In McGinn v. State, 961 S.W.2d 161, 169 (Tex. Crim. App.) we stated that "future dangerousness is, in essence, an issue of prediction," as opposed to "an issue of historical fact." 961 S.W.2d at 168. We explained: "Findings of historical fact are either right or wrong at the time of trial. But, predictions are not right or wrong at the time of trial--they may be shown as accurate or inaccurate only by subsequent events." Id. We held that a Clewis review of the future dangerousness issue is impossible because it would require us to assign some evidence mitigating value and to substitute our judgment for that of the jury. Id. We concluded that the Jackson standard should instead be used because it views the evidence in the light that supports the jury's verdict and asks only whether circumstances are present that a rational person somewhere could find a probability of future dangerousness beyond a reasonable doubt. Id. In McGinn, only one of the cases cited in Clewis demonstrated that a factual sufficiency review may have been applied to a punishment issue. 961 S.W.2d at 167. That case involved the sufficiency of the evidence to sustain a jury finding of "malice" in a murder case in which the defendant was given the death penalty. Villareal v. State, 140 Tex. Crim. 675, 146 S.W.2d 406 (Tex. Crim. App. 1940), overruled on other grounds, Mays v. State, 563 S.W.2d 260, 264 n. 5 (Tex. Crim. App. 1978). We distinguished McGinn from Villareal by stating that "malice" was an issue of historical fact, but future dangerousness is an issue of prediction. McGinn, 961 S.W.2d at 168.