Intro to LacrosseReference
I thought it would be useful to lay out a bit of Lacrosse Stats 101 so that you can get up to speed or refresh yourself on some of the more foundational things we do around here stats-wise.
First of all, LaxElo is the model that I use to compare the strength of two teams. It’s based on a model that was originally created for chess, but it’s been used in lots of sports. Here’s the original write-up of how the model works. Unlike media polls, it’s forward looking. LaxElo isn’t about rewarding past performance, it’s about trying to estimate what a team will do.
The LaxElo ratings at the start of a game are the key input that underpins my Win Probabilities, which provide an estimate of how likely both teams are to win before or during a game. Every game with a live stats page will have a live win probability estimate published on the site in real-time. Second screens anyone?
But in order to create a win probability during a game, you have to be able to project how likely it is that Team A will score a goal in the next 60 seconds. How does it change if Team B commits a penalty? These questions required an estimate of the “value” of each type of play. The happy side effect is that those play values can be summed up for every player to provide an estimate of how much value they’ve created. It’s call Expected-Goals-Added as the good folks at USA Lacrosse Magazine allowed me to explain. (It also inspired the name of this very newsletter.)
The nice thing about having a value model for individual players is that it allows us to go beyond point totals and compare how much players in different roles are contributing to their team’s success. How does a finisher with a ton of assisted goals but few ground balls compare to a midfielder who may create a lot of their value through ground balls and assists? But it still leaves the fact that some players just have more opportunity than others. That’s why I created a metric called usage-adjusted-EGA that accounts for opportunity and gives us a way to quantify player efficiency, which is what really matters. How much value did you produce given the opportunity that you had?
This is a bit on the lighter side, but “Friends Don’t Let Friends Cite Goals/Game“. I’m not really the crusading type, but if I were, this would be my crusade. Goals-per-game is a terrible stat and we can do better. Call it out whenever you see it.
And if there is ever a stat or concept that you aren’t sure about, there’s a good chance that it’s got an entry in my glossary.
Ok, class dismissed!
Wait, not dismissed! Go sign up for the Expected Goals Daily newsletter. It’s free. The good, the bad, and the ugly about every team. No fluff.