Gerald R. Hooper appeals from the district court judgment denying his petition for a writ of habeas corpus. In 1979 a Martinsville, Virginia grand jury indicted Hooper on charges of murder, malicious wounding, and using a firearm in the commission of a felony. Hooper was indigent, and the Virginia state court appointed two attorneys to represent him. Pursuant to a plea agreement, Hooper pled guilty to a reduced charge of second degree murder and to the two other charges. He was sentenced to serve three consecutive terms totaling forty-one years. In the district court proceedings on his petition for a writ of habeas corpus, Hooper claimed that his plea was invalid because it was the product of ineffective assistance of counsel. The district court found the representation Hooper received was not constitutionally deficient and denied his petition.

On appeal, Hooper argues his counsel was ineffective because they failed to investigate the possibility of an insanity defense and allowed him to accept the plea agreement before obtaining an evaluation of his mental condition as of the time he committed the crime. Although we find that the legal representation provided by Hooper's counsel fell below the level of reasonable competence required of attorneys representing a criminal defendant, we conclude that the attorneys' errors did not prejudice Hooper. Accordingly, we affirm the district court's denial of his petition.

On December 25, 1978, Hooper shot and killed a fifteen-year-old girl and seriously wounded an eighteen-year-old boy. Hooper had never met the young couple, but apparently believed the boy was the person with whom he had argued on a citizen's band radio. The victims were parked in a car in a "lovers' lane" near Martinsville, Virginia. Hooper approached the car and fired several shots into it, striking both occupants. The girl's body was found a few feet from the car, and the evidence indicated that she had been shot in the head from close range after she had left the car. Hooper admitted to his sister that he had committed the crimes, and his sister provided this information to the police.

The state circuit court appointed two lawyers to represent Hooper, Junius Warren and Howard Beck. Hooper's family members told the lawyers that they believed he was mentally ill, and Hooper told them that he wanted psychiatric treatment. The lawyers also learned that in 1974, during a prior period of imprisonment, a prison psychologist had diagnosed Hooper as having a personality disorder. After receiving this information, Hooper's lawyers attempted to have him examined at the Institute of Law, Psychiatry and Public Policy at the University of Virginia (the Institute) to determine whether he was fit to stand trial and to evaluate the possibility of an insanity defense.

Doctors at the Institute examined Hooper and determined that he was competent to stand trial. They refused, however, to discuss the shootings with Hooper or to evaluate Hooper's sanity as of the time of the shootings. The doctors realized the prosecutor would have access to their report and were concerned that Hooper's fifth amendment rights be protected. The Institute suggested to Hooper's lawyers that they obtain a court order allowing a sanity examination but preventing disclosure of the results to the prosecution. The lawyers delayed three months before eventually filing a motion seeking such a protective order. The district court had yet to rule on the motion at the time Hooper entered his plea.

In the course of the lawyers' contacts with the Institute, one of the lawyers discussed Hooper's situation with the Institute's legal counsel, who told him it was unlikely that further examination would indicate Hooper was legally insane at the time he committed the crimes. The lawyer did not confirm this report by talking directly to the psychiatrists. According to the lawyer's testimony at the habeas corpus hearing, he and his associate did not discuss with Hooper Hooper's mental condition at the time of the shootings, and for that matter, purposely avoided discussing with him his version of the crime.

While Hooper's motion for a psychiatric evaluation with a protective order was pending, the prosecutor approached Hooper's lawyers and offered the plea agreement that Hooper ultimately accepted. The agreement provided that Hooper plead guilty to second degree murder instead of first degree murder and also plead guilty to the charges of malicious wounding and use of a firearm in the commission of a felony. The State agreed to recommend that he receive a twenty-year sentence on each of the first two charges and a one-year sentence on the third charge. Two days later Hooper's lawyers presented the prosecution's offer to him. They explained the available options and told Hooper the chances were poor that he would receive a diagnosis from the Institute that would be helpful in his defense. They made no recommendation whether he should accept the offer, but advised him that he must decide promptly because they believed the offer might soon be withdrawn. Hooper chose to accept the plea.

In the course of the 1986 federal habeas corpus proceeding, Hooper again was evaluated at the Institute. This time, a psychiatrist attempted to reconstruct Hooper's mental condition at the time of the shootings and concluded that Hooper's description of the incident was "consistent with his having been in a dissociative state for a portion of the time at issue." The psychiatrist stated in his report that this diagnosis would be probative of criminal insanity as it is defined in Virginia.1 He also stated, however, his opinion that if he had examined Hooper in 1979 and had received the same responses he received from Hooper in the 1986 examination, he would not have concluded that Hooper was legally insane at the time of the shootings. Hooper testified at his habeas corpus hearing that if in 1979 he had had available psychiatric evidence similar to the 1986 evaluation, he would not have accepted the plea bargain, but instead would have gone to trial.

Appeals from lower court habeas corpus decisions involving ineffective assistance of counsel claims are seldom routine. Guidance from the Supreme Court is by now so well-developed that we have little problem understanding the applicable legal standards. The Supreme Court enunciated the basic test for ineffective assistance of counsel in Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). A convicted defendant claiming ineffective assistance "must show that counsel's representation fell below an objective standard of reasonableness," id. at 687-88, 104 S.Ct. at 2064-65, and that "there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different," id. at 694, 104 S.Ct. at 2068. Applying these standards to the facts of individual cases, however, practically always requires judicial submersion into the circumstances in which decisions were made and, more particularly, a detailed analysis of how a defendant's lawyers prepared and presented his defense.

Hooper's primary allegation of ineffectiveness is that his lawyers failed to investigate his potential insanity defense. He complains that his lawyers never discussed with him the details of his crime or his mental state at the time of the shootings. Hooper stresses that they were aware of the possibility he was legally insane at the time of the shootings and that insanity was the only conceivable defense available to him. Nonetheless, the lawyers failed to obtain an evaluation of Hooper's sanity as of the time of the shootings, which would have allowed a reasoned evaluation of the likelihood that an insanity defense would have been successful. Hooper acknowledges that his lawyers did speak with an attorney on the Institute's staff, who told them the chance of obtaining a helpful diagnosis was slim. He points out, however, that they did not confirm this report with any of the Institute's psychiatrists and that they misled him concerning the source of the information.2

We agree with Hooper that his lawyers' representation was " 'not within the realm of competence demanded of attorneys in criminal cases,' " Strickland, 466 U.S. at 687, 104 S.Ct. at 2064 (quoting McMann v. Richardson, 397 U.S. 759, 771, 90 S.Ct. 1441, 1449, 25 L.Ed.2d 763 (1970)). Effective assistance includes the "duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary." Id. 466 U.S. at 691, 104 S.Ct. at 2066. Hooper's lawyers knew that Hooper might have an insanity defense and that, realistically, he had no other defense. They determined that a sanity examination was necessary, but failed initially to file an appropriate motion that would have allowed the examination to take place without revealing the resulting report to the prosecution.3 They compounded this error by delaying the filing of a corrected motion. The result was that Hooper made his choice without the benefit of a psychiatric evaluation of his mental condition at the time he committed the crimes.

Under the circumstances, the minimum standards of competency required Hooper's counsel to exhaust the possibilities inherent in such an examination. An evaluation might not have produced results helpful to Hooper's defense. Indeed, the statement of the Institute's legal counsel suggests that would have been the case. Nevertheless, knowing the results of such an examination, whatever they might have been, would have assisted Hooper in making his choice. Even in light of the "heavy measure of deference" we must accord attorneys' judgment in evaluating the effectiveness of representation, Strickland, 466 U.S. at 691, 104 S.Ct. at 2066, the failure of Hooper's counsel adequately to investigate the possibility of an insanity defense fell below the required standard of competence.

Having found that the lawyers' deficient investigation constitutes ineffective representation, we must next consider the second prong of the Strickland test--whether this deficiency prejudiced Hooper. See 466 U.S. at 693-694, 104 S.Ct. at 2067-68. When a defendant challenges a conviction entered after a guilty plea, "prejudice" prong of the test is slightly modified. Such a defendant "must show that there is a reasonable probability that, but for counsel's errors, he would not have pleaded guilty and would have insisted on going to trial." Hill v. Lockhart, 474 U.S. 52, 59, 106 S.Ct. 366, 370, 88 L.Ed.2d 203 (1985).

Although this modification focuses the inquiry on a subjective question, the answer to that question must be reached through an objective analysis. The Supreme Court outlined the appropriate mode of analysis in Hill:

[W]here the alleged error of counsel is a failure to investigate or discover potentially exculpatory evidence, the determination whether the error "prejudiced" the defendant by causing him to plead guilty rather than go to trial will depend on the likelihood that discovery of the evidence would have led counsel to change his recommendation as to the plea. This assessment, in turn, will depend in large part on a prediction whether the evidence likely would have changed the outcome of a trial.

Id. Hooper's lawyers made no recommendation in favor of pleading guilty. Consequently, our prejudice analysis test must center on whether knowledge of additional information would have led Hooper to decline the plea agreement or would have led his attorneys to recommend that he do so.

Like the magistrate, we do not put great weight in Hooper's representation that if, prior to accepting the plea bargain, he had received a psychiatric report similar to the one issued in 1986, he would have chosen to go to trial. Although it carries some probative value, such a statement suffers from obvious credibility problems and must be evaluated in light of the circumstances the defendant would have faced at the time of his decision.

The circumstances Hooper confronted at the time he accepted the plea agreement were grim. Hooper was aware that he had committed a heinous crime and that the evidence against him was overwhelming. The 1986 psychiatric evaluation was equivocal evidence at best. Although the report stated Hooper sometimes was in a dissociative state, it would have remained his burden to show he was in such a state at the time of the shooting. Even if the jury found he had been in such a state, whether it would have found he was legally insane also remains in doubt. While the psychiatrist's ultimate conclusion that Hooper was not legally insane at the time of the shooting would not have been admissible, see Waye v. Commonwealth, 219 Va. 683, 251 S.E.2d 202, 210, cert. denied, 442 U.S. 924, 99 S.Ct. 2850, 61 L.Ed.2d 292 (1979), it is indicative of the difficulties Hooper would have faced in pursuing an insanity defense.

The burden is on Hooper to establish a reasonable probability that if his lawyers had obtained a psychiatric report, he would have rejected the plea agreement. Hill, 474 U.S. at 59, 106 S.Ct. at 370. Even if Hooper had had a psychiatric report similar to the one produced in 1986, his only rational conclusion would have been that he faced serious obstacles in attempting to pursue an insanity defense. By accepting the plea agreement, Hooper was able to have the principal charge against him reduced to second degree murder. Moreover, Hooper's speculation on what results a complete 1979 psychiatric examination would have produced is tenuous at best.4 Examining the 1986 psychiatric report against the background of the circumstances Hooper faced at the time he decided to plead guilty, we are not persuaded that it is sufficient evidence to establish a reasonable probability that had Hooper's lawyers obtained a sanity evaluation in 1979, Hooper would have declined the plea bargain in favor of a jury trial.

Hooper also contends that his lawyers were ineffective in allowing him to accept the plea. He argues that in presenting a sudden offer and urging him to act on it hastily without waiting for the psychiatric report, the lawyers fell below the required level of professional conduct. We disagree.

In resolving ineffective assistance claims, we evaluate "[t]he reasonableness of counsel's performance ... from counsel's perspective at the time of the alleged error and in light of all the circumstances and the standard of review is highly deferential." Kimmelman v. Morrison, 477 U.S. 365, 106 S.Ct. 2574, 2586-87, 91 L.Ed.2d 305 (1986) (citing Strickland, 466 U.S. at 689, 104 S.Ct. at 2065). We find the lawyers' decision not to advise Hooper against accepting the plea bargain entirely reasonable under the circumstances. One of Hooper's lawyers testified that he believed the prosecutor offered the plea agreement out of concern the report from the Institute would support an insanity defense. At that stage in the proceeding, Hooper's lawyers possessed information of which the prosecutor was unaware--that the Institute's report was unlikely to provide a strong basis for such a defense. They also believed that the prosecutor would withdraw his offer if Hooper did not accept it promptly. They explained to Hooper the consequences of the plea and the possible sentences he faced if he were convicted at trial. Under the deferential standard set out in Strickland, 466 U.S. at 693-94, 104 S.Ct. at 2067-68, the lawyers' decision not to recommend that Hooper reject the plea agreement was not ineffective representation.

Although the failure of Hooper's lawyers to investigate his defense fell below the objective standard of reasonableness mandated by Strickland, we cannot conclude there is a reasonable probability that but for their deficiency, Hooper would have chosen to reject the plea agreement. Accordingly, the district court's denial of the writ is affirmed.

AFFIRMED.