Sunday, March 17, 2013

Finding an Upset: Three Pointers

Every year in the NCAA tournament, inferior team beat superior teams. Norfolk State was not a better team than Missouri, they just scored more points over a 40 minute period. The cause behind these upsets is variability. It's no shocking revelation that three point shots increase variability. The three point shot has long been known as a great equalizer.

Underdogs need to increase variability by taking riskier strategies (i.e. shooting three pointers). If each team plays at their average, the favorite will win. Threes do NOT ensure an underdog will play better, but if you're going to go out you might as well go out swinging.

To determine which tournament matchups should have the most variability from a three point shooting perspective, I looked at the percentage of three point attempts taken and allowed for both teams (the favorite and the underdog). 3PA% is different from 3P%. I am not concerned with how often a team makes threes. In one single game that number will be difficult to predict. 3PA%, on the other hand, identifies just how often threes are attempted.

Key Takeaways

New Mexico and Florida each rely the most heavily on threes of all the favorites. However, they were fortunate enough to draw opponents that don't rely on them. Florida got one of the most unique teams in the country in Northwestern State. Which leads me to my next point...

Northwestern State is an interesting basketball team. Not only do they play at the fastest pace of any team in the country, but they also don't rely on the three at all. They don't take many and they don't let their opponents take many either.

Miami and Georgetown may be the "best" bets at becoming this year's Duke and Missouri. That would make Pacific and Florida Gulf Coast this year's Norfolk St. and Lehigh. I put best in quotes, because obviously all the games we are dealing with in this analysis are long shots. However, the numbers say that Miami and Georgetown are still relatively more likely to have an off night than others top teams.

Valpo and Davidson may want to consider changing up their defenses. Marquette and Michigan State rarely take threes on the offensive end. Coincidentally, Valpo and Davidson rarely allow threes on the defensive end. However, both Marquette and MSU shoot over 50% from two. If the two underdogs want a better shot at reaching the round of 32, changing their normally great defensive strategy might be a good idea.


  1. Interesting analysis. I agree with your premise off the top (more 3s = more variables = more upsets)... but do you have numbers for 1st round "upsets" in the past?

    Basically, is it really true that teams who shoot tons of 3s are upset more?

  2. Nick -

    I'm guessing someone has already looked at your question in further detail, but I just quickly looked up Missouri and Duke from last season.

    Missouri's 3PA% was 37.9% and Duke's 3PA% was 38.6%. Meaning they would be second and third on this year's list (only behind Florida). Obviously just looking at two examples is by no means definitive. Still, I am pretty confident that although top seeds who shoot tons of 3s aren't upset *often*, they are upset *more*.