I picked this up because I’d enjoyed reading Standard Deviations: Flawed Assumptions, Tortured Data, and Other Ways to Lie with Statistics by the same author last year. This one was lighter all round; more accessible probably but also focused on just one concept really.

It’s a couple of hundred pages of exploration of regression to the mean. An example he uses is when you praise one group of students for doing well, and shout at a group who do badly. When you come back the next day the praised group generally do worse, and the shouted at group do better. Does this mean that shouting is a more effective tool than praise? No, it’s just that generally people who turn in performances far from the average are really more average than their performance showed. If they do really well they will do worse next time whatever, and the poor performers will probably improve anyway. (Side note: so just be nice to everyone ok? Not mathematical insight, just human.)

Like his other book this was a bit heavy on the examples from American sports for me, but there were plenty of other topics covered. I didn’t need convincing of the theory behind regression to the mean but I enjoyed the read anyway and the surprisingly takeaway for me was how uplifting the idea that “it’s more average than you think” could be.