It does this by comparing the prediction glitches of the two styles above a specific period. The take a look at checks the null hypothesis the two models provide the exact same performance on regular, versus the alternative that they do not. Should the check statistic exceeds a significant price, we reject the null speculation, indicating that the