Win probabilities

March 14, 2014

I started working today on a new project. I’m trying to build a dataset that allows somebody to get the probability of a team winning a game, given the current score, time remaining, PP situation, and other variables. So far I’ve built the score and time remaining into my model, although hopefully in the next few days I’ll add PP/SH info.

I also am trying to find a nice looking way to present the results. Below is what I have so far. This is the win probability from the exciting game 6 of the Stanley Cup Finals last year, in which Chicago trailed by a goal with less than 2 minutes left and ended up winning the game in regulation.

I’d love anybody’s feedback (via email, Twitter, Facebook, wherever) on where you think I should go with this or how I can make the graphs look nicer.

In a crazy twist, it looks like Extra Skater has rolled out a similar metric today as well. I think mine looks fancier though.