A New Way to Measure Defense: Introducing the Defensive Efficiency Rating

Three games in under the new regime and the 2013 Chivas USA team bears very little resemblance to last year's model. In wanting to compare last year's squad with this year's squad, I encountered a conundrum: Robin Fraser's Chivas squad permanently used a four-man backfield while Chelis prefers a three man look paired on the flanks by two wingbacks. How could I fairly discriminate against two completely different systems?

The simple solution would to merely tally the goals allowed. This is hardly an ideal solution at all, however. A goal given up by a defensive blunder counts just as much as the goal scored on brilliant strike.

The simple solution also negates the value of the goalkeeper. In 2012 Chivas USA was simply dependent on Dan Kennedy and no MLS netminder faced more shots last year than the Chivas keeper. It seems unfair to reward the defense when its actually the man between the pipes who's the difference maker.

For another thing, defense is not, and should never be, specifically delegated to four or five players on team. Defense ought to be viewed holistically, not limited to only a group of 4-5 defenders.

In such, there is a void. So let's ask the question: What is it that good defenses do? Specifically, what is that good defenses do that we can track statistically? Good defenses disrupt their opponents' passing and limit their possession. A good defense prevents chances on goal. Finally, because the defense limits the opponent's possession and chances on goal, good defenses don't require a goalkeeper to make a lot saves.

Maligned as it is, I like how the quarterback rating equation takes four separate values (touchdowns, yards, attempts and interceptions) and gives you a meaningful number. So I took a page from the NFL and sum the following numbers to create a team's Defensive Efficiency Rating:

1. The other team's passing accuracy
2. A value I call "Possession with Purpose" multiplied by goals scored
3. Goalkeeper work rate i.e. is the goalkeeper cleaning up messes or bored stiff?

After each game I sum the numbers and average each team's score. And since we're rating defense, the lower the number, the better. I should also note that for my purposes I use only data posted on the MLSsoccer.com website. I'm sure Opta can serve these numbers on a platter in no time. I'm afraid I lack the time, resources, luck, patience and Brad Pitt-like stubble to search each match's chalkboards.

Three weeks into the season the median score has worked out slightly below 50, a nice round number but we'll see how this charts out over the course of a 34 game season. Perhaps this is good start or perhaps this is wild goose chase. Time will tell!

In a future post I might give a better idea of my methodology but in the meantime, posted below are the MLS Defensive Efficiency Ratings through Round 3 of the 2013 MLS season. Enjoy!

Some Observations

1. Some will be surprised to see the Galaxy tops in this list. First of all they've played two games while most of other teams have played three (note: Seattle has also played two games and they are 16th). But most notably they have limited shots on goal and their keeper, Carlo Cudicini, has only had to make one save.
2. If you notice a discrepancy between the number of goals conceded and their rating, it could very well be due to the heroics of a goalkeeper. Montreal, D.C. United, Real Salt Lake and, of course Chivas USA are tops amongst saves leaders.
3. Is the key factor bad defense or simply playing good offenses? Chivas USA has faced three of the top five squads ranked by shots on goal: FC Dallas (19, 1st), Columbus (16, 3rd) and the LA Galaxy (13, t-5). We might find out this weekend as the Chicago Fire (6) are behind only the New England Revolution with five shots on goal.

What do you think? Leave a comment below!