(concurring in the judgment, with whom Botsford and Lenk, JJ., join.) I agree with the court that, evaluating this claim on the merits, the court must affirm the defendant’s conviction. In the circumstances of this case, there was no substantial risk of a miscarriage of justice arising from the judge’s error in instructing the jury that, to prove voluntary manslaughter, the Commonwealth bears the burden of proving beyond a reasonable doubt that the defendant injured the victim in the heat of passion or by using excessive use of force in self-defense. By finding the defendant guilty of murder in the first degree under the theory of extreme atrocity or cruelty, the jury essentially rejected the defendant’s version of events beyond a reasonable doubt, including the version that suggested that the killing occurred in the heat of passion on a reasonable provocation or through excessive use of force in self-defense. I write separately because I do not agree with the court’s conclusion that this issue is not “new” and therefore need not be decided on the merits.
In his closing instructions, the judge told the jury:
“The factor that distinguishes voluntary manslaughter from murder is not the absence of intent but rather the absence of malice aforethought. So in order to prove a defendant guilty of voluntary manslaughter, . . . the Commonwealth must prove three elements beyond a reasonable doubt, first that the defendant inflicted an injury upon the victim . . . from which he died; secondly, that the defendant injured [the victim] ... in the heat of passion or using excessive force in self-defense and, thirdly, that the homicide was committed unlawfully without legal excuse or justification.”
Later in his instructions, the judge said: “The Commonwealth has the burden of proving beyond a reasonable doubt that the defendant did not act in the heat of passion or sudden provocation and that there was a causal relation between any such provocation and the killing.” The first instruction is error because it incorrectly told the jury that they could not convict the defendant of voluntary manslaughter unless the Commonwealth proved beyond a reasonable doubt the absence of malice, which here meant that the defendant acted in the heat *5of passion on reasonable provocation or used excessive force in self-defense. Commonwealth v. Acevedo, 427 Mass. 714, 716 (1998) (Acevedo). The second instruction correctly instructed the jury that the Commonwealth was required to prove beyond a reasonable doubt that the defendant did not act in the heat of passion or from sudden provocation, id., but then the judge erred when he placed on the Commonwealth the burden of proving beyond a reasonable doubt “a causal relation between any such provocation and the killing.” To find such a causal relationship beyond a reasonable doubt, the jury would have to find provocation beyond a reasonable doubt, which infects with error even the correct instruction in the first part of the sentence.
On direct review, after rejecting the defendant’s claim that the judge’s instruction on self-defense resulted in a substantial likelihood of a miscarriage of justice, the court characterized as a “somewhat more plausible claim” that the defendant acted in the heat of passion and therefore was guilty of voluntary manslaughter. Commonwealth v. Johnson, 426 Mass. 617, 621 (1998). The court noted that the defendant argued that the judge’s charge to the jury “did not adequately make clear that the Commonwealth has the burden of disproving that the defendant killed in the heat of passion on reasonable provocation.” Id. The court rejected the argument in two sentences:
“But the judge twice stated in his charge that the Commonwealth has the burden of proving beyond a reasonable doubt that the defendant did not act in the heat of passion or on sudden provocation. Once again, the defendant did not object to these instructions, and viewing them as a whole we see nothing approaching a substantial likelihood of a miscarriage of justice.”
Id. The court’s characterization of the judge’s charge was factually incorrect. Not only did it fail to acknowledge that the judge had erred in instructing that, to prove voluntary manslaughter, the Commonwealth has the burden of proving heat of passion on reasonable provocation beyond a reasonable doubt, but the court declared that the judge had twice provided the correct instruction, where the judge had only once provided a proper instruction and even that was later infected with error.
Under G. L. c. 278, § 33E, after a defendant’s conviction of *6murder in the first degree has been affirmed, an appeal from an issue raised in a subsequent motion for a new trial may be decided by the full court only if a single justice grants leave “on the ground that it presents a new and substantial question which ought to be determined by the full court.” G. L. c. 278, § 33E. Here, the single justice reserved and reported the case to the full court, limited to review of the Acevedo error in the judge’s jury instructions, and allowed the full court to decide whether the issue presented is “new and substantial,” and, if it is, to resolve the appeal on the merits.
The first question, then, is whether an issue is “new and substantial” under G. L. c. 278, § 33E, where the issue was briefed and addressed by the court on direct appeal, but where the court misstated the essential facts underlying the claim of error and therefore never truly confronted the issue. If we are to remain true to the spirit of § 33E, we must recognize that the issue here is “new and substantial” under § 33E, and should be decided on the merits.1
Under § 33E, we review the entire record of a conviction of murder in the first degree, examining both the law and the evidence, considering the issues raised on appeal, the issues raised through objections at trial, and the issues that reasonably should have been raised on appeal and objected to at trial, to ensure that there has not been a miscarriage of justice. See Commonwealth v. Gunter, 459 Mass. 480, 485-486, cert. denied, 132 S. Ct. 218 (2011). “This uniquely thorough review ... is warranted by the infamy of the crime and the severity of its consequences.” Dickerson v. Attorney Gen., 396 Mass. 740, 744 (1986). Because our plenary review under § 33 is intended to be comprehensive, collateral review is intended to be narrow, limited to questions that are “new and substantial.” G. L. c. 278, § 33E. “In this manner, a full quorum of the Supreme Judicial Court is spared the task of hearing frivolous or *7duplicative claims.” Dickerson v. Attorney Gen., supra at 744-745.
I agree that an issue is generally not “new” under § 33E, where it has been raised by the defendant and addressed by the court in its opinion on direct appeal. See Commonwealth v. Gunter, supra at 487. I also agree that an issue is generally not “new” under § 33E, where it is not specifically addressed by the court in its opinion on direct appeal but where there is reason to be confident that the issue has been considered and rejected by the court under applicable law through its plenary review. See Commonwealth v. Smith, 460 Mass. 318, 320-321 n.1 (2011). But this case presents the unusual, hopefully extraordinary, case where the issue was raised by the defendant on direct appeal and addressed by the court in its opinion, but the facts underlying the issue were incorrectly stated by the court and therefore the court never confronted the trial judge’s error. In such unusual circumstances, there is no reason to be confident that the legal issue was fairly considered and rejected by the court. According to the court’s incorrect characterization of the judge’s charge, the judge correctly described the government’s burden so there was no reason to consider or decide whether the Acevedo error resulted in a substantial likelihood of a miscarriage of justice. For all practical purposes, then, the Acevedo error raised by the defendant on collateral review is “new” in that it has yet to be fairly considered by the court. The claim is not “duplicative” because the court on direct appeal never decided the claim based on an accurate understanding of the judge’s erroneous charge.
A few months ago, we stated, “Regardless of the nuances of the ‘new and substantial’ standard and its application in different contexts, the court in the past has expressed confidence — and we remain confident today — that single justices faced with gatekeeper applications under § 33E will allow cases to proceed to the full court in all meaningful matters . . . and will correctly screen out those that do not warrant another appeal.” Commonwealth v. Smith, supra at 322. See Leaster v. Commonwealth, 385 Mass. 547, 550 (1982) (“We have every confidence that the single justices in all such cases will allow access to the full court in meaningful matters . . .”). The court *8repeats that assertion today. Ante at 3 (“We continue to adhere to the principle that single justices . . . faced with gatekeeper applications under § 33E, will allow cases to proceed to the full court in all meaningful matters”). By this assertion, we declared that we expect single justice gatekeepers to find an issue “new and substantial” in “all meaningful matters,” which in practice means in all matters where there is a meaningful risk of a miscarriage of justice. But this confidence is misplaced and this expectation unjustified if we define “new” so narrowly that the full court never gets the opportunity to decide legal issues that were raised on direct appeal but never fairly considered because the court mistakenly stated the facts and therefore concluded that there was no legal issue. Under the court’s reasoning, the single justice gatekeeper should have denied an appeal to the full court because the issue was not “new,” even if the gatekeeper believed that there might be a meaningful risk of a miscarriage of justice arising from the Acevedo error. We cannot have it both ways: we cannot assert that gatekeepers will grant leave to appeal from any petition that demonstrates the potential of a meaningful risk of a miscarriage of justice while we assert a legal standard for a “new” claim that would bar such an appeal.
Section 33E was amended twelve years after the execution of Nicola Sacco and Bartolomeo Vanzetti following their convictions of murder in the first degree. The spirit of the new legislation was succinctly summarized in a report of the Judicial Council of Massachusetts in 1927, which proposed the statutory language that was ultimately enacted into law: “It is vital that our Courts do justice; it is also vital that people know that they do justice.” Third Report of the Judicial Council of Massachusetts, Pub. Doc. No. 144, at 40 (1927). We neither ensure that we do justice in a case of murder in the first degree nor ensure the public’s confidence that justice is done where we fail to address on the merits an issue that was never fairly considered because the underlying facts were mistakenly presented by the court on direct appeal. Where, as here, a judge’s error potentially may create a meaningful risk of a miscarriage of justice, we ensure that justice is done only by confronting the error and *9determining on the merits whether the error in fact did create a substantial risk of a miscarriage of justice.2
I join in the court’s judgment because I agree that, in view of the evidence in the case and the jury’s finding of extreme atrocity or cruelty, the error here did not produce a substantial risk of a miscarriage of justice. But I do not agree that justice would be done in this case if the full court had not made such a determination after fairly considering the error in the judge’s charge.
I will focus only on whether the issue is “new” because it is plain that the issue is “substantial,” as that term has been defined in the context of the gatekeeper provision of G. L. c. 278, § 33E. In Commonwealth v. Gunter, 459 Mass. 480, 487, cert. denied, 132 S. Ct. 218 (2011), we declared that an issue is “substantial” if it presents “a meritorious issue in the sense of being worthy of consideration by an appellate court,” and recognized that this hurdle to obtain full court review was “not high.” See Dickerson v. Attorney Gen., 396 Mass. 740, 743-744 (1986).
We would not open the door to a plethora of frivolous collateral appeals if we were to conclude that the issue here is “new and substantial.” Because of the care we take in cases of murder in the first degree, I have no reason to believe there are many cases with a factual inaccuracy central to a legal claim as significant as that found here.