CLARK, Circuit Judge,
dissenting,
in which JOHNSON, Circuit Judge, joins:I dissent from the court’s order denying in banc consideration of the panel’s decision in this case. The panel affirmed a denial of Gates’ petition for writ of habeas corpus which was grounded upon a claim of ineffective assistance of counsel. At the core of effective trial advocacy is the duty of trial counsel to know the legal rights of his client and to explore the possible factual underpinnings that would support an effort to secure those legal rights. Yet it appears that this court finds reasonable an attorney’s unresearched decision to forego a challenge to the racial composition of a jury based substantially on the attorney’s belief that such a challenge could only alienate the jury that would eventually try the defendant.
Gates was convicted of murder and sentenced to death in a case in which all of the evidence pointed to guilt. The only real issue in the state trial was whether Gates would receive the death penalty. A panel of our court found that Gates’ attorney, William Cain, was not ineffective for failing to investigate the possibility that Mus-cogee County, Georgia, had unconstitutionally composed grand and petit jury pools from which Gates’ juries had been drawn. Furthermore, defense counsel decided to forego this investigation of Gates’ rights without consulting Gates. Without undertaking any investigation, Cain foreclosed the possibility of knowing whether a jury challenge would be successful.
The panel concluded “Cain could have stated a prima facie case of jury discrimination.” 863 F.2d 1492, 1498 (11th Cir.1989). Gates was tried before a jury of twelve white persons drawn from a pool of sixty veniremen containing three or four black persons. The prosecutor peremptori*294ly challenged those blacks who reached the jury box.
The panel found that Cain's decision not to investigate was reasonable based on Cain’s experience and observations. Specifically, the panel held that Cain was experienced and conscious of the possibility of a jury composition challenge. It then held that there was only a slight inference that a jury challenge would have been consistent with Cain’s overall defense strategy. Finally, the panel held that the foreseeable prejudice that would have resulted from the challenge mitigated against launching the challenge.
Cain did not explore the jury composition issue. His preparation and investigation of Gates’ case consisted of going to the jail several times to talk to the defendant, getting the complete prosecutor’s investigative file, following up on Gates’ assertion that another individual was responsible for the crime, and interviewing most of the State’s witnesses. (Cain Dep. at 6-13.) His testimony as to his awareness of a jury challenge at the time of Gates’ trial is revealing:
My experience has shown that I do not believe that [sic] any intentional discrimination against blacks, young woman or men from either the Grand Jury array or petit juries in Muscogee County.... I do not understand that [percentage disparities] ha[ve] ever been held to be a basis of setting aside the Grand Jury array or petit jury.... In the final analysis of the question, I realize I could challenge the Grand Jury array; I realize there were certain objections that could be made as to the petit jury. I did not believe them to be in order. And from experience day in and day out over many years, we have many, many juries where the panel is at least 50 percent women, and some instances more. We have panels where a substantial percentage — even in some instances close to 50 percent— are black... ,1
Cain Dep. at 35-37.
Cain’s “practical” experience mistakenly led him to believe a jury composition challenge would be unavailing. Cain, however, conducted no investigation into the jury composition issue to determine whether such a challenge would be successful. Such an investigation would have borne out that a prima facie case of jury discrimination did exist and the challenge did have a chance of success.
Cain may have been generally aware of the availability of jury composition challenges, but the facts of this case reveal that he was not aware that such a challenge was available in Muscogee County. Cain’s assertion that he had seen juries with acceptable ratios in other cases simply begs the question; the issue is the jury composition in general, not in specific cases, especially where the jury that tried Gates was all white.
The panel considered whether the rejected line of defense was inconsistent with Cain’s overall defense strategy. However, Cain had no strategy with regard to the composition of the jury; in fact his testimony reveals that he had no real strategy at all. Thus, it is difficult to understand the panel’s conclusion that a jury composition challenge would have been, at most, only minimally consistent with Cain’s overall defense strategy. Cain testified that it was his belief that Gates had no defenses and that Gates’ only hope was that the state would not prove intent to the degree required by law. In addition, Cain testified that his only plan for the trial and the sentencing hearing was to invoke sympathy in the jury for Gates. Thus, contrary to the panel’s assertion that a jury chal*295lenge was only minimally consistent with Cain’s overall defense strategy, such a challenge was very consistent indeed.
Under Georgia law at the time of Gates’ trial, a single vote against the death penalty would have spared Gates’ life. See Hill v. State, 250 Ga. 821, 301 S.E.2d 269 (1983). Thus, Cain’s “mercy strategy” could have worked if there were only one sympathetic juror. In a case where the evidence of guilt is so overwhelming, one would expect trial counsel to seek sympathetic jurors in anyway possible.
The panel also found that the foreseeable prejudice of launching a jury challenge was substantial enough to conclude that Cain was reasonable in not pursuing this line of defense.
Cain was concerned that the jury that eventually tried Gates would become alienated against Gates if they became aware that a challenge to the county jury list had been made. This was no small concern in a case in which the only real hope of saving the defendant from the death penalty was through a plea of mercy. If such a challenge were successful, the jury commissioners would have had to create a new master jury list and to discard the old one. This would have resulted in delay and the dismissal of potential jurors already called to the courthouse under the old list. There was no way that Cain could be guaranteed that the final jury that would decide Gates’ fate, a jury selected from a new list, would be unaware of the previous challenge. The jury might very well view the challenge as an attempt by Gates to avoid punishment on a “technicality.” Lawyers and judges may realize that “technicalities” generally are legal safeguards designed to ensure that the accused receives a fair trial. Experienced criminal defense lawyers recognize, however, that this is not always the view of the layman. They therefore often go to extremes to avoid any possibility that the jury might learn that they are raising a “technicality.”
863 F.2d at 1499 (footnote omitted).
First, it is not at all clear that Cain’s rationale is valid. The chances of a new jury finding out about the challenge is minimal and is speculation on the part of the panel. Furthermore, the chance that the new jurors would be annoyed by the challenge even if they did discover it would also be minimal because the challenge would cause no imposition on the new jury.
A more fundamental reason exists, however, for rejecting such a rationale as a reasonable basis for not investigating and/or pursuing a jury challenge. Should a federal court be guided by the possibility that jurors might react adversely if efforts are made to enforce the constitutional right to a racially fair jury? This question was answered in the negative over 20 years ago by Judge Wisdom in Whitus v. Balkcom, 333 F.2d 496 (5th Cir.), cert. denied, 379 U.S. 931, 85 S.Ct. 329, 13 L.Ed.2d 343 (1964):
The petitioners and their attorneys had no desire to give up their right to be tried by a jury chosen without regard to the race of the jurors. It was not to their interest to do so — except as a choice of evils. A choice of evils was indeed the only state remedy open to them. The petitioners could choose to be prejudiced by the hostility the attack on the all-white jury system would stir up. Or they could choose to be prejudiced by being deprived of a trial by a jury of their peers selected impartially from a cross-section of the community. This is the “grisly”, hard, Hobson’s choice the State puts to Negro defendants when it systematically excludes Negroes from juries; white defendants are not subjected to this burden.
The constitutional vice is not just the exclusion of Negroes from juries. It is also the State’s requiring Negro defendants to choose between an unfairly constituted jury and a prejudiced jury. We hold that this discrimination violates both the equal protection and the due process clauses of the Fourteenth Amendment.
Id. at 498-99 (footnotes omitted). See also Wells v. Wainwright, 488 F.2d 522 (5th Cir.1973) (remanding case to district court for hearing to determine if petitioners’ fail*296ure to bring jury challenge was excusable; one question to consider was whether challenge would have caused such hostility in community that any resulting jury would be prejudiced against defendant). Because Cain was hoping to save Gates’ life, not to acquit him, even one sympathetic juror would have made the challenge worthwhile. Although a chance existed that a jury composition challenge would have been unsuccessful, Cain's decision cannot be deemed reasonable when one considers that in choosing between these “two evils,” Cain conducted no investigation and he never consulted Gates.
The panel’s acceptance of jury hostility as foreseeable prejudice that justifies the failure to investigate and/or pursue a jury composition challenge is disheartening, especially where the foreseeable prejudice in not bringing the challenge is so great. Gates was tried by an all white jury and the prosecution peremptorily challenged the three or four prospective black jurors who were on the original panel. Unlike Birt v. Montgomery, 725 F.2d 587 (11th Cir.), cert. denied, 469 U.S. 874, 105 S.Ct. 232, 83 L.Ed.2d 161 (1984), in which the attorney’s defense strategy was purposefully employed to result in a racially fair jury, Cain’s strategy was blatantly deficient of any such goal.
Although the panel dismissed, in a footnote, the idea that Cain was more concerned about alienating jurors in future cases than he was about alienating the jury in Gates’ case, it does so without explanation. The panel fails to address Cain’s comments in his deposition that:
It is my lawyer’s responsibility to handle any case as if it were his only case and, regardless of what the consequences are, to win that one no matter what. Yet on the other hand, some of these challenges that are suggested to be made in each of these instances — and time and time again in my professional opinion — would only alienate the juror and the whole judicial system in the community so that your chances, in my opinion, of getting a fair trial are less than if you did not do it. I still say that if proper grounds are there, it should be done. But it’s not just as easy as to say it should be done in every case.
(Cain Dep. at 37.) In addition, although Gates was tried by an all-white jury, Cain could not remember how many blacks were on Gates’ jury and he had no feeling that their number was unusually small. Id. at 56, 58.
Cain’s failure to investigate a potentially meritorious claim did not fall within the range of competence demanded of attorneys in criminal cases and did not conform to professional standards of reasonable investigation of the facts and understanding of the law. Birt 725 F.2d at 597.
Under any inquiry into the adequacy of Cain’s representation, Cain’s “experience” and “observations” did not render his failure to investigate reasonable. Cain asserted that he did not challenge the jury composition because it was known among the criminal defense attorneys in the county that it was not “the thing to do.” In other words, such challenges were not favored by the local bar because it was the bar’s analysis “that more harm than good can perhaps come from such a procedure.” (Cain Dep. at 37, 54.) This “logic” is perhaps the worst of all. If such a rationale is deemed “reasonable,” then this court will sanction the perpetuation of unconstitutionally composed juries. The incumbent defense attorneys will not challenge the jury composition and any new attorney that comes into the county will quickly learn that such a challenge is “not the thing to do.” Yet this court would uphold such provincial wisdom as reasonable and with its sanction, Muscogee County may never be forced to face the potential unconstitutionality of its system.
The situation in the present case is somewhat analogous to Amadeo v. Zant, 486 U.S. 214, -, 108 S.Ct. 1771, 1776, 100 L.Ed.2d 249 (1988), in which the Supreme Court held that interference by officials that made compliance with state procedural rules impractical constituted cause for purposes of federal habeas review. In this case, the defense bar has effectively precluded the assertion of the constitutional *297right to a fairly constituted jury. Cain testified that he had never challenged the composition of either the grand jury or the traverse jury. He also testified that he did not believe that such a challenge had been made within the last 4 or 5 years of the date of his deposition (in 1980). (Cain Dep. at 53, 90.)
“At the heart of effective representation is the independent duty to investigate and prepare. ‘[T]he cornerstones of effective assistance of counsel’ are the ‘[ijnformed evaluation of potential defenses to criminal charges and meaningful discussion with one’s client of the realities of his case.’ ” Goodwin 684 F.2d at 805 (citations omitted). The uninformed decision not to pursue a meritorious claim was unreasonable in this case. The situation is similar to that in Smith v. Wainwright, 111 F.2d 609 (11th Cir.1985), cert. denied, 477 U.S. 905, 106 S.Ct. 3275, 91 L.Ed.2d 565 (1986). In that case we found that, depending on what an evidentiary hearing would reveal, the attorney’s failure to conduct any investigation into the possibility of suppressing his client’s confessions because the attorney believed that a motion to suppress would be frivolous and because his client never complained of coercion could amount to ineffective assistance.
Cain conducted no investigation and a prima facie ease did exist in which Cain could have asserted a challenge to the jury composition. Cain did not discuss this potential defense with his client. As demonstrated above, the failure to investigate this meritorious claim was inexcusable. By all accounts, Cain was ineffective.
. Cain’s statement, "If in fact ... the jury lists are not truly representative of some classes specified in the petition, that could create a problem, except that I do not understand that that has ever been held to be the basis for setting aside the Grand Jury array or petit jury,” demonstrates that Cain misunderstood the law upon which he could have based a successful prima facie claim of jury discrimination in Gates’ case. See Berryhill v. Zant, 858 F.2d 633, 638 (11th Cir.1988) (bare showing of underrep-resentation of protected group enough to sustain prima facie case of discrimination and, if the state offers no plausible explanation, it is enough to hold as a matter of law that the exclusion of the group is unfair and unreasonable).