I've been thinking a lot lately about how to apply the principles of sabermetrics to soccer. For those who don't know, sabermetrics is a set of advanced statistics computed against a player's performance in baseball. Starting in the 1980's, a scattered group of dedicated baseball fans realized that traditional baseball statistics, like ERA and batting average are not good measures of a player's true performance on the field. So they came up with new statistical methods which gave a more accurate picture of how valuable a player is in a game.
One such example is On-Base Percentage plus Slugging (OPS). Traditionally, a player's offensive abilities is measured through batting average. However, a player's BA is measured as their hits divided by their at-bats. Hits are not the only way a player can get on base. Players can get on base by getting hit, or by getting walked. You may have a fearsome hitter who gets intentionally walked, resulting in a lower average, even though he's consistently making it to first base. (Insert your own crude joke here about balls, first base, and hitters.)
What happens when you start to measure players through sabermetrics is you find players who are undervalued by traditional thinking, but are highly valuable during the game. It was the application of sabermetrics that allowed the Oakland Athletics to win consecutive World Series championships with no-name players on a shoestring budget.
So with this high-level overview, you can start to see the appeal of sabermetrics, especially if they can be applied to another sport. It lets you see things in different way. You feel like Neo in the Matrix. You see the code underneath the game. Maybe a player that appears undersized on the field actually prevents twice as many goals against your team compared to the player who looks more robust. Maybe a player doesn't score ever but assists on every goal. She would be just as valuable up front during an attack as a player who makes the goals. The new insight lets you reasses your players and their abilities. As a geek and as a soccer coach, this appeals very deeply to me.
Obviously the key difficulty is whereas measuring a single player's performance in baseball is relatively easy (each player performs, in essence, individually), it's harder to do for true team sports like soccer.
But I have been putting some thought into how to do it, and I think I have some good equations to start with.
I will begin posting the equations here for discussion and refinement.
I hope you're willing to join me on this discussion.