Analyzing Major League Soccer’s Castrol Index

Soccer, whether in MLS or other professional leagues, is a notoriously difficult sport to quantify through data analysis.  The other major team sports have highly-evolved analytics for comparing the performance of players, such as: baseball’s exhaustingly numerous Sabermetrics measurements, football’s Total Quarterback Rating and basketball’s Player Efficiency Rating.  The problem with creating similar tools for measuring soccer is that very little of what happens on the pitch in a given match is quantified with statistics beyond shots, goals, assists and saves.  So for instance, defensive midfielders or attacking wingbacks or creative forwards that can have a major impact on the outcome of a given match will very often not be listed in any way on a stat sheet.

Major League Soccer’s Castrol Index attempts to rectify this by analyzing an average of 1800 player movements per match.  You can read about all the geek behind it here, but in a nutshell the index does the following:

“The Castrol Index tracks every move on the field and assesses whether it has a positive or negative impact on a team’s ability to score or concede a goal.  A key factor for all areas of performance in the Castrol Index is which zone on the pitch the action takes place.”

And they are not only tracking things like successful passes and tackles, but also weighing these activities based on their proximity to either goal.  In theory, it should be ranking the best players in the league based on their comprehensive performance.  I decided to take the Castrol Index a step further and see if analyzing the player rankings on a team-by-team basis would be indicative of a team’s position in the standings.  In other words, does the information the Castrol Index is tracking for individual players translate to team wins?

Methodology

First of all, I threw all of the teams into one table and ranked them on a points-per-game (PPG) basis rather than overall number of points.  Teams have played anywhere from eight to thirteen games, so PPG is a more accurate comparative measurement of the results teams have been getting.  Then for each team I tracked the number of Castrol Index players in the overall league top ten, top 100 and top 200 to see if any noticeable patterns developed.  Finally, I took the average Castrol Index score of the top fifteen players on each team as well as the median.  I chose fifteen because it covers the first choice starters and regular subs for each team.

Results

Below is the resulting chart (colors indicate conference):

Analysis 

It very quickly becomes obvious that, for the most part, having more highly ranked players on the Castrol Index does translate to results on the field:

• Three of the top four teams in the PPG standings (Red Bulls, Earthquakes, Sporting KC) are the only teams that each have two players in the overall top ten.  The bottom nine teams have a combined two players in the overall top ten
• The top three teams are the only three with ten or more players each ranked in the overall top 100.  Every team in the bottom half has five or fewer players ranked in the overall top 100
• The top ten teams in the standings each average 450 or higher on the index with their top fifteen players.  Conversely, five of the bottom nine teams are below 450

The following charts are each team in MLS plotted first based on points per game and second on the average of their top fifteen players on the Castrol Index:

The charts don’t line up identically, but for the most part the teams near the top in average Castrol Index score for their top 15 players are near the top in points earned per game, and the same goes for the bottom.  There is a lot more movement in the middle where there is little difference in the average scores between teams.

Further Implications

• They may not have any top ten studs, but Seattle Sounders are the best team in the league right now and only San Jose comes close to matching their depth.  That depth should allow them to make yet another run at the US Open Cup and have a strong showing in CONCACAF Champions League
• It’s easier to see why Los Angeles Galaxy crashed out of CONCACAF Champions League and are near the bottom of the standings – they just don’t have enough players playing at a high level
• FC Dallas’ recent six-game winless streak is not hard to understand since they only have three top 100 players (Blas Perez, Ricardo Villar and Brek Shea) to start with.  Villar has been out for each match in the streak with an elbow injury and Perez and Shea have played the majority of a match together only twice during the streak.   If Shea and Perez can stop getting suspended, once Villar is back in a few weeks and if former league MVP David Ferreira finally makes his season debut as expected later this summer, Dallas will be a handful in the stretch run
• New York’s two top ten players (including overall #1 Thierry Henry) seem to be carrying the team at first glance, but they’ve managed to win three in a row since Henry’s hamstring injury.  A consistent ten through fifteen that doesn’t significantly decrease in Castrol Index score may be the reason why
• If I were a Sporting fan, I’d be very concerned about any injuries to their first-choice eleven and/or fatigue later in the season.  After Roger Espinoza, ranked eleventh on the team on the Castrol Index, their scores drop off of a cliff.  Their twelve through fifteen are by far the worst in the league
• There may be hope for the Timbers Army.  Portland is tied with Real Salt Lake for the third most top-200 players and have the ninth best top fifteen average.  They’re definitely underachievers in the standings based on their Castrol Index scores
• Toronto FC is absolutely terrible any way you slice it.  Aron Winter will not make it through the end of this season as coach

I’m not ready to create any Laws of the Castrol Index based on these results as there are some glaring exceptions to the correlation between having more highly ranked Castrol Index players and a teams’ points earned per game (most notably New York and Portland).  But the Castrol Index does overall give a good indication of the relative strength of a given team based on their individual components.

Ready to join the SportDFW Nation? Check us out on Facebook, Twitter or contact us at therealsportdfw@gmail.com