Forgot your password?
typodupeerror
Math Stats The Almighty Buck

Mathematician Gives Tips On How To Win $1 Billion On NCAA Basketball 76

Posted by samzenpus
from the winning-the-pool dept.
Hugh Pickens DOT Com (2995471) writes "Jake Simpson reports at The Atlantic that Mathematician Tim Chartier, a Davidson College professor who specializes in ranking methods, teaches a math-heavy form of bracketology — the science of predicting the annual NCAA college basketball tournament at Davidson College in North Carolina. Chartier's academic research is in ranking methods where he looks at things like the page-ranking algorithms of Google. 'In 2009, my collaborator Amy Langville said: "You know what? ESPN has this huge online bracket tournament. Let's create brackets with our ranking methods, just to see if it's creating meaningful information."' Chartier's formula, an evolving code-based matrix that ranks each of the 68 tournament teams, has helped several Davidson students score in the 96th percentile (or higher) in ESPN's bracket challenge and this year, Chartier's goal is to help someone win the $1 billion prize offered by Warren Buffett to anyone who correctly predicts all 63 games of the men's tournament.

Chartier uses two methods. One is the Colley Method, named after astrophysicist Wesley Colley who developed a method used by the BCS for college football (PDF). His basketball method only counts wins and losses, not margin of victory. The other method is the Massey method created by sports statistician Kenneth Massey (PDF), which does integrate scores. Chartier has not been banned from any office pools — at least none that he knows of. But as a result of coming pretty darn close to filling out a perfect bracket just by crunching the numbers, brackets have become a labor of love. 'Now that the brackets are actually out, I've had students in and out of my office all week, sharing new ideas,' says Chartier. 'For me, that's more fun than filling out a bracket. They will all be filling out brackets, so it's like I'm doing parallel processing. I know what might work, but watching them figure out the odds, is a thrill.'"
This discussion has been archived. No new comments can be posted.

Mathematician Gives Tips On How To Win $1 Billion On NCAA Basketball

Comments Filter:
  • Model Worship (Score:5, Interesting)

    Not everything can be reduced to numbers, factored and condensed down to a single answer or list of probabilites. I'm a methematician and I'm here to tell you that a lot of what is presented as "mathematical" modelling in the modern world is little short of numerology and data massage.

    Eventually, if you go deeply enough into these kinds of models, you will forget that there is an actual game of basketball, being played by real human players. The instant that happens, you've become a numerologist and cargo-cult scientist. My opinion is that this is occuring in an increasingly large number of "clever geeks" now equipped with powerful computers and sophormic mathematics.

  • Re:Model Worship (Score:4, Interesting)

    by rmdingler (1955220) on Friday March 21, 2014 @09:03AM (#46542335)
    ...being played by real human players.

    You're right that the human element of any contest cannot be completely quantified. The outcome of every contest cannot be predicted by mathematics, but it is possible to identify trends, predispositions, and characteristics that generate more victories over an elongated number of Contests.

    It's more about using numbers to gain an edge, often for good value (another quantifiable datum) in sports, over a larger body of work. The 162 game baseball season has been adeptly exploited by sabremetric gurus like Bill James and Billy Bean. And Vegas sports books continue to make money because they do the math better than anyone else over the long haul.

  • Re:Model Worship (Score:5, Interesting)

    by Wescotte (732385) on Friday March 21, 2014 @09:04AM (#46542343)

    I would argue that professional baseball is ripe for statistical analysis because they play tons of games per season and players have very long careers. College basketball play very few game per year and the players are out after four years no matter what. You probably just don't have enough data to find anything meaningful before the team is comprised of completely new players rendering your data useless.

ASCII a stupid question, you get an EBCDIC answer.

Working...