Thursday, April 12, 2018

Is a #16 Seed Beating a #1 Seed Shocking?

Last month, a #16 seed beat a #1 seed in the Men's NCAA tournament for the first time ever.  Note the word "Men's" however.  A #16 seed beat a #1 seed in the Women's NCAA tournament way back in 2008.  So let's be careful by tossing around the word "first". 

The NCAA expanded the men's tournament to 64 teams in 1985, thus creating 4 brackets with 16 teams each.  The first round of the tournament has 4 games in which a #16 team faces off against a #1 team.  Until this year, no men's #16 team had ever defeated a #1 team.  It took 134 games, given the timing of this year's games, for this upset to happen.  Everyone is super shocked, but should they be shocked?  In the regular season, a team ranked in the top 4 quite often plays a team ranked in the 60's.  Usually the top team wins, but not always.  Upsets happen every year and quite often every year. 

Nate Silver, over at fivethirtyeight.com wrote about this game and in the last paragraph of the article, linked below, calculated the odds of it taking this long for a #16 seed to win, at just 3.7%.  Not the odds of winning, but the odds of it taking this many games to occur.  In other words, Silver has calculated a 96.3% chance that a #16 seed should have won BEFORE this year.

How UMBC Did The Unthinkable — And The Inevitable

Statistics are fun and often go against common sense.  See the birthday matching problem I posted yesterday for another example of this.


No comments:

Post a Comment