Aka the "hungry judge effect": https://en.wikipedia.org/wiki/Hungry_judge_effect. But it's not like you can only fill your stomach up at a fancy restaurant (neither that you claim that nor that I find it being something bad).
FWIW, the hungry judge effect is almost certainly nonsense.
The original research assumed the case were randomly ordered, and so everyone should have an equal shot at parole—-but they weren’t! Instead, all the cases from one particular prison were heard together, with breaks usually occurring between prisons. Within a prison, cases were arranged by lawyer, with prisoners representing themselves going last. If people without professional representation fare worse (and they do), then…that’s the whole effect. PNAS published a “rebuttal” article where someone actually interviewed court staff who reported this, but it’s been cited like…70 times vs thousands for the original article.
There are other reasons to think the original result wasn’t true too. The “effect” didn’t occur when considering wall-time (i.e., judges were similarly severe at 9:30 and 11:30), only order (first vs last), but you’d expect hunger, blood sugar, etc to track time elapsed.
Sorry for the aside, but the fact that people still cite this drives me nuts.
