A New Look at Player Value

meolaCommissioner Garber spoke on Wednesday and much of his talk focused on some big news in the world of soccer business.  We will have much more detail on some of the salient points from his address via Google+.  However, with the new season set to being, we once again offer the floor to Dave Laidig with some thoughts on player value.

Inspired by metrics created by baseball’s quantitative gurus, I have attempted to apply the concepts of the Wins Above Replacement (WAR) statistic to soccer performance.  In short, this statistic compares a player to a hypothetical replacement; valued in contribution to wins.  Thus, the number should represent a player’s value (in wins) above a stereotypical fill-in player.  And when applied to soccer, this concept becomes the Points Above Replacement (PAR).  Simple on its face, quantifying the parts are difficult.

Using the Castrol Index – previously demonstrated to be related to season results – we can show how much the performance ratings reflect winning.  I don’t wish to cover material again, especially since much of the data has been previously posted on Footiebusiness. (See a recap in the background on the PAR).  However, with metrics that correlate to winning, we can assess player value.  The PAR may be useful because it blends playing time and performance level (i.e., a stellar game will not have a better PAR than a solid player over a whole season).  Ideally, the PAR will spur unique perspective on analyzing player performance.  The PAR cannot replace good scouting and traditional evaluation methods, but may be included in a broader analysis of players

On the eve of the MIT Sloan Sport Analytics Conference, there is a growing effort to apply advance techniques to understanding the game.  And from this knowledge, teams can gain a competitive advantage or at least allocate their resources more efficiently.  With the MLS season starting again, it is worth viewing the season from a different perspective.

Advertisements

Remebering Sandy Hook: Soccer Auction

quinnAuction link here.

Soon after the tragedy in Newtown, Dave Clarke, the head coach of the Qunnipiac University Women’s soccer reached out to discuss some of the initiatives the school was working on in the aftermath of the terror at Sandy Hook Elementary.  One of these was the extremely powerful Soccer Night in Newtown.

Coach Clarke has again reached out to discuss a fantastic soccer themed fund raiser.

Starting on March 14, 2013, Quinnipiac Soccer will be auctioning off #26 shirts donated by soccer teams from around the world to raise money for a scholarship fund in memory of Rachel Marie D’Avino. Rachel was one of the 26 victims of the tragic shooting. Rachel’s cousin, Lauren Carmody-Grenier was team captain in 2002 while the 2011 captain, Kyla Miles attended Sandy Hook and knew many of the victims.

The proceeds of the auction will be used to establish a scholarship fund in memory of the 20 students and six teachers who died at Sandy Hook and will be dedicated in honor of Rachel.  Bidding will open in March 14, but for now, you can peruse some of the available items here.

I want to take this opportunity to commend those involved in this event for their good acts and offer my hope that the night brings some joy to those in that community.  The business of soccer includes the charitable good works of its teams and players, and these events are a shining example of good that athletes and professional sports can provide in the community.

Selling Tickets in MLS

revs jerseyEvery few weeks during the season we take a look at the promotional efforts used by Major League Soccer teams to sell tickets.  Over the four plus years we have been writing this blog, the number of teams that rely upon promotions have declined, with teams like Portland, KC and Seattle filing their stadiums without running such promotions.  However, some teams still use such devices to entice fans and fill stadiums.

The Revs are running a promotion for their home opener by offering four and six game “First Kick” packages.  Starting at just $86, the packages include a special Revs t-shirts and allows fans to select a series of other games during the course of the season. Package holders can also purchase discounted tickets to other games and receive preferential treatment for other games.  For more on this promotion, click here. One thing sure to bother Revs fans is that the link to the sale connects to Patriots.com.

The Rapids are offering a five game plans through their website that also gives fans a choice of games.  Purchasers also receive a limited edition Rapids scarf.  The order page is somewhat confusing and price points aren’t easily determined.  However, the seats range from $20 per game to $36 and fans can seemingly mix and match their sections.

Finally, Chivas USA is offering 3 and 5 game discount mini-plans. Fans can receive a substantial discount off of full ticket prices.  Tickets start at just $45 for the 3 game   packs and $60 for the 5 game packages.  The team also offers four ticket family packs.  For more, click here.

Monday After

san-joseTime for the last Monday After of the off season.  As regular readers know, we tend to use Mondays to look at a few big business stories from American soccer.  During the season, that look includes a recount of attendance figures and an examination of the national broadcasts from the last round of games.  For now, I hope everyone enjoyed the last weekend without MLS and here are some business stories to enjoy.

It is no secret that MLS is wed to the idea of expanding into New York City for its 20th franchise.  Yet over the last few months, Orlando has been the city generating the most ink around the possibility of moving into the top level of American soccer.  Over the weekend, the Orlando Sentinel provides a detailed feature on Flavio Augusto da Silva, a Brazilian businessman who has dumped tens of millions of his own money into the team. One interesting business not from the feature comes courtesy of Orlando City owner John Bonner, who indicated that the team had choices when looking at potential big money investors into the franchise.

On Sunday the league announced that Commissioner Garber will participate in a “March to Soccer” via the league’s YouTube channel and via Google+ hangout.  The event will be hosted by Google’s New York office.   This event will replace the annual teleconference with the media and will be a first for a league commissioner.  The devil is in the details, but it will be interesting to see if this replaces the usual phone interactions on a permanent basis.

One final note. After much trial and tribulation, construction is set to begin on the Earthquakes’ new stadium in San Jose.  This week the building process will take a big leap forward with the team on track for a 2014 opening.  For more on the stadium financing and permitting process, click here.

From the Vault: Money & Performance in MLS

Dave Laidig is back with the third part of his series looking at the use of statistics and numbers in soccer.  For Part I, click here. For Part II, click here.  So far, this series has focused on analyzing objective measures of
performance.  In Part 1, we covered the Castrol Index and an adjusted
index that allows meaningful comparisons on overall contribution to
wins between positions.  In Part 2, we used this information to
determine the potential impact of field players from different
positions on wins.  However, in the business of soccer, resources are
limited.  And one must get the maximum value for their investments in
players.  Here, we review some of the financial aspects of obtaining
the performance levels discussed in earlier installments of this
series.

The theme for this part is return on investment: knowing what can be
obtained for a given price.  And to get this analysis, I made some
assumptions.  For example, I used the 2011 “guaranteed salary”
reported by the MLS Players’ Union, instead of base salaries.  I
believe teams likely know which contract incentives will probably be
met.  Thus, I treat the salaries guaranteed as of Sept. 2011 as
expected by teams, and use them for my salary analyses.  Also, I treat
all guaranteed salaries above the designated player (DP) threshold as
DPs.  I understand there is room for teams to buy-down salary cap
values using allocation money.  But sticking to my dollars and cents
theme, I classify DPs based on actual expenditures, and not salary cap
rules.

With the background aside, we turn to the role of money on
performance.  First, player salaries are poor predictors of team
success.  Between 2007 and 2010, total team expenditures were not
significantly correlated to league points (.193).  Considering
individual players, guaranteed salary was not correlated with Castrol
Index scores or adjusted index ratings (.166 and .172 respectively).
And if we ran the equations in Part 2, with salary replacing the
performance indices as a predictor of team points, there is no
significant relationship to league points and the model R-squared was
a paltry .27 (compared to .78 of a maximum possible 1.0 for the
adjusted index weighted by playing time).  Further, I created an
effective average salary for each team (avg. salary weighted by
minutes), and that was not significantly related to points either.

These results inform us that more money does not lead to more wins in
MLS.  In contrast, in the 2010 EPL season, team salary costs were
highly correlated (.85) with league points.  As a rough indicator of
the value of large salaries, consider whether Designated Players (2011
salaries above $335,000) are more likely to be in the top 20% of
performers.  There were 31 non-goalie DP salaries in 2011, and 11 of
these were in the top 20% of their position group.  This is a
statistically significant result (Chi-square = 4.75, df 1), but the
size of the effect is modest in comparison to the wages.  A randomly
selected DP has about a 35% chance of being a top performer, while the
rest of MLS players have a 19% chance of being a top performer.  And
of course, one could sign several other players for the typical DP
salary.  In MLS, one can obtain high quality player performance
without spending more than opponents.  In short, there is room for
more efficient player spending.

But knowing there is room for improvement and actually improving are
two different things.  A standard is needed to measure the value of
performance, and not just for DPs.  As an initial step, the average
salary per adjusted index point is $31,228 for forwards, $23,425 for
midfielders, $17,849 for defenders, and $18,861 for goalkeepers.  The
median salary per point is about $12,000 for the field positions;
which is interesting even though the average performance index and the
wages differ for each position.  Also, the range of dollars per index
point is very wide.  Indeed, the field players with the greatest value
(typically key starters on a minimum salary) are in the $4,200 per
point range; while the egregious examples can be over 600k or 800k per
point.  And with such incredible variability, I use the median values
as the basis for calculating value.

In addition, I chose to examine a subset of the top players as well.
Some economists have suggested that performance at the top-end is
disproportionately rewarded; possibly due to the all-or-nothing nature
of sports.  Thus, considering the top 20% of performers at each
position, we find their average salary per adjusted index point for
forwards are $84,175 (median $ 13,243), midfielders $40,689 (median $
14,880), defenders $17,339 (median $17,954), and goalkeepers $10,197
(median $ 7,728).  When compared to the entire position groups, it
becomes evident that purchasing higher end talent is slightly more
expensive.

These data points are involved in creating wage standards for
performance levels.  For example, the median wages per point
multiplied by the median points creates an “efficient salary” for a
50th percentile player.  With the math, an efficient salary for a
mid-level forward would be $87,800 (7.48 * $11,738), a midfielder
would be $91,318 (7.44 * $12,274), and a defender would be $91,116
(7.71 * $11,818).  These “efficient salaries” are slightly below the
median position salaries reported in Part 1.

Similarly, we can calculate an “efficient salary” for a top player
(80th percentile) using the median wages per point for the top 20% of
players.  An efficient salary for a top-level forward would be
$107,665 (8.13 * 13,243), a midfielder would be $117,254 (7.88 *
14,880), and a defender would be $146,863 (8.18 *17,954).  Using these
salaries, one can start to analysis the value of a player contract.
These standards represent what a performance increase alone would
justify, based on the current MLS market.  Any expenditure beyond
these levels would require an additional justification.

And soccer is a business.  Any salary or wage must be justified; but
increased performance is only one justification for a DP.  Obviously,
anything else a player contributes to increased revenue would support
extra wages (beyond that supported by performance).  And because fans
will buy Donovan jerseys over Franklin jerseys, the Galaxy are
justified in paying Donovan more, even if both contribute the same to
wins.  Further, there are less tangible benefits as well.  A DP may
attract better competition for friendlies, or lead to more TV
exposure.  And other players may accept less for the chance to play
alongside a star player.  All of which may affect the bottom line.
Consequently, a DP decision process should consider the value
justified by performance (rough estimate of performance * $ per point)
and projected revenue (additional jerseys and tickets) and as well as
the more speculative benefits.

And by using a quantitative method to account for the various buckets
of player value: teams may be able to make better business decisions
by recognizing where their purchase price is going (performance,
merchandising, or improving other players’ performance) and then
evaluate the success or failure of the results.  Over time, one can
quantify the risk involved for each category and improve the market
efficiency of player acquisition.

Columbus Business Bits

columbusWith the season is continuing to creep up on us, we thought this would be a good time for a team specific soccer business bits.  The featured team tonight is the Columbus Crew, the home team in Major League’s Soccer’s first soccer specific stadium.   The Crew have just announced two new corporate partnerships with well known food franchises White Castle and Papa Johns.   Both are Ohio based companies and both will now be featured at Crew Stadium.   According to the official press release, “Papa John’s becomes the official pizza of the Columbus Crew and Crew Stadium, and will be sold in its own permanent in-stadium concession stand – located in the southeast corner of the main concourse – and in-stadium concession sales at southwest & northeast stadium stands. White Castle hamburgers will be offered on the Crew Stadium plaza as well as in the southwest and northeast concession stands.”  Both brands will activate around the Crew with a number of promotions and events.

From the dinner plate to broadcast television, where the Crew have issued a press release about their 2013 broadcast schedule.  Fox Sports Ohio will televise 32 Crew matches during the 2013 regular season.  The rest will be part of national telecasts. Per the presser, “610 AM WTVN returns as the club’s English-language radio partner in 2013, with the network carrying the club’s complete MLS regular season schedule. Neil Sika resumes his play-by-play role alongside analyst and Ohio State Men’s Head Soccer Coach John Bluem.  103.1 FM La Mega, the club’s Spanish-language radio partner, will also carry the club’s complete 2013 slate of matches. Carlos Cordova calls the games with Benny Pietrangelo serving as the network’s analyst. Juan Valladares is set to provide additional commentary from the sideline.”

 

Laidig Speaks: Adjusted Castrol Index and Creating a Predictive Framework

Footiebusiness Contributor Dave Laidig weighs in with his latest manipulation of Castol data.  Dave’s statistical work is ground breaking stuff and represents the cutting edge of soccer analytics and stats based crunching.   Read closely and drop Dave a line.  Also, check Dave out at Par Stat, his innovative website looking at Points Above Replacement in soccer.

Castrol Index scores are represented as performance measures by Opta, the publisher of the data.  Past analyses have shown that Castrol Index scores, positions, and playing time reflect league results (See Footiebusiness and A Beautiful Numbers Game).  And while the scores appear to match up with the league results pretty well of over time (R squared in excess of .73 for each of the three seasons analyzed), the year to year consistency of the data has not been established.  Consequently, this analysis addresses the year to year relationships between adjusted Castrol scores (Castrol scores adjusted to remove the “punishment” for lesser playing time).

stats

The first step in comparing the 2011 and 2012 MLS Castrol Index is to convert the scores to a standardized scoring system. The Castrol Index changed scoring systems between 2011 and 2012; a 3.4 – 10 point scale became a 0 – 1000+ point scale.  Similarly, although the adjusted scores narrowed the ranges, the adjusted Castrol scores (the key component of this analysis) were also different.  Thus, the 2011 and 2012 scores are not directly comparable. Consequently, this analysis converts the reported scores to Z-scores (by position), which then allows for year to year comparisons.

Next, several year over year relationships were examined; starting with 2012 playing time.  In a mild surprise, the biggest predictor of 2012 minutes, was not 2012 salary (r = 0.129), or even 2011 adjusted Castrol scores (0.16), but was 2011 minutes (0.625).  This may reflect that a manager’s comfort level plays a greater role in lineup decisions that otherwise expected.  One might hope that teams play their best performers, but this is not borne out by the data.  And to give managers a break, 2011 performance may not be related to 2012 game time for a variety of competitive or health reasons.  Alternatively, a cynic may assume that a manager may put his most expensive players on the field.  But again, the data does not support this idea.

Turning one’s attention to 2012 performance, the best predictor (although mild) is 2011 performance (r = 0.343).  2011 playing time has no relationship with 2012 performance (-0.01), and 2012 salary is not meaningfully related either (.18).  While we know that with the DP system in MLS, salaries often reflect business value more than on-the-field performance.  Thus, a small salary-performance correlation is not much of a shock.  Here, the key concern is the disappointing relationship between the 2011 performance Z-score and the 2012 Z-score.  A quick rule of thumb would predict that this relationship would explain about 10% of the 2012 performance.  It’s not zero, but only a small benefit, and would likely be captured by other performance evaluation standards.

But one can look at the situation from a larger perspective.  Instead of using the 2011 scores to predict a 2012 score, one may use the 2011 score to classify players into a couple groups, and then determine if this classification helps determine which players will turn in desirable or undesirable performances.  For the statistically minded, this would convert a continuous variable into a discreet variable.  And taking Nate Silver’s admonition to think probabilistically to heart, one can determine whether the 2011 performance data improves the chances of predicting a good 2012 performance.

Here, of the 2012 MLS field players (i.e., excluding goalkeepers), 46.6% had above average adjusted Castrol Scores, and 53.3% had below average performance scores.  In other words, if one had no other information, and randomly picked players, they would select a desirable performance (i.e., an above average adjusted Castrol Score) about 46.6% percent of the time.  We can call this the hit rate.  Any data we can use to improve this hit rate may be of value in personnel evaluation and selection.

Using the adjusted Castrol scores, we can classify the players into two groups based on their 2011 performance.  Roughly speaking, we will call them the good 2011 performers, and poor 2011 performers.  Here, the good 2011 performers are defined as those players with Z-scores at or above 0.5, and the poor performers are those players with Z-scores at or below -0.5.  Under a normal distribution, this corresponds with the top and bottom 30% of scores.  However, because of turnover in the MLS player pool, these cutoffs correspond with the top 28% and the bottom 33% of the players with both 2011 and 2012 scores.

2011 performance

2012 performance

All Field Players

Difference

 

(% above avg)

(% above avg)

 

Top 28%(Z= +0.5)

60.8%

46.6%

14.2%*

(% below avg)

(% below avg)

Bottom 33%(Z= -0.5)

68.8%

53.3%

15.5%*

*Statistically significant; α ≤ .05, two-tailed

Top 2011 performers had a good 2012 performance over 60% of the time, significantly better than the population as a whole (46.6%).  Looking at the other direction, bottom 2011 performers had a poor 2012 performance over 68% of the time, significantly worse than the population as a whole (53.3%).  Thus, the 2011 rating is associated with 2012 performance.  While a good or poor 2011 score does not match up with 2012 performance 100%, nor does it help differentiate between players within the 2011 performance group, knowing the previous performance level increases the chances of getting the decision right.  And with limited roster space, and limited financial resources, relatively small advantages can lead to meaningful advantages.

Another way to look at the data would be to line up the three Z-score groups with the probability of an above-average performance the following year.  Those players with a 2011 score 72nd percentile and above, between 72nd and 33rd percentile, and those below 33rd percentile had a probability of a good 2012 performance of 60.8%, 46.6%, and 31.2% respectively.

2011 Performance Probability ofgood 2012 performance
Z > .05(72nd percentile and above) 60.8%
0.5 > Z > -0.5(between 33rd and 72nd percentile) 46.6% (default %)
Z < -0.5(33rd percentile and below) 31.2%

By broadly categorizing the 2011 performance level, one can determine the probability of a good performance the next year.

For one last comparison, we can compare the predictive value of 2011 performance versus player 2012 salaries (which are negotiated and signed prior to taking the field for the 2012 season).  And there is a common theme that spending on players dictates success on the field.  However, 2012 salary does not predict 2012 performance as well as adjusted Castrol Index scores from the previous season.  A top salary is associated with a good 2012 performance 55.7% of the time, which is 9% above chance.  And a bottom salary is associated with a poor 2012 performance about 55% of the time, only 2% different than chance.  Both of these values are not statistically significant.

2012 salary

2012 performance

All Field Players

Difference

 

(% above avg)

(% above avg)

 

Top 28%

55.7%

46.6%

9.1%**

(% below avg)

(% below avg)

Bottom 33%

55.8%

53.3%

2.5%**

**Not statistically significant; α ≥.05, two-tailed

All in all, the MLS data suggests that knowledge of the previous year’s adjusted Castrol Scores can significantly increase the probability of a hit (i.e., selecting a field player who will turn in a good performance) as well as significantly reduce the chances of a miss.  Further, the adjusted Castrol scores are better predictors than player salaries.  It remains possible that additional data – and larger sample sizes – will allow for refined probabilities.  In the meantime, one can layer in an objective measure into their player performance predictions.