WAS THE RED SOX COLLAPSE A ONE IN 278 MILLION EVENT? The 278 million figure is fun to think about, and the events were dramatic. One home run by potentially the last batter in one of the games was by a hitter with a .108 batting average for the year. An unlikely event. The Red Sox won only 7 of their last 27 games, and that seems unusual for a team that had won over 60% of its games going into September. But there are arguments that those 27 games were not statistically independent. To begin with, there are the psychological issues. I posted here that I was reconsidering my belief that sportswriters have exaggerated the importance of streaks because people don’t realize that random number sequences will generally include some long streaks. It seems to me that there are teams that “catch fire” after they get a lead in the standings, and win a high percentage of their games the rest of the way—the Detroit Tigers this year or the 1969 Mets. The games during these streaks are not independent events because each success makes these teams more likely to win the next game. If I believe that psychological factors can lead to hot streaks, I have to accept that they can also lead to losing streaks. Another explanation is that the Red Sox pitching staff wore out. It’s a long year, and pitchers wear down. One of the other famous collapses, by the 1964 Phillies, came when their pitching also wore down. If the Red Sox pitching was weakened in September, it’s a mistake to use their high success rate in the early months to estimate their probability of winning a game in September. Further, if their relief pitchers were heavily used in one game, the probability of losing the next couple games would increase with the weakening of the bullpen—another reason why the games in September were not independent events.

This entry was posted in Baseball, Economics, Science, Sports. Bookmark the permalink.

Leave a Reply

Your email address will not be published.