concurring.
I concur in the court’s result, but not in the analysis that leads to it. The key to that analysis is the proposition that Ivie Clay’s second lawyer, Elson, rendered ineffective assistance of counsel as a matter of law. This proposition is untenable on its own terms, and is also inconsistent with the rule announced in United States ex rel. Spurlark v. Wolff, 699 F.2d 354, 357-61 (7th Cir.1983) (en banc), that a state criminal defendant may not seek federal habeas corpus on grounds he could have raised, but failed to raise, in an appeal to a state court from his conviction.
*435Miss Clay pleaded guilty, almost certainly because she was incompetently represented by her first lawyer, Friedman — for she had excellent defenses, state and federal, to the charges against her. Under Illinois law, if she wanted to appeal her conviction, entered as it had been on the basis of a guilty plea, she had first to file with the trial court a motion to vacate the conviction. Friedman did file such a motion, but it was completely inadequate and was denied. It was at this point that Elson began actively representing Miss Clay. He faced a dilemma. If he appealed, he had little or no prospect of winning, because Friedman had failed to make a record substantiating the claim that the proceedings leading up to Miss Clay’s conviction had been constitutionally infirm. As we said the last time this case was here (an unbelievable four years ago), “Given the general grounds in [Friedman’s] motion, and the fact that no argument was presented in the trial court, there would have been no basis on which the appellate court could have reversed.” 631 F.2d 516, 523 (7th Cir.1980). The best ground for an appeal would have been Friedman’s incompetent representation of Miss Clay (and if this could be shown, presumably she would be given an opportunity to repair the deficiencies in the record on her other constitutional claims); but of course Friedman had not used the motion to vacate to make the case for his own incompetence. Elson might still have been able to raise on appeal the issue of Friedman’s incompetence, but without a record demonstrating it (the kind of record later made in this habeas corpus proceeding) he would not have had a decent chance of prevailing on the appeal even to the extent of getting an order for an evidentiary hearing on the issue. Elson’s alternative to appealing was to file an amended motion to vacate. This was a long shot, because Illinois law makes no specific provision for such an animal; but Elson was greatly encouraged when Judge White set the matter for a hearing.
The .scheduling of the amended motion for a hearing made Elson’s dilemma exquisite. If he appealed from the denial of the first motion to vaeate (as he still had time to do), then under Illinois law the trial court would lose jurisdiction and Elson would lose the hearing that seemed his best shot at getting his client’s conviction overturned. But if he did not appeal, the time for appeal would lapse before the hearing on his amended motion was held. He decided not to appeal, and unfortunately for him and his client the judge before whom the hearing had been set (a different judge, Judge Rogers) decided that the amended motion was procedurally improper and dismissed it out of hand. It was now too late for Elson to file an appeal from the denial of Friedman’s original motion to vacate the conviction; and taking an appeal from the denial of the amended motion was unpromising since the denial (even assuming it was an appealable order, which is unclear) had been based on a procedural ground unlikely to be overturned. It was against this complicated background that Elson decided to seek federal habeas corpus without having first attempted to raise on appeal to a state court the constitutional claims that he would be asking the federal court to decide.
I do not see how Elson’s conduct in the state court can be called ineffective assistance of counsel as a matter of law. He acted reasonably in the unusual circumstances in which he found himself. Only after a judge had actually scheduled a hearing on the amended motion to vacate did Elson give up what we said in 1980 would have been a futile appeal. He cannot be faulted for not having anticipated that a different judge would cancel the hearing and throw out the motion as procedurally improper. If a judge offers a lawyer a hearing on a motion, not one lawyer in a hundred will pause to question the judge’s authority to grant the motion should the hearing produce evidence showing that the motion is meritorious. It is true that Illinois law makes no specific provision for an amended motion to set aside a conviction, but Elson would not have been unreasonable in thinking that Illinois judges might have implicit authori*436ty to accept an amended motion, relating back to the date of filing of the original motion, in an appropriate case. Anyway, it was better than nothing. And there was no obvious way in which Elson could preserve both his (probably worthless) right to appeal and what seemed the much more promising option of a hearing on the amended motion to vacate the conviction. Of course with the benefit of hindsight we know he should have gone ahead and appealed, not because it would have gained him anything but because it would have complied formally with the requirement, imposed by Spurlark, of giving the state appellate court a chance to correct any constitutional error committed in the trial court. But the proper perspective in evaluating a claim of ineffective assistance of counsel is ex ante, not ex post. We should not make lawyers always go by the book, when an unconventional approach holds more promise for the client, as it did here. Legal imagination should not be penalized by a concept of ineffective assistance of counsel that discourages all risk-taking on a client’s behalf.
Maybe with even more imagination Elson could have escaped from his dilemma. Maybe he should have moved to accelerate the hearing scheduled before Judge Rogers, or filed a late notice of appeal, or tried to appeal from Judge Rogers’ order dismissing the amended motion without a hearing. But a lawyer is not ineffective just because he fails to make extraordinary efforts on his client’s behalf; the constitutional standard is minimum rather than maximum compliance with the standards of the profession.
As Elson was not in fact incompetent in his efforts on behalf of Miss Clay, I am led to wonder whether my brethren may not regard any failure to appeal an appealable judgment (such as the order denying the original motion to vacate) as incompetence per se, no matter how reasonable it actually is in the circumstances. There are hints of this position in the majority opinion, but I hope I am misreading it, for automatically equating failing to appeal with failing to render effective assistance of counsel would do great damage to our recent en banc decision in Spurlark. It is true that the precise issue there was the slightly different one of a defendant’s failing to include in his state-court appeal grounds that he later tried to raise in a habeas corpus action in federal court. I assume my brethren would not think it incompetence per se for a lawyer who had taken an appeal for his client not to include every possibly meritorious ground in the appeal; the lawyer might have made a professionally sound judgment that his chances of winning would be greater if he limited his appeal to his best grounds. But nothing in Spurlark suggests that that decision should be limited to cases in which an appeal is filed but some grounds for appeal are omitted — quite the contrary, see 699 F.2d at 359-61. In any event the extension of Spurlark to the case where the lawyer fails to appeal at all — if extension it be— was taken in Williams v. Duckworth, 724 F.2d 1439, 1443 (7th Cir.1984); see also United States v. Correa-De Jesus, 708 F.2d 1283, 1285 (7th Cir.1983) (dictum). Contra, Holcomb v. Murphy, 701 F.2d 1307, 1310-12 (10th Cir.1983). Williams would be unfathomable if the court had thought that not appealing an appealable judgment is always and everywhere the infallible sign of lawyer incompetence. Then the defendant could always show good cause — the incompetence of his counsel — for failing to appeal his conviction to a state appellate court, and the failure to take such an appeal could never bar a defendant from seeking federal habeas corpus.
The idea that failure to appeal is ineffective assistance of counsel per se not only guts Spurlark as elaborated in Williams, but is wrong on its own terms. If an appeal would be futile, the failure to appeal is not a sign of incompetence. This is such a case. Williams gives another example: forgoing an appeal because of a recent and dispositive adverse precedent in the state supreme court. See 724 F.2d at 1443 n. 5.
Even though Elson acted reasonably throughout, Miss Clay has demonstrated good cause for not having pursued her state appellate remedies — which is why I concur in the majority’s result. The combi*437nation of Friedman’s incompetence with the inconsistent signals given by the two state judges made it impossible (or virtually so) for her to pursue her appellate remedies (remedies that in any event Friedman’s incompetence made unlikely to succeed despite the great substantive merit of her constitutional claims) without giving up what seemed a better remedy in the trial court — a hearing on the amended motion to vacate. True, the mere fact that making an objection seems futile on the basis of existing state-court precedents does not in itself establish good cause for failing to make it, Engle v. Isaac, 456 U.S. 107, 130, 102 S.Ct. 1558, 1573, 71 L.Ed.2d 783 (1982), and I shall assume that the same principle applies to taking an appeal. But the principle has reference to cases where there is no great cost to making the objection or taking an appeal. If Elson had taken a futile appeal, he would have given up something worth more to his client: a hearing that a state judge had offered him, at which he would be able to present the evidence essential to show that his client had been wronged. As Elson has established good cause for failing to pursue state appellate remedies, and there is no doubt that denying Miss Clay access to federal habeas corpus would do her great prejudice, the judgment of the district court must be reversed and the case remanded for a consideration — at long last — of the merits of her constitutional claims. But there is no need for us to embrace the factually and legally unfounded proposition that a decision not to appeal an appealable judgment is conclusive proof of incompetence of counsel.