Don’t Trust Teams That Have Outperformed Expectations

by William Ezekowitz

Preseason rankings. Irrelevant in professional sports, weirdly important in college basketball. I have shown in the past that they overvalue previous year’s tournament success, so they clearly aren’t perfect. But they also are as predictive as pre-tournament rankings in determining who will make the Final Four, so they must have some value. I decided to look into just how important preseason rankings are, specifically by looking at whether teams who outperform their preseason expectations regress in the Tournament.

To do this, I looked at all of the teams who were unranked in the preseason and were ranked pre-Tournament (i.e. teams who have played better than expected). In order to gauge how a team should do in the Tournament, I borrowed Neil Paine’s win expectation chart by seed, listed in this article. I then tested whether or not the teams (from the 2007 Tournament onwards) that fit my definition for outperforming expectations did better or worse relative to win expectation than everyone else. Here are the results:

Screen Shot 2016-03-15 at 1.04.11 AM

The tested group did in fact do worse than everyone else. More importantly, a two sample t-test confirmed that the difference between the two groups is significant at the 95% level, with a p-value of 0.003, well below the threshold of 0.05. So teams who are unranked in the preseason but ranked pre-Tournament do worse on average than all other teams. It should also be noted that a test of the converse group, i.e. ranked teams in the Preseason dropping to unranked pre-Tournament, produced no difference between win expectation by seed and actual wins.

For reference, there are seven teams this year who have gone from unranked in the Preseason to ranked pre-Tournament:

Screen Shot 2016-03-15 at 1.08.16 AMNote: Oregon, WVU and Texas A&M were 26th, 27th and 29th, but you’ve got to draw the line somewhere, and where we drew led to statistically significant results.

It should be noted, though, that this trend refers to teams in general and not in particular. This result does not mean that every team will do worse than its seed expectations, but rather, it suggests that all teams on average will fare worse. The 2011 UCONN Huskies notably bucked the trend when they went from unranked Preseason to National Champs. While these teams will almost certainly on average do worse than their 13.1 projected wins as a whole, that doesn’t mean that one or two of them won’t outperform expectations—statistically speaking, a couple probably will.

So when you’re filling out your bracket, take care not to have all of these teams advance too far. Odds are, this group as a whole will underperform.

About the author


View all posts

1 Comment

Leave a Reply

Your email address will not be published. Required fields are marked *