Predicting the wins and losers of March Madness is such a daunting invite that it allures math nerds like Starfleet voyagers stringing up at Comic-Con. Statisticians, economists, Silicon Valley coders, the PhD quants at hedge funds and gambling syndicates: They’ve all tried to “solve” the outcome of the annual college basketball tournament’s 63 matchups.
“Every kid who takes a numerical modeling class and who’s a college basketball devotee, the first thing they want to do is prophesy the NCAA tournament, ” says Ken Pomeroy, a former meteorologist who has become arguably the foremost college basketball digits guru. His famous KenPom ratings evaluate the strength of all 351 NCAA Division 1 basketball teams consuming an old-school regression procedure known as “least squares, ” which psychoanalyzes statistical variabilities in teams’ past concerts and facilitates predict the winners in two-team matchups.
But to generate entire brackets is to tangle not just with the randomness of the game itself, but with the randomness of your speculation pool–the luck guesses made by all the people you’re rivalling against to predict the greatest number of champions. Microsoft investigates have loosed their machine-learning instrument Bing Predicts on March Madness projections, and various independent researchers, such as the chief data scientist of a big protection consultant, have used neural networks to entwine discrete predictive modelings into “ensembles” that spit up probabilities. But some of the most intense March Madness study is being done by David Hess. He’s a 36 -year-old with grades in neuroscience from Johns Hopkins and NYU who’s also from Kansas, and is thus “a huge college basketball fan.” In 2011 he went to work at a plays prediction site called Team Rankings, where he set out to build a tool to produce optimized NCAA tournament brackets for paying customers.