Posts

With the Sabres’ training camp starting today, I thought it would be a good idea to lay out some updates I hope to make on my live goalie evaluation project during this coming season.

Introduction Last night, Dean Evason and the Minnesota Wild raised some eyebrows by opting to start Marc-Andre Fleury over Filip Gustavsson, who turned in a 51-save performance en route to a double-overtime win at Dallas in Game 1.

In a previous post, I motivated and created a theoretical win probability model. I explored a few different ways to choose parameters that the model depends on, but one way that I haven’t written about (but have started using) is using goalie-specific priors.

Introduction If you haven’t read my previous posts motivating the idea of optimizing goalie decisions and developing a goalie-dependent win probability model because they were too mathematical – good news!

Introduction In the public sphere, many live win probability models already exist for NHL games. Brian Burke, now doing NFL analytics for ESPN, created a theoretical model that models NHL scoring as a Poisson process.

Introduction You just remembered that your favorite hockey team is playing tonight. You turn on the game during the first intermission to find that your team’s starting goalie has given up 3 goals on only 8 shots, putting them in a 3-0 hole after the first period.

This project was originally posted on Hockey Graphs. Introduction Hockey, by nature, is a fast-paced sport that can be difficult to represent by discrete situations. While most other professional sports can be viewed as combinations of distinct in-game events – at-bats in baseball, plays and series in football, and even possessions in basketball – hockey is extremely fluid and the game state is constantly changing.