Very easily. You wouldn't see a lot of data points, and you wouldn't be looking for it. You'd have on parole hearing maybe every few weeks. It'd be extremely hard to piece together that the time of day had a significant effect on the outcome, especially when you want credit for good results (and generally want to blame the judge for bad results). You probably don't have many lawyers in a position to observe enough parole hearings to really form a suspicion.