The women's basketball team at the University of Connecticut (the UConn Huskies) won its 89th consecutive game last night. The victory, over 22nd-ranked Florida State, by a lopsided score of 93-62, broke the previous record held by the UCLA men's team, which won 88 games in a row back in the 1970s. UConn's record-tying 88th consecutive win, over 12th-ranked Ohio State, was by a score of 81-60. Last year UConn had its second straight undefeated season and, naturally, its second straight national championship. The Huskies won the championship game of their regional tournament by forty points and their national semifinal by twenty points before prevailing in the championship game by just six points over Stanford.
There's been a lot of talk about just why UConn is so good. No doubt they have a good coach, who is probably an even better recruiter, and their past success makes it easier to attract the best players. But I wonder, too, whether there isn't a kind of pointy-headed, mathematical explanation akin to Stephen Jay Gould's account of why no baseball batters hit .400 anymore.
It used to be fairly common for major-league baseball batters to hit for an average above .400, but no one has now done it since 1941, when Ted Williams averaged .406. It probably isn't the case that Williams, Hornsby, Cobb and others were better hitters than the best contemporary batters. But they were playing in a day when baseball was still a comparatively young game, the best strategies had yet to be devised (or discovered), and, consequently, there was more variation within the system of all batting averages. And more variation meant the outliers were "outier," farther from the mean, possibly even north of .400.
The UConn women's team, in this view of things, is not just a great team--it is a great team playing in the (relatively) early development of the women's game, when the distance between the best team and most teams is greater than it is in the men's game now and will be in the women's game sometime in the future. More variation within the system of women's collegiate basketball means that the very best team is less apt to suffer an upset, because the gap between the best and the typical is greater. A long winning streak is more likely to occur.
If you are skeptical, you probably won't be persuaded by the fact that the general theory would predict that UCLA's record, if ever surpassed, would be surpassed by a women's team. After all, there is around a fifty-fifty chance of that without introducing any other considerations. But what if the average margin of victory in all collegiate women's games was somewhat greater than the average margin of victory in all men's games? I think that would be strong evidence of greater variation within women's basketball.
I don't have the energy to track the results of all collegiate basketball games. If you'd like to, please tell me what you discover, as I'd be very interested. Meanwhile, I think I will track the score differential in all men's and women's games played this season in the Big Ten Conference. Then I will do the same for the NCAA tournament games. That should give some idea. I'll report in on what I find out.
POSTSCRIPT: For what it is worth, I've looked at the scores in last year's NCAA Division One Basketball Tournaments. In the women's games, the average margin of victory was 16.4 points (14.5 points in games not involving UConn). Whereas in the men's tournament, the average margin of victory was just 10.4 points. It's not a lot of data--just 63 outcomes in each tournament--but the women's games were on average considerably more one-sided, suggesting more variation among the entrants.
Comments