While I am what he calls an "accidental moderate", there is an argument to be made for the intentional moderate stance. If you think the damage from your error

While I am what he calls an "accidental moderate", there is an argument to be made for the intentional moderate stance. If you think the damage from your error is proporational to the square (or some higher exponent) of the actual error, then an intentional moderate will have far fewer (perhaps no) fatal errors. They may choose 50 when they should have chosen 30, but they won't choose 100 when they should have chosen 10. If the damage from the one is like (50-30)^2=400, and the damage from the other is (100-10)^2=8100, then you can have a lot of modest, intentional moderate errors that don't add up to the damage of the one thing an accidental moderate is really, really wrong about. As usual, there is a relevant xkcd: https://xkcd.com/1170/ https://xkcd.com/1170/ Again, I myself am mostly an accidental moderate. But I think he's underselling the case for intentional moderates.