Quote:
Originally Posted by madJ
Because real life MLB managers wouldn't do that, just reiterates their overvaluation of the closer role and their willingness to lose the game because the closer would be pissed if taken out of a close game when he just sucks.
Again, nothing wrong with the AI decision. The problem in this case are the decisions by real MLB managers.
|
The problem with the AI decision is it is dumb. It is based on an extremely small sample size. Why take out your best reliever because of one bad plate appearance? That is just short sighted thinking. Would you remove Mariano Rivera in his prime if he gives up a leadoff home run? I mean shouldn't the AI understand better than humans the idea of variance? Removing your best pitcher because of one bad plate appearance in a close game makes no logical sense when evaluating from a purely statistical standpoint or when evaluating from the standpoint of this being a sim which should simulate the actions of real life managers.