This article shows that the first autocorrelation of basketball shot results is a highly biased and inconsistent estimator of the first autocorrelation of the <italic>ex ante</italic> probabilities with which the shots are made. Shot result autocorrelation is close to zero even when shot probability...