Quote:
Originally Posted by Curve Ball Dave
That's thinking like a pitcher, not a statistician. Your job is to win the game, not make a sabermatrician happy. Watch the smart pitchers and they do just what you described. They give up a lot of meaningless solo home runs (statistically that's bad) but they give them up early in the count so they don't throw a lot of pitches and expend energy on one hitter who can't cost them the game no matter what he does. When all is said and done the smart pitcher's team wins, and the smart pitcher makes some sabermatrician unhappy.
|
FWIW, some sabermetricians are now starting to try to measure just what you described by using a concept of
Win Probability Added. It's supposed to determine how much a particular play is likely to impact the outcome of a game. So Cabrera's solo home run that makes the game 10-2 wouldn't negatively impact the pitcher's WPA nearly as much as a solo home run would if the pitcher's team had just 3 runs instead of 10. The WPA for a particular play is higher when the score is close, and higher the later in the game it is, so the highest WPA plays are late in tight ballgames.