A fun graph: a chess example of correlation vs causation
Today's example of why correlation does not equal causation comes to us from the comparison of move time to average centipawn loss (a measure of accuracy: lower is better) from my games. Notice that the more time I spend thinking about the moves, the less accurate I am. Of course, the conclusion isn't that I should blitz out moves to be more accurate, but that I think more in harder-to-evaluate positions. Just a fun chart if you're teaching an intro course in causal inference, statistics, or anything in empirical economics.
