There is a simmering debate online right now about the style of soccer played by Real Salt Lake and Sporting Kansas City.
It kicked off with this Grantland article, which, among other things, quoted Garth Lagerwey as saying:
“We play diametrically opposed styles of soccer. We want to keep the ball, we want to possess, we want to attack. Kansas City want to kick people … I think the league has to decide do they want that kind of physical, rock’em-sock’em style, or do they want to play more of a passing, possession, beautiful game. And I think that’s really why the rivalry comes into focus, because you see these two different directions the league could go.”
Greg Lalas – editor-in-chief of MLSsoccer.com, all-around good guy, and brother to the best hair / worst mustache in the history of soccer – responded with this video arguing that Lagerwey is incorrect.
I get what Greg is saying, but I think he is fundamentally wrong here. There are a couple of reasons why:
Playoff stats are inadequate – Greg relies heavily on them, but there simply isn’t a large enough sample size to control for the random variation that exists between matches. Season-long statistics are much better.
Take these playoffs. Sporting Kansas City has found itself chasing each series it has been in and facing opponents that were bunkering in, just trying to hold on to leads or draws. That’s a recipe for a really high possession percentage. It is also why raw possession percentages by themselves do not predict the winner of soccer matches very well. Most matches end with one team bunkering and the other attacking like crazy in an attempt to score. Inevitably the possession stat skews toward the attacking (usually losing) team for this last portion of the match, which impacts the total percentages that are often reported.
In contrast, Real Salt Lake dominated the Portland series from beginning to end, and actually played the Galaxy pretty evenly once Jason Kreis went back to the diamond midfield about 60 minutes into the first leg. That meant Salt Lake was the team trying to weather the storm at the end of both series, having the possession numbers skewed against it.
As a note: A better option to see the value and style of possession would be to calculate possession statistics while a match is tied and to check the minutes immediately preceding a goal to see how these numbers impact results.
Problems with the Possession Stat – All of what I just wrote is fine and good, but how exactly is the possession stat measured? Believe it or not, it isn’t just a raw percentage out of 90 minutes (+ stoppage). Actually, soccer statistics right now treat each pass (and therefore player possession) as an equal amount of time. So whether a player takes one touch of the ball or 10 before passing to a teammate, it counts equally in the possession metric. The statistical tools then take the number of events in a match and divides them down to take the possession stat.
Why does that matter? Sporting Kansas City spends a lot of time knocking its opponents to the ground. The time between when the whistle blows and when the free kick is taken do not count on the possession stat. Since Sporting’s general defensive philosophy is to cut down opponents as soon as they get into the attacking half (hence the fouls stat), this skews things a lot. Teams are inevitably going to play far more direct with a free kick because they can. That impacts opponents’ possession, and therefore Kansas City’s as well.
The Fouls Metric is King – The MLS average for fouls committed (and suffered) for the season was 422.26. That means, over a 34 match season, the average team commits and suffers 12.4 fouls per match. In contrast, Kansas City commits 15 fouls per match, and its opponents commit 13.2 fouls per match.
For reference, the next highest fouling team (Chivas USA) averaged 13.6 fouls per match.
Kansas City commits a lot of fouls, but they also suffered the third most in the league. If “Kansas City’s Opponent” was one team, it would be the fourth highest fouling team in the league. There are two reasons for that.
- People who get kicked tend to kick back when it keeps happening.
- Referees have a bias toward evening out the number of fouls called on each side.
But even considering that, Kansas City’s fouls committed to suffered disparity is absurdly high. The team stats page on MLSsoccer.com records that 15 of the 19 MLS squads committed and suffered almost the same number of fouls over the course of the season (within 25 total over a 34 match season). The four outliers were:
- Montreal Impact (110 more fouls suffered than committed – over 3 per game)
- FC Dallas (64 more fouls suffered than committed – nearly 2 per game)
- Sporting Kansas City (59 more fouls committed than suffered – nearly 2 per game)
- Chivas USA (63 more fouls committed than suffered – nearly 2 per game)
Chivas has that fouls committed stat because half of its team doesn’t belong in MLS, and the squad had nothing in the way of defensive tactics for the first half of the season. (Thank you Chelis!)
So that leaves Sporting as the only legitimate MLS side that committed substantially more fouls than it suffered.
What about Real Salt Lake? Jason Kreis’ men are below average in both fouling (400) and fouls suffered (421). That means the average Kansas City match would see 28 stoppages for fouls, while the average Salt Lake match would see 24. That is enough to change a game.
Your Eyes Don’t Deceive You
The eyeball test, which is what most media use in reporting that Kansas City and Real Salt Lake play different styles of soccer, is 100 percent correct. The numbers back it up.
Real Salt Lake plays a very aesthetically pleasing style of soccer, while Kansas City seeks to muck up the pitch with foul after foul. If this MLS Cup really is a referendum on which style of soccer ought to dominate MLS, then you can bet the national television partners will be cheering for Jason Kreis’ men to win the day.
Pretty much every neutral watching this match will probably feel the same way.