I enjoy sports and specifically I played soccer, fencing, karate and rugby throughout high school and college. I’m fond of watching hockey and local AA baseball here in Connecticut (soon to be known as the Hartford Yard Goats). Despite my familiarity and general appreciative take on organized sports I have never indulged in March madness bracket pools until this year. It could be because of my laser-focused interest in all things data science or because my co-workers were nice enough to ask me to join either way I’m in the mix and find myself jarring on about #14 upsets over #2 seats or more interestingly enough HOW I selected my bracket as a first time March Madness participant.
To quote my co-workers in reference to myself:
“I don’t know sports”
So I went to the best resource I knew at the time which was Nate Silver’s Five Thirty Eight blog. I’m a fan of their other non-sports related posts as well. They have a beautiful vizualization of the brackets and if you hover over each team you are presented with a statistical probability of performance in each round generated from a forecasting model that incorporates power rank, pre-season results, player injuries, geography, NCAA’ s-curve from available data. I went through and aligned my bracket based each teams probable chance of advancing in each round per region and who they would be playing in each round. In the end I’ve got Kentucky to win the entire pot!
I have not watched a basketball game in several years but as of yesterday at noon I was leading my office bracket pool at 12 points. I also think it’s worth mentioning my co-workers are mostly avid sports aficionados who made their brackets based on historical knowledge or gut feelings (or even mascots) which makes this victory over them even sweeter (or statistically satisfying). I’ll be sure to post my results at the end of the tournament. I chalk this success up to me being a big ole Data Nerd!