he had previously stolen, matched the murder weapon.3

CLAIMS THAT ARE BARRED

1. Claims that were previously litigated and decided on direct appeal are barred because it is well-settled that “[a]fter an appellate review the same issue[] will not be reviewed on habeas corpus.”4 Mobley has asserted on habeas corpus that the trial court’s discovery order violated Rower v. State.5 Rower overruled Sabel v. State,6 which had held that any report by a defense expert had to be reduced to writing and turned over to the state, even if the report was unfavorable and the defense elected not to call the expert as a witness. Before trial, the trial court had issued a discovery order pursuant to Sabel. This Court issued Rower shortly after Mobley’s trial and Mobley raised this issue on his direct appeal. This Court determined that the Sabel discovery order was harmless error in Mobley’s case.7

In later cases, this Court has found that a Sabel discovery order can have a “chilling effect” when it forces the defense to choose against consulting an expert for fear that unfavorable reports would have to be given to the state.8 Although the habeas court acknowledged in its final order that Mobley did not raise the chilling effect argument on direct appeal, the habeas court ruled that the chilling effect argument found in later cases changed the test for harmless error with regard to Sabel discovery orders, and that this intervening change in the law overcomes the bar to reconsidering issues previously litigated.9 Later cases did not change the harmless error test *637for Sabel discovery orders. The error is harmless if we can conclude beyond a reasonable doubt that the result would have been the same absent the trial court’s error. This Court applied this test with regard to the trial court’s discovery order and concluded on direct appeal that any error was harmless.10 Therefore, this issue could not be revisited by the habeas court.11

CLAIMS THAT ARE DEFAULTED

2. A habeas petitioner who fails to raise an issue that he could have raised on direct appeal defaults the issue on habeas corpus, unless he can meet the cause and prejudice test.

[A] failure to make timely objection to any alleged error or deficiency or to pursue the same on appeal ordinarily will preclude review by writ of habeas corpus. However, an otherwise valid procedural bar will not preclude a habeas corpus court from considering alleged constitutional errors or deficiencies if there shall be a showing of adequate cause for failure to object or to pursue on appeal and a showing of actual prejudice to the accused.12

To show cause, Mobley must demonstrate that “ ‘some objective factor external to the defense impeded counsel’s efforts’ to raise the claim that has been procedurally defaulted.”13 To show prejudice, he must demonstrate actual prejudice that “ ‘worked to his actual and substantial disadvantage, infecting his entire trial with error of constitutional dimensions.’”14 The only exception to the cause and prejudice test is the granting of habeas corpus relief to avoid a “miscarriage of justice,” which is an extremely high standard.15

Mobley has raised the following claims for the first time on habeas corpus that he could have raised on direct appeal: pros-ecutorial misconduct for allowing the former district attorney to allegedly testify falsely as to his reasons for rejecting Mobley’s plea offers, prosecutorial misconduct for failing to correct the false testi*638mony of a jailhouse informant, and the unconstitutionality of the Unified Appeal Procedure. Mobley has not shown sufficient cause to overcome his procedural default with regard to these issues. The habeas court thus correctly ruled that these three claims are procedurally defaulted.

INEFFECTIVE ASSISTANCE OF COUNSEL

3. Mobley’s claim of ineffective assistance of counsel is neither barred nor defaulted. An ineffective assistance of counsel claim does not need to be raised until trial counsel no longer represents the defendant.16 The record shows that Mobley’s trial counsel represented him through his direct appeals, and his separate habeas counsel raised this claim at the first available post-conviction opportunity.17 Therefore, ineffective assistance of counsel remains a viable claim.

In order to prevail on a claim of ineffective assistance of counsel, Mobley must show both deficient performance and actual prejudice.18 To show deficient performance, he must demonstrate that his counsel’s performance was not reasonable under the circumstances confronting his counsel at the time, without resorting to hindsight.19 Mobley’s burden is high because his counsel is “strongly presumed to have rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment.”20 To show actual prejudice, he must demonstrate that “there is a reasonable probability (i.e., a probability sufficient to undermine confidence in the outcome) that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.”21

In its final order, the habeas court ruled that trial counsel had been ineffective in investigating and presenting mitigating evidence in the sentencing phase. Specifically, the habeas court found that trial counsel failed to uncover significant mitigating evidence regarding Mobley’s alleged traumatic upbringing, failed to seek funds for an independent psychologist, failed to accept Mobley’s father’s funds to conduct a genetic test, and inadequately presented an unorthodox mitigating defense that attempted to show a possible genetic basis for Mobley’s conduct. A claim of ineffective assistance of counsel is a *639mixed question of law and fact.22 The proper standard of review requires that we accept the habeas court’s factual findings and credibility determinations unless clearly erroneous, but we independently apply the legal principles to the facts.23 We conclude that Mobley’s counsel was not ineffective in the sentencing phase because trial counsel’s mitigation strategy was reasonable and supported by adequate investigation and because Mobley cannot show actual prejudice.

A. The investigation.

Attorneys Dan Summer and Chuck Taylor were appointed to represent Mobley at his trial.24 Summer assumed the role of lead counsel. Summer is a former assistant district attorney and an experienced criminal defense attorney. At the time, he had served as co-counsel for the state on two capital trials. Summer made repeated attempts to persuade the district attorney to allow a guilty plea in exchange for multiple life sentences, but to no avail.

Having concluded that Mobley would be convicted due to the overwhelming evidence of his guilt, trial counsel focused on the sentencing phase and the development of mitigating evidence. Trial counsel interviewed Mobley’s relatives, who told them that Mobley had been a difficult child who had frequent problems with lying, cheating, stealing, vandalizing, setting fires, and being cruel to animals.25 Trial counsel also subpoenaed all the records from the numerous schools and institutions that Mobley had attended and thoroughly reviewed all the records. The school and treatment records reveal numerous instances of Mobley’s misbehavior, such as assaulting other students, using profanity to teachers, lying, stealing, and destroying property, and his resulting expulsions.

During his childhood, Mobley had been treated and evaluated by a number of psychologists and social workers. Trial counsel read all of these psychological records and consulted some of the mental health professionals who had treated Mobley. The psychological reports showed that Mobley had an average IQ and a conduct disorder manifested by a lack of a true values system. He was reported to be manipulative, self-centered and impulsive. Although some psychological reports early in Mobley’s childhood suggested that he *640might have a learning disability or organic brain disorder, later reports found no evidence of either.

Mobley’s counsel sought to find what they called “traditional mitigation evidence.” They interviewed Mobley’s mother, father, and sister, all of whom reported that there was no physical or sexual abuse in Mobley’s childhood. Mobley’s parents also denied any birth trauma or alcohol abuse. Mobley himself denied that he had ever been physically abused, sexually abused or neglected. Additionally, poverty was not a mitigating issue because Mobley’s childhood standard of living had ranged from middle class to affluent. Trial counsel also found that Mobley had only one real friend, who was the owner of the car that he had used to commit the armed robberies (and who had told him that he could use the car to visit his family and look for a job).

The habeas court’s final order faults trial counsel for failing to interview Mobley’s maternal grandmother, who would have provided information that Mobley’s mother had been a perfectionist who sometimes yelled at Mobley when he was a child, and often sent him to his room. The habeas court further notes that trial counsel failed to obtain his birth records, which showed that Mobley may have had some difficulty breathing when he was a newborn. In addition to the grandmother’s testimony and the birth records, there was evidence presented that Mobley’s parents had used inconsistent parenting styles in trying to deal with their difficult son. Mobley’s father had tried spanking him, sometimes with a belt, but had given that up when it did not work. A psychologist testified that Mobley’s family had been a “classic family of alcohol,” but Mobley’s father denied being an alcoholic or having a drinking problem.

The adequacy of a trial attorney’s investigation into a defendant’s background is judged by a standard of reasonableness.26 In the present case, Mobley’s counsel thoroughly investigated the information reasonably available to them before making their strategy decisions.27 Trial counsel subpoenaed every available childhood psychological and school record and thoroughly reviewed them. They consulted doctors who had treated Mobley as a child for possible psychological mitigation evidence. They questioned most of the available Mobley family members, including Mobley himself, and all denied the existence of any “traditional mitigation evidence.” We conclude that trial counsel’s investigation was more than adequate and was reasonable under all the circumstances. In cases where we have held *641that lawyers failed to conduct an effective investigation, trial counsel either did no investigation or had obvious avenues of investigation available to them that reasonable counsel would have pursued and they did not pursue.28 The failure of trial counsel to uncover every possible favorable witness does not render their performance deficient. We recognize that post-conviction counsel will almost always be able to identify a potential mitigation witness that trial counsel did not interview or a record that trial counsel did not obtain.

Further, even if the mitigating evidence presented at the habeas evidentiary hearing had been presented at trial, Mobley has not shown actual prejudice. The facts presented at the habeas hearing, considered collectively, show no extenuating circumstances in Mobley’s childhood, especially in light of Mobley’s early violent behavior. Additionally, the state presented substantial evidence in aggravation, including his prior convictions, lack of remorse, boasting of the crime, contemptuous comments regarding the victim, and sexual assaults of other inmates while awaiting trial. The aggravating evidence presented by the state clearly showed Mobley’s proclivity toward violence, future dangerousness, and utter lack of remorse over the murder. We conclude beyond a reasonable doubt that the limited additional mitigation evidence concerning Mobley’s childhood presented at the evidentiary hearing would not have changed the outcome of Mobley’s trial.29

B. Failure to seek funds for an independent psychologist.

In its order, the habeas court found that Mobley’s counsel were ineffective for failing to seek funds to hire an independent psychologist. The trial court had ordered a mental evaluation of Mobley and the state psychologist reported that Mobley was sharp-witted with above-average intelligence. Mobley had no delusions or hallucinations and was competent to assist in his defense.

The evidence shows that Mobley’s counsel had reviewed all the previous psychological reports and had consulted several doctors who had treated Mobley as a child. In addition, trial counsel planned to call a psychologist who had treated Mobley when he was sixteen years old to testify on Mobley’s behalf in the sentencing phase. Trial counsel considered hiring an expert psychologist but decided against *642it because the previous psychological reports from Mobley’s childhood were not favorable to the mitigation case and any future tests were therefore unlikely to be favorable. Thus, the decision not to pursue an additional psychological examination was a reasonable strategic decision. Although other attorneys might have explored the mental issue further, we cannot conclude that trial counsel’s psychological investigation and tactical judgment were outside the wide range of reasonably effective assistance.30

C. Failure to accept Mobley’s father’s funds for a genetic test.

While searching for a reason to explain Mobley’s behavior, Summer read an article about a researcher in the Netherlands who had studied a possible genetic basis for violent behavior. The study tracked several generations of a family that had a number of members with histories of violent and antisocial behavior, and the researcher discovered a possible link between this behavior and a point mutation in a structural gene for monoamine oxidase. According to the article, this study was the first time researchers had discovered a genetic disturbance directly related to violent or antisocial behavior.

Summer investigated the use of this genetic link to violence as a possible mitigation defense. He contacted a professor of genetics at Harvard Medical School and reviewed the case with her. She informed Summer that he needed to conduct a Mobley family history to determine if there was a prevalence of violent behavior that might suggest a genetic predisposition. Summer then spoke with Joyce Ann Elders, Mobley’s father’s cousin, who was considered the family historian. She revealed to Summer that a number of people in the last three generations of the Mobléy family had exhibited aggressive and antisocial behavior. These people included a murderer, a rapist, an armed robber, several substance abusers, and several spouse abusers.

Summer then contacted the National Psycho Pharmacology Lab and a doctor at the Emory Nuclear Medicine Laboratory about the cost of preliminary testing to determine if Mobley had the same gene mutation as the people in the Netherlands study. Summer filed a “Motion for Continuance and Motion to Provide Funds for Expert Witness Assistance to Conduct Prehminary Analysis for MAOA Deficiency and other Genetic Analysis as Information Becomes Available” and supported the motion with his research, affidavits from doctors, and testimony from the family historian. After a hearing, the trial court denied Mobley funds to hire a geneticist and funds to con*643duct the genetic testing because the Netherlands family had a prevalence of mental retardation and Mobley’s family did not.31 Mobley’s father offered to pay for the genetic testing himself, but Summer declined. The habeas court ruled that trial counsel was ineffective for failing to accept Mobley’s father’s offer to pay for the genetic test.

The evidence shows that trial counsel decided against going forward with the genetic test because (1) they believed that there was only a slim chance that the genetic test would reveal that Mobley had the same genetic defect as the people in the Netherlands study, (2) they intended to present a genetic defense without the geneticist, by using the family historian, and (3) they believed that the trial court’s denial of funds might constitute reversible error and they wanted to preserve the issue for appeal. The first two reasons for declining the father’s offer were reasonable conclusions. Trial counsel’s desire to “preserve” the issue for appeal, however, was not a reasonable basis for the decision. We do not condone trial counsel foregoing a vigorous defense in the hope that a ruling might constitute error on appeal. If Mobley had prevailed on this issue on appeal, the reversal of the trial court would only result in a new trial in which counsel would have the assistance of a genetics expert, which is the same relief Mobley’s father offered in the beginning.

Nevertheless, we conclude that the decision regarding the genetic test was a reasonable strategic decision and therefore did not constitute ineffectiveness by trial counsel. The presence of one flawed reason does not render counsel’s otherwise reasonable strategic decision deficient performance. Additionally, Mobley was able to present the genetics theory and there has been no showing that a geneticist would have offered additional significant evidence. Therefore, even if the decision constituted deficient performance, Mobley has not demonstrated prejudice.

D. The mitigation defense.

The habeas court ruled that trial counsel presented an inadequate mitigation defense. Specifically, the habeas court found that, after being denied the funds for a geneticist, Mobley’s counsel nonetheless persisted with presenting an unconventional genetic theory for Mobley’s behavior, without the assistance of an expert witness. We disagree because this ruling fails to take into account the fact that the genetic defense was only one part of trial counsel’s overall mitigation strategy.

Mobley’s counsel’s strategy in the penalty phase centered around the following theme: Mobley has a personality disorder that has affected his behavior since he was a child, this behavior may be the *644result of a genetic problem that he cannot control, the jury should show him mercy because people with personality disorders tend to “mellow out” as they age, and Mobley has accepted responsibility for his crimes by cooperating with the police and offering to plead guilty. Trial counsel adequately prepared the witnesses for their testimony and called Mobley’s father as the first mitigation witness. He testi- . tied about Mobley’s childhood problems and all the schools and juvenile institutions that Mobley had attended. He stated that a doctor had told him that his son would “mellow out” by age 40, and he also pleaded with the jury to spare his son’s life. The next witness was Dr. J. Steven Ziegler, who had evaluated Mobley when he was almost seventeen years old. Dr. Ziegler testified that Mobley had a conduct disorder that would develop into a personality disorder and that a person who has this disorder displays antisocial personality characteristics. Dr. Ziegler further testified that people with conduct or personality disorders tend to display fewer problems as they age, and that there were some experimental drugs that had shown promise in treating people like Mobley. Next, Summer himself testified that Mobley had been willing to accept responsibility for his actions by pleading guilty and accepting the maximum years for each crime.32 The following witness was another lawyer, who testified that it would have been unprofessional for Summer to plead Mobley guilty without some type of assurance that death would not be imposed. Lastly, Joyce Ann Childers testified about the prevalence of people with criminal or behavioral problems in the last three Mobley family generations. She also testified that these people had tended to exhibit fewer aggressive tendencies (“mellow out”) as they reached middle age. In his closing argument, Summer argued that Mobley’s personality disorder had a genetic basis, that the jury should not sentence him to death for a disorder that may lessen with age, and that medical technology may be able to treat in the future.

The appropriate test for determining whether trial counsel was constitutionally effective is “ ‘whether some reasonable lawyer at the trial could have acted, in the circumstances, as defense counsel acted at trial.’ ”33 The purpose of an analysis of a defendant’s Sixth Amendment right to effective counsel is not to grade trial counsel’s performance, but is simply to ensure that the adversarial process at trial worked adequately.34 For this reason, we must be “highly deferential *645to those choices made by defense counsel in the conduct of a trial that are arguably dictated by a reasonable trial strategy.”35

Although the habeas court focused on trial counsel’s genetic theory, it was only a part of the overall mitigation strategy. Further, a novel defense or mitigation theory is not unreasonable simply because it is unorthodox. We also do not believe that it was unreasonable for trial counsel to proceed with the introduction of the genetic theory without a geneticist, by using only the family historian. The simple argument that characteristics or traits, including violent behavior, may be passed from one family generation to another is not so technical that a layperson could not grasp it without the assistance of an expert. Therefore, we conclude that Mobley’s counsel adopted a mitigation strategy that was not unreasonable under the circumstances.

Since the evidence at the habeas hearing does not support Mobley’s burden, we reverse and reinstate the death sentence.

4. The habeas court correctly ruled that trial counsel had no conflict of interest during Mobley’s trial.

Judgment reversed in Case No. S98A0410 and affirmed in Case No. S98X0411.

dissenting.

I need not address the habeas court’s grounds for granting the writ and vacating Mobley’s death sentence in this case because I would hold that Mobley’s death sentence should be set aside and the case retried on the sentencing phase for the reasons set forth in my dissent in Mobley v. State, 265 Ga. 292, 301 (455 SE2d 61) (1995). Accordingly, I must dissent to the majority’s opinion.