Right. It's one of the side-issues pointed out in Moneyball: it's always safer to 'go with tradition'. Contracts are incentive laden (you get money for saves, not for the team's record), and managers will get pummeled if they don't 'go by the book' even if the book is known to be false. My favorite dumb-ass version of this was a d.b. football announcer commenting on the wisdom of the "prevent defense" (which in this case, cost a game): "But still, it works more often than not." (well, yes: the team that is ahead late in the game wins more often than it loses; just as a baseball team up by a run or two in the 9th inning wins more often, granting its closer a S, than when the game is tied).