The new paper found that practicing muslim judges were more lenient on a major religious holiday involving fasting. I wish the study looked at whether the results generalized to other religious holidays without fasting.

> The original also found a marked impact of hunger on acquittal rates.

Yes, but the opposite one, and it's been completely discredited. What this article calls out as a clever gotcha is in fact obvious to any subject matter expert: case assignment isn't random. People without lawyers were scheduled for right before lunch, and that's what correlated with worse outcomes.