Concerns about hindsight in the law typically arise with regard to the bias that outcome knowledge can produce. But a more difficult problem than the clear view that hindsight appears to provide is the blind spot that it actually has. Because of the conventional wisdom about error review, there is a missed opportunity to ensure meaningful scrutiny. Beyond the confirmation biases that make convictions seem inevitable lies the question whether courts can see what they are meant to assess when they do look closely for error. Standards that require a retrospective showing of materiality, prejudice, or harm turn on what a judge imagines would have happened at trial under different circumstances. The interactive nature of the fact-finding process, however, means that the effect of error can rarely be assessed with confidence. Moreover, changing paradigms in criminal procedure scholarship make accuracy and error correction newly paramount. The empirical evidence of known innocents found guilty in the criminal justice system is mounting, and many of those wrongful convictions endured because errors were reviewed under hindsight standards. New insights about the cognitive psychology of decision-making, taken together with this heightened awareness of error, suggest that it is time to reevaluate some thresholds for reversal. The problem of hindsight blindness is particularly evident in the rules concerning the discovery of exculpatory evidence, the adequacy of defense counsel, and the harmfulness of erroneous rulings at trial. The standards applied in each of those contexts share a common flaw: a barrier between the mechanism for evaluation and the source of error. This essay concludes that reviewing courts should consider the trial that actually occurred rather than what “might have been” in a different proceeding and proposes some new vocabulary for weighing error.