*716¶ 65 I join the court's opinion except for its conclusion that the evidence offered by Mr. McAlister is cumulative. I agree with the court's observation that "defining when such evidence is cumulative is difficult," majority op., ¶ 38, but a fairly straightforward test can establish that the evidence here cannot be so characterized.
¶ 66 The point of evidence is to give the trier of fact the raw material upon which to exercise his judgment in deciding whether a particular fact is true. We presume he will act in good faith, and will conclude that a fact is true if presented with sufficient credible evidence. It's easy enough to say that anything beyond this quantum is cumulative. The tricky part is determining whether new evidence is cumulative when the fact finder has already determined the old evidence was insufficient to establish the contested fact.
¶ 67 There is, however, a test that can unmistakably identify new evidence as non-cumulative, and we should have applied it here. It is this: If the trier of fact were to believe the new evidence, would he necessarily conclude the disputed fact has been established? If so, then the new evidence cannot possibly be cumulative because it is capable of producing a result the old evidence did not. I suspect it will be a rare piece of information that will satisfy this criterion, in which case other "cumulativeness" tests may be employed. But evidence that does meet this standard definitively answers the cumulativeness question.
¶ 68 Mr. McAlister has presented such information here. At trial, he introduced evidence suggesting that Messrs. Jefferson and Waters had a strong motivation to falsely accuse him of involvement in the crimes. The jurors could have believed this evidence-that is, they could believe the witnesses had good *717reason to lie-and nonetheless conclude that, upon the event, they told the truth. The result of the trial suggests this is, in fact, what they did. *93¶ 69 Mr. McAlister's new evidence is not of the same nature. The affidavits he now presents claim that Messrs. Jefferson and Waters admitted they made up a story about Mr. McAlister's involvement in the crimes. If the jurors were to credit this new evidence, they could not simultaneously believe that Messrs. Jefferson and Waters' trial testimony was truthful. That is to say, the new evidence is capable of producing a result the old evidence did not. Therefore, it cannot be cumulative.
¶ 70 Nevertheless, I agree with the court's conclusion that this evidence requires corroboration before Mr. McAlister is entitled to an evidentiary hearing on his request for a new trial. This new information is in the nature of "recantation" evidence (for which we have always required corroboration), even though it presents as a pre-existing decision to commit perjury rather than a post hoc confession. The justification for requiring corroboration is the same-the inherent unreliability of what often looks like a grown-up version of Kipling's "just-so" stories.
¶ 71 There is one other aspect of the court's opinion that bears comment. The court took the State to task for conceding that Mr. McAlister's new evidence is not cumulative: "We are always disappointed when counsel concedes a difficult issue, as counsel for the State has done here." Id., ¶ 36 n.7. I disagree.
¶ 72 One of the distinguishing characteristics of an accomplished and wise advocate is knowing when to concede a point. And, having come to the realization that a previously-defended position is not actually defensible, it takes courage and humility to say so.
*718This is the type of candor we should be encouraging, not condemning. Just because the court (mistakenly, in my view) disagrees with the State's position does not mean the State conceded for some reason other than its professional, good faith assessment of the issue's merits. And yet the court's rebuke implies that very thing. However, institutional litigators (like the State) should make certain that a concession truly is the result of a good faith assessment of the issue's merits, and not an attempt to steer the court away from issues it would prefer not to address.
¶ 73 I respectfully concur.