I know that kind of remark is considered just funny, but this is exactly the trend. "Way back" when Ken Pomerance (sp?) began his types of analyses to finely break down individual play, plus subtle winning actions, plus "tendencies", this computer analysis has been creeping further and further into even top level sports (esp. see baseball, but also basketball and football) and the computer statistical advantage is replacing more and more "coaching decisions."
I wonder, if we imagine a "stack" of coaches from pee-wee ball through NFL, how far up that stack we could simply void out the coaching in-game decisions by computer already. To put it another way: how many coaches are already irrelevant as to in-game play-call decisions? ... or at least POORER choices than the best computers with available data. I wouldn't be surprised if that "irrelevant" number reached far up even to the collegiate regions. [[ remember I'm talking about only play calling in game.]]
As an irrelevant aside: I DON'T like the computer take-over in any sport. I want to see how PEOPLE do, not machines. That includes coaches and human strategists and the occasional true human intuitive geniuses.