clock menu more-arrow no yes mobile

Filed under:

MLS Statistical Analysis, 3.15: Offense, Defense, and Goalkeepers

A look at the league's offense, defense, and goalkeeping numbers for the 2015 season.

Aaron Doster-USA TODAY Sports

Now that the 2015 MLS season has come to a close, so now is as good a time as ever to dive into Statistical Analysis. Here's the data:

Games Played / Goals / Shots / Shots on Goal / Wins-Draws-Losses / Points


Things are much easier to analyze when every team has played the same number of games. How about a shout out to the Chicago Fire, who reached the 20 loss plateau. From this info, we get these numbers, sorted by PPG:

Goals per Game / Shots per Game / % of Shots on Goal / % of Shots = Goal / % of Shots on Goal = Goal / Possession % / Pass Completion % / Points per Game


When I did this at the All-Star break, Colorado was the only team scoring less than 1 goal per game (0.90), and Seattle was the only team shooting less than 10.0 times per game (9.91). Both of those things remained true at the end of the year, though Colorado was able to improve their goalscoring to 0.97 GPG. Seattle, on the other hand, saw their SPG number drop to 9.74.

Three teams saw less than 10% of their shots result in goals -- Colorado (7.91%), Chicago (9.21%), and Portland (8.44%). One of these things is not like the others.

At a glance, DCU is going to rank at or probably just below average in most scoring-related categories, and we see for sure below, ranked by PPG.


Portland's low goal scoring, low shot accuracy, and low shot conversion is not a surprise. They were 18th in each of those categories at the All-Star break. What's surprising is they got worse in two of those three categories and improved their PPG from 1.45 at the break to 1.56 at season's end. At the same time, they increased their SPG from 13.73 (4th) to 14.29 (1st). This was not the reason for their MLS Cup win.

Down toward the bottom of the PPG table is NYCFC, despite their decent offensive statistical performance. This generally points toward bad defense or goalkeeping. Joining them toward the bottom of the table was Colorado, who were offensively offensive, with only their SPG sparing them a small number of blushes.

United were 10th or worse in all categories, with the worst being 18th (tied with RSL) in SPG. Apparently having Bill Hamid behind the defense and being average in shot accuracy and conversion is good enough to make the playoffs.

Now let's go on to the defense, starting with this data:


Three teams allowed 58 goals -- NYCFC, Chicago, and Toronto. Toronto was a playoff team, so it goes without saying that Sebastian Giovinco earned his MLS MVP award. Speaking of NYCFC, they were the only team to allow more than 500 shots (504).

From these numbers, we get these stats:


No team allowed 1.0 GPG or less, but Seattle and Vancouver were close with 1.06. That's a strong GPG for those defenses and goalkeepers Stefan Frei and David Ousted. Not pictured: Luis Robles.

Toronto's high shot number was compounded by the fact that 40% of their shots were on goal, the worst number in the league. Allowing high numbers of shots and shots on goal is not necessarily bad if your goalkeeping is good, but Toronto just released two of their goalkeepers, so you can guess what that means.

United's numbers, principally their GPG and SOG=G% would be slightly better if Bill Hamid had played a full season. He missed 9 matches, and while Andrew Dykstra was not a bad replacement, the team faced less shots per game and allowed less goals per game with Hamid behind the defense.

Let's get to the rankings.


Columbus stands out as a successful team, reaching MLS Cup without a great defense. Excepting PPG, they were no better than 11th in any category. Having an attacking four of Justin Meram, Ethan Findlay, Federico Higuain, and Kei Kamara really helps.

Portland's goalkeeper Adam Kwarasey is rarely mentioned without also mentioning his number of clean sheets. It's a notable statistic, but it's a bit easier to keep the other team out and keep clean sheets when a league low 31.44% of opposing shots are actually on goal, which masked Portland's de facto save percentage ranking of 9th.

Mired in the middle of the defensive table, but high in terms of shots resulting in goals and shots on goal resulting in goals were Seattle (1st in both), D.C. (2nd in both), and Vancouver (3rd in both). That reflects well on goalkeepers Stefan Frei, Bill Hamid, and David Ousted. Not pictured: Luis Robles.

The remainder of United's defensive rankings are not great. They allowed a lot of shots (19th), and didn't make many tackles (19th), while being average or below average in anything else. There's an explanation for this that will come with tomorrow's D.C.-specific breakdown.

Let's go on to goalkeeping, sorted by save percentage.


I almost always say here that save percentage is a difficult thing, because not all shots are the same. That said, look at the Red Bulls and Luis Robles, with a save percentage of 65.65%. Compare this with NYCFC, whose goalkeeping faced 43 more shots this season and posted a better save percentage.

Down at the bottom of the league is Philadelphia, who used four goalkeepers that played 5 games or more. Those four players combined to save 60.69% of the shots they faced, which is remarkable considering one of them (Andre Blake) saved 76.5% of the 34 shots he faced in 540 minutes.

At the top of the league you see Seattle, D.C., Vancouver, and San Jose, the only teams to register a save percentage over 70%. The DCU goalkeepers get a lot of credit for posting such a high save percentage despite facing almost 200 shots.

That's it for the league as a whole. Tomorrow we'll go through all of DCU's players in offense, defense, and goalkeeping.