Following on from my previous post on the relationship between Major League Soccer payrolls and on-field performance, I consider how efficient league teams have been in the 2011 regular season. I seek to answer the question, What is an MLS team's cost per point relative to the league average?
In order to answer this question, I perform a quantitative performance benchmark analysis modeled after Bill Gerrard's study of the Oakland A's in his Moneyball paper. Performance benchmarking is the practice of assessing an organization's performance relative to some defined baseline. (It doesn't have to be an organization; such analyses are common in manufacturing and the computer software and hardware industries.) The simplest approach is to use a ratio analysis that calculates the amount of output per unit of input, or amount of input per unit of output, whatever quantity makes more sense to researcher and practitioner. Ratio analysis has the advantage — and disadvantage — of being simple: it is transparent and simple to understand, but breaks down for organizations and systems with multiple inputs and outputs. Moreover, I've already demonstrated that payroll is weakly connected to on-field performance in MLS. However, its simplicity makes ratio analysis a good starting point in performance benchmarking analysis.
I calculated team payrolls using player base salary data from the MLS Players Union website, and I used total points won over the 2011 MLS season. I set the payroll-performance baseline to $2 million and 25 points. Then the marginal payroll cost-per-point ratio is the difference between the total payroll and its baseline, divided by the difference between league points and the baseline total. By comparing the marginal cost to the league average cost per point, one obtains a measure of a team's payroll-performance efficiency gain. Let's call it "front office efficiency."
Below I present the marginal payroll costs of MLS teams in the 2011 season, ranked by front office efficiency gain. The league average cost per point is $141,578.26, so marginal costs below that amount are considered positive and costs above it are negative.
|Seattle Sounders FC||$3,033,191.00||63||$27,189.24||80.80%|
|Real Salt Lake||$2,946,031.00||53||$33,786.82||76.14%|
|Sporting Kansas City||$3,109,530.00||51||$42,674.23||69.86%|
|San Jose Earthquakes||$2,600,712.00||38||$46,208.62||67.36%|
|New England Revolution||$2,698,503.00||28||$232,834.33||-64.46%|
|Los Angeles Galaxy||$12,830,222.00||67||$257,862.43||-82.13%|
|New York Red Bulls||$12,583,039.00||46||$503,954.24||-255.95%|
|Vancouver Whitecaps FC||$3,566,997.00||28||$522,332.33||-268.94%|
I've looked at this chart a few times, and one of the things that strikes me is how few of the MLS clubs are grossly inefficient. In Gerrard's paper, roughly half of the Major League Baseball teams in the study spent more per point than the league average. The large payrolls of LA and New York, and to a lesser extent Toronto and Vancouver, contribute to skewed average. If one uses a median payroll cost, the efficiency gains change dramatically but the team orderings remain the same.
As one would expect, the Galaxy and the Red Bulls — the teams with the largest payrolls in the league — are among the most inefficient sides in the league according to the ratio analysis method. However, this approach is biased against big-spending teams to the extent that even if such a team wins the league they are still described as inefficient. Toronto and Vancouver were also hugely inefficient, but they were also among the worst teams during the regular season so perhaps there is a nugget to truth to the results. Fifth from bottom is the New England Revolution with an efficiency loss of 65%, who also had a very poor season.
Beyond the big-spending Galaxy and Red Bulls, front-office efficiency gain does appear to track well with on-field performance. The two major exceptions were the Chicago Fire and the Portland Timbers. Both missed the playoffs yet had marginal efficiency gains at the top of the league, which indicate that they extracted the most out of a somewhat limited squad. Chicago were down for most of the season and recovered late, but Portland were at the edges of the playoffs for a long time and were eliminated in the penultimate match. A lot of people in the Pacific Northwest will make some noise if I don't point out that Seattle Sounders were #2 on the front-office efficiency list.
Now, if you regenerate the table relative to median payroll cost-per-point, a different picture emerges. The median cost-per-point for the league is $44,599.30, with marginal costs below that amount expressing an efficiency gain.
|Seattle Sounders FC||$3,033,191.00||63||$27,189.24||39.04%|
|Real Salt Lake||$2,946,031.00||53||$33,786.82||24.24%|
|Sporting Kansas City||$3,109,530.00||51||$42,674.23||4.32%|
|San Jose Earthquakes||$2,600,712.00||38||$46,208.62||-3.61%|
|New England Revolution||$2,698,503.00||28||$232,834.33||-422.06%|
|Los Angeles Galaxy||$12,830,222.00||67||$257,862.43||-478.18%|
|New York Red Bulls||$12,583,039.00||46||$503,954.24||-1029.96%|
|Vancouver Whitecaps FC||$3,566,997.00||28||$522,332.33||-1071.17%|
As I said before, the order of the efficiency rankings relative to the league median does not change, but there is an even number of teams that are efficient and inefficient and the percentage magnitudes are very different. Chivas' inefficiency becomes more apparent, as do — in a very big way! — that of the New York Red Bulls and Vancouver Whitecaps. At the other end of the table, it becomes very clear just how much the Timbers got out of every point earned in MLS, despite not making the playoffs.
Now, payroll is but one variable that determines the performance of a sports team, and in Major League Soccer, it's not all that definitive. However, this analysis serves as a starting point for multi-year, and multi-variable, analyses.
CORRECTION: I miswrote "cost-per-win" when I meant to write "cost-per-point" in several places in the post. I've corrected the text above. Thanks to a couple of people on my Twitter feed who pointed that out.