Shouldn't team performance decrease the season after a coach change since the players have to learn a new system? This seems to say that the cost of firing a poor coach is zero because at worst you will do the same as before in the short term, with a great potential upside for longer time spans.
These teams have large coaching staffs that don't get fired and change more slowly. I would guess that a coach's impact is often more long term (in training, recruitment of players and staff).
