dissenting. I disagree with the majority’s conclusion that in order to prove prejudice under Strickland v. Washington, 466 U.S. 668, 104 S. Ct. 2052, 80 L. Ed. 2d 674, reh. denied, 467 U.S. 1267, 104 S. Ct. 3562, 82 L. Ed. 2d 864 (1984), the petitioner must estab*466lish that, because of appellate counsel’s deficient performance, “there remains a probability sufficient to undermine confidence in the verdict that resulted in his appeal.” In other words, instead of only proving that the ineffective assistance of appellate counsel would have produced a different result in the appeal, the petitioner must go further and prove that the appellate counsel’s failure leaves in place “an unreliable determination of his guilt.”
The majority finds support for its conclusion in an Iowa Supreme Court decision; Gering v. State, 382 N.W.2d 151, 156 (Iowa 1986); and rejects the judgment of the only two federal circuits that have ruled on the issue. See United States v. Williams, 934 F.2d 847, 853 (7th Cir. 1991) (under Strickland, “result of the proceeding” means outcome of appeal); Sharp v. Puckett, 930 F.2d 450, 452-53 (5th Cir. 1991) (under Strickland, prejudice means that outcome of appeal would have been different); Lockhart v. McCotter, 782 F.2d 1275, 1283 (5th Cir. 1986), cert. denied, 479 U.S. 1030, 107 S. Ct. 873, 93 L. Ed. 2d 827, reh. denied, 480 U.S. 911, 107 S. Ct. 1360, 94 L. Ed. 2d 530 (1987) (petitioner failed to demonstrate reasonable probability that, but for appellate counsel’s errors, result of appeal would have been different); Gray v. Greer, 778 F.2d 350, 353 (7th Cir. 1985), vacated, 478 U.S. 1017, 106 S. Ct. 3328, 92 L. Ed. 2d 734 (1986), on remand, 800 F.2d 644 (7th Cir. 1986) (question is whether appellate counsel failed to raise issues that “would have been clearly more likely to result in reversal or an order for a new trial”); see also Miller v. Keeney, 882 F.2d 1428, 1434 n.9 (9th Cir. 1989).
It stands to reason that, when determining whether there was ineffective assistance of appellate counsel, we should focus on whether the result of the appeal *467would have been different.1 “The right to appellate counsel is now firmly established. . . . Strickland v. Washington, [supra,] established the standard for ineffective assistance of counsel, and though it is phrased in terms of ineffective assistance of trial counsel, it can be used as a basis for establishing a standard for effective assistance of appellate counsel. . . . Had appellate counsel failed to raise a significant and obvious issue, the failure could be viewed as deficient performance. If an issue which was not raised may have resulted in a reversal of the conviction, or an order for a new trial, the failure was prejudicial.” (Citations omitted.) Gray v. Greer, supra, 800 F.2d 646. In the present case, appellate counsel failed to raise an issue that, had it been raised, would have resulted in a new trial for the petitioner. That should be the remedy on this appeal. The practical result of the majority’s decision today is that if the defendant is fortunate enough to have competent appellate counsel, he or she will get a new trial when there is reversible error. If, however, the defendant is unfortunate to have been assigned or has retained appellate counsel who fails to raise the same issue that, if properly raised, would on appeal have resulted in a new trial, he or she will have the additional burden of proving that the counsel’s failure undermined the reliability of the conviction.
Finally, it appears to me that the majority in its footnote 11, is setting the stage for a 180 degree reversal *468of State v. Rodgers, 198 Conn. 53, 502 A.2d 360 (1985), for direct appeals. In Rodgers, the trial court, while instructing the jury on circumstantial evidence, stated that an inference may be drawn if “ ‘it is more probable that the fact to be inferred is true.’ ” Id., 57. This court reversed the defendant’s conviction on the basis that this instruction, which is identical to the one at issue in the present appeal, impermissibly diluted the state’s burden of proof. In reversing the conviction, we refused to follow the traditional route of reading the charge in its entirety to determine whether the jury was apt to have been misled by the instruction. We specifically held that “[t]here exists such a contradiction between the erroneous instruction and the admixture of correct doctrine that as a whole it cannot be said that the incorrect statement did not influence the jury.” Id., 59. In State v. Whelan, 200 Conn. 743, 757, 513 A.2d 86, cert. denied, 479 U.S. 994, 107 S. Ct. 597, 93 L. Ed. 2d 598 (1986), this court reasoned that “[wjhere the principal factual issue is intent, which is characteristically proven by circumstantial evidence . . . the court’s instructions regarding the use of circumstantial evidence as proof of this essential element are subject to close scrutiny.” (Citation omitted; internal quotation marks omitted.) Because intent is routinely proven by circumstantial evidence, the need for the rule is evident; without it, there exists the likelihood that a jury could convict a defendant of the crime charged without the state proving each element beyond a reasonable doubt. By today’s decision the majority eliminates this per se Rodgers-Whelan rule for collateral attacks, but claims that it is still adhering to the rule for direct appeals. That defies logic. If on a direct appeal we are concerned, as we should be, that such an incorrect statement by the trial judge probably influenced the jury, notwithstanding other ameliorating instructions, so must we also be concerned when the same *469issue is presented on habeas corpus. Accordingly, I fear that the stage has been set for the eventual demise of the Rodgers-Whelan rule.
I dissent.
To support its conclusion, the majority notes that it is “significant” that the petitioner sought “a new trial by way of a writ of habeas corpus.” It should, however, go without saying that a claim of ineffective assistance of counsel may only be pursued on a petition for a writ of habeas corpus. State v. Leecan, 198 Conn. 517, 542, 504 A.2d 480, cert. denied, 476 U.S. 1184, 106 S. Ct. 2922, 91 L. Ed. 2d 550 (1986). Indeed, when the claim is one of ineffective assistance of appellate counsel, the only possible avenue of review is by way of habeas corpus. Therefore, I fear that the true purpose of this dicta is the beginning of an assault on the writ of habeas corpus as a vehicle for justice.