I was thrilled when my boss here at Stone Ward asked me to write a blog post on March Madness and get an early start on our contribution to the $1.7 billion in lost productivity created by the NCAA Basketball Tournament this year. Every March, sports fans are inundated with expert tournament bracket picks, simulations and odds. I wanted something that allowed for a little more customization, so I built an experimental bracket generator that allows users to create completed brackets based on factors of their choosing.


I settled on four simple factors that users can control (with the sliders at the top of the screenshot above) when generating their brackets:

Seeding – While the assigned #1-16 seeds don’t mean much in and of themselves, they’re a good representation of what the tournament selection committee thinks about a given team and expects from them in the coming weeks.

Travel Distance – Statistics show that teams who play their tournament games close to home perform significantly better than those that have to travel long distances (potential factors include increased travel time, time zone differences, and increased fan presence for de facto home teams). For the purposes of this site, distances are calculated based on latitude/longitude of home stadiums and tournament venues.

RPI – RPI (or Rating Percentage Index) is a metric used by the tournament selection committee that combines a team’s winning percentage with that of its opponents (and opponent’s opponents). I pulled RPI values from ESPN.com.

Upset Likelihood – Upsets and underdogs are a huge reason why March Madness is so popular. No bracket is complete without a few Cinderellas.

Click here to give it a try. You can quickly generate brackets with your choice of factors, and run it 100 times until your favorite team wins it all.