After the official 2013 March Madness bracket surfaced on Sunday, basketball fans and non-fans alike started selecting the teams they hope will be dancing their way to a championship. Actually making an accurate bracketology prediction is never easy, however, forcing us to look to statistician Nate Silver, ESPN’s Joe Lunardi or, most recently, a Harvard statistics teaching fellow to make March Madness picks.
Harvard Extension School published a blog post this week highlighting how not to fill out your NCAA Bracket. The writer interviewed Kevin Rader, who teaches the summer course Introduction to Quantitative Methods, to discuss what mistakes to avoid when sealing the fate of your simulated Sweet 16.
The first rule? Don’t let your emotions fill out your bracket.
“I have to admit, I let my emotions and rooting interest get the best of me,” Rader says. “I cannot see these teams as nameless entities. I have to cheer against (and pick against) Duke and the favorites, and root for the underdogs. It’s more fun the way.”
Emotions won’t win you a $50 prize, though -- well, unless you’re playing against the BostInno team.
There are going to be teams you love, and teams you love to hate, but just because your alma mater is say, Harvard, doesn’t mean they won’t get beat by New Mexico on Thursday night in Salt Lake City. Sometimes you just need to go against your passions and tune in to that voice echoing, “You won’t cash in on the Crimson,” inside your head. (Sorry, Harvard. I still love you.)
Also, stop hating so much on Duke. You know they’ll survive at least the first round.
Rader’s second rule? Learn what LRMC stands for: the logistic regression/Markov chain.
As Rader explains:
The logistic regression aspect of it is a model that uses key statistics from NCAA teams to try to predict the probability that each team will win a match-up. The Markov chain part of it is just the implementation of the model on any particular bracket that it is given to simulate who will win the tournament. It’s like choosing the winner of any game by flipping a weighted coin with probability p of the favorite winning that game (where p was estimated from the logistic regression model).
Basically, this just reiterates: don’t use your heart over your head.
Here are a few other mistakes to avoid, not courtesy of Rader.
Be Prepared for an Upset, But Don’t Pick Your Bracket Around It
The motto for BostInno’s March Madness Bracket Challenge is, “Your guess is as good as ours.” Honestly, anything can happen. “Anything” meaning an upset, or a pretty predictable four week span of basketball. Don’t make rash decisions because you think a No. 8 seed could upset a No. 1. That logic is just irrational.
Don’t Only Follow the Biggest Trends
Perhaps this is contrary advice, but while you don’t want to solely pick for an upset, you don’t want to solely pick for the big names either -- and that refers to both the teams and the gurus making the calls. Will Nate Silver’s predictions outscore mine? I’ll stay confident for now and repeat: “Anything can happen.”
Stop Overthinking It
If you’ve made it through this entire article, thank you. But, you’re definitely already overthinking that bracket.
