According to statistician Nate Silver, those are the odds that he predicted the outcomes of this year’s March Madness games with perfect accuracy.
No fan has ever verifiably completed a perfect bracket. An average March Madness tournament has between 16 and 20 upsets.
In case you’re a little behind on your math education: one in one-point-six-billion is not good odds.
For a better chance of picking a winner, let’s look to another college sport with a March championship: the Pan-American Collegiate Chess Tournament.
Chess is a good choice for predictability. The first round of this year’s “Pan-Am” (featuring 45 teams) resulted in zero upsets.
What accounts for the difference?
First, chess is fundamentally more predictable. The outcome of the game results entirely from the decisions of the players.
One group from Stanford University created a system that can predict individual chess players’ wins and losses based on past wins and losses with greater than 85% accuracy. Ranking systems abound, some of which have been generalized to other competitions.
Second, the chess tournament structure itself minimizes randomness.
The Pan-Am is structured as a Swiss-system tournament, which leads to fewer upsets that March Madness’s single elimination system.
Single elimination makes for easy to understand brackets, and mathematically “efficient but unfair” outcomes (as compared to “fair but inefficient” league play).
What’s harder to measure is the influence of unpredictability on fan enjoyment.
So let us know in the comments: how much do you like being right?
