MON, 4PM ET
QPR2
AST0
TUES, 3:45PM ET
SHR0
CHE1
TUES, 4PM ET
LIV0
SWA1
WED, 3:45PM ET
MCFC
NUFC
WED, 4PM ET
TOT
BRI
THURS, 3:45PM ET
VER
LAZ

The Americanization of Statistics By Premier League Clubs


4414502116 daf19113b4 The Americanization of Statistics By Premier League Clubs

Photo credit: John Marcus from http://www.jwmarcus.com

Was Thierry Henry lazy? Is Florent Malouda the most valuable player in the English Premier League? Are NBA analytical models being used in the EPL?

All were questions and discussion points during the “Soccer Analytics” panel held Saturday at the 5th annual MIT Sloan Sports Analytics Conference in Boston. The conference, which has grown rapidly over the last five years, drew more than 1,500 sports analysts, executives, managers, coaches, agents and owners over the weekend from more than 47 sports franchises across the world. Baseball, which was has been a central topic of the conference since its inauguration, along with basketball, were major topics once again, but international football was also a hot topic.

While European traditionalists would be quick to mock the Yankee-created panel title, the participants and discussion points gave it a very British tea and crumpets feel. With crucial, late campaign matches taking place across the Premier League over the weekend, several executives from top EPL clubs took time to cross the Atlantic to discuss analytics. The global game of football, unlike baseball and basketball in America, has yet to reach its analytical tipping point. However, it is ripe for change. With a growing American influence on the business of football across England, the revolution is coming. Steven Houston of Chelsea and Gavin Fleig of Manchester City, who participated on the panel, were hard pressed to argue with that during the day’s discussion. You’d also find them hard pressed to argue that working with American analysts within the National Basketball Association (NBA) and Major League Baseball (MLB) has been a good thing for their respective clubs and the overall future of the Premier League.

“We like working with American sports franchises,” noted Houston. “One benefit, we’re not competing with them. Secondly, they do so much analysis.” It shouldn’t be surprising to hear Houston say that. The head of technical scouting and data analysis for Chelsea has a deep understanding of statistics in American sports. He cut his teeth within the NBA for the Houston Rockets as an analyst – applying data analytics to international basketball prospects. At which time, he worked for Daryl Morey, current General Manager of the Houston Rockets, Co-Chair of the Sloan Sports Analytics Conference and the focus of a New York Times feature by the author of Moneyball on his use of analytics in basketball.

Following that impressive apprenticeship, Houston moved to Chelsea in 2009. He now works closely with senior management, including Michael Emenalo and Carlo Ancelotti, on data modeling and visualization, statistical and video analysis, and developing technologies. However, it hasn’t been a necessarily easy transition. Although he joined the English Premier League several years after the introduction of statistical analysis at clubs like Bolton, he didn’t have the data he worked with in the NBA. With the Houston Rockets he had pre-established values associated with literally hundred of points, passes and rebounds. Chelsea, like many other English Premier League clubs, didn’t have league available data and had far less scoring to attribute events to than the NBA or MLB. To solve the problem, he along with other analysts at Tottenham and Fulham – that were also in attendance at Sloan – have worked within their respective clubs to set values for connecting a pass, intercepting a pass, completing a tackle, winning a header and much more.

They’ve certainly made strides since then. Fleig, who worked with Sam Allardyce from 2004 – 2007, has made strides as well. A strong proponent of Allardyce’s use of analytics, he followed him to Newcastle and eventually made his own way to Manchester City. In fact, he credits Allardyce’s early introduction of analytics at Bolton for its long run in the top division of English football. During his time with the Wanderers, Fleig was part of a financially-driven, multi-year effort to develop a model of where and when on the pitch games were won. Specifically he analyzed the differences in statistics between clubs that got relegated versus those that escaped relegation. The approach at the time assisted Bolton in acquiring undervalued assets – like 34 year old Gary Speed. At that age, Speed appeared to be an untouchable player at his requested price, but Bolton had statistical evidence that illustrated his play was not on the downturn.

While Fleig acknowledges that the nature of football doesn’t permit the same statistical analysis of American sports, he believes there is a definite use for the analytical approach in the EPL. “It’s true, but the idea is to understand the characteristics of the team and develop a plan to make optimal use of it – driven by analytics.” With Fleig, Houston and other analysts setting the precedent for video and data analysis, the rest of the Premier League has followed suit. Today, all Premier League clubs have cameras to track match data and roughly 95 percent of Championship clubs do as well. At Manchester City, Fleig has a team of seven analysts with him working with the first team all the way through to the u-9 boy’s squad. Without Fleig and Allardyce, Bolton also continues to develop their data modeling and other top of the table clubs like Liverpool and Arsenal are catching up behind the leadership of analytical minds like John Henry and Arsene Wenger.

So, have the elite sports franchises in England caught up to the elite sports franchises in America? Not yet. Houston notes that there still is an education process required to sell analytical approaches to front-office people and scouting heads within Premier League clubs that isn’t as widespread in America. They both also cautioned analytics advocates, echoing the idea that data is just one piece of the managing process. Owners and managers still need to apply their football knowledge to decisions based on what they see on the field, with the assistance of analytical data. In addition, all the panelists stressed the need to improve their systems in order to adequately compare players across global professional leagues that are facing various levels of competition. The need will grow even more important if UEFA financial regulations are imposed in the future and clubs with seemingly endless funds are forced to finally uncover undervalued assets like more fiscally responsible clubs are today.

“You don’t need analytics to know that Messi and Rooney are great players,” chimed Houston. “Analytical systems are useful to find the best role players, or the best players for a particular team.”

Follow Kyle Austin’s updates on the business side of football and soccer at http://socceronomist.com and @Coach_Austin

This entry was posted in General, Leagues: EPL. Bookmark the permalink.

17 Responses to The Americanization of Statistics By Premier League Clubs

  1. Bishopville Red says:

    It’s true that football has nowhere near the statistics of North American sports, but that’s not necessarily a bad thing. While Baseball has a number for everything under the sun (or lights – night games can change players stats), the crux of the issue with MONEYBALL was that people put faith in the wrong stats. Without decades of bias, football can cut through the myths and get to the relevant much quicker – if it’s willing to keep an open mind (wait and see on that last point!)

    The biggest problem you face is getting stats on players outside your club. It’s great to track and analyze players in your system as early as 9 years old, but the more immediate value of this insight will come with more effective transfer market moves on both sides of the ledger.

    With the staggering amounts getting thrown around for players today, it would be fantastic to have some data that supports a scout’s instincts, or reinforces a manager’s valuation of a player. But in order for that to happen, you’d either have to track almost the entire planet (depending on the size of your club) or rely on others to do that level of work along with (re: for) you. For a variety of reasons, that’s not going to happen, probably ever.

    • BMcKeon says:

      “the crux of the issue with MONEYBALL was that people put faith in the wrong stats”

      You are kind of right, but miss the major point of Moneyball, which is its most valuable lesson. It’s not that the wrong stats are wrong necessarily, it’s that teams put their faith solely in those stats along with their eyes . Thus, players who weren’t necessarily pretty on the eye and didn’t put up eye-popping stats (the outdated Avg, HR, RBI, Runs Scored) were extremely undervalued in the marketplace. They could be had for bargain prices, allowing a team to build a squad of competent and dependable role players without breaking the bank and allowing flexibility to spend money on other things.

      It’s asset allocation; putting your assets to work in the most efficient, market-friendly way.

  2. Jason says:

    I’m in the opinion that stats don’t lie. However, they can be badly misinterpreted. Stats are only helpful if can undestand the context and relative position of that stat as it relates to the player and the organization.

  3. SeminoleGunner says:

    Enjoyable article.

    I think a lot of American fans have been shocked by the relatively small role that statistics and video analysis (tools that are so massive in American sports) played in soccer in the past. It’s exciting to hear that may be changing.

    The next step I would like to see are the advanced statistical tools being made available to fans.

    • R2Dad says:

      I would settle for any statistical data/information that would be available to coaches of the USMNT–they certainly could use some. The faster/taller/stronger metrics that have been used for the past 20 years have got to give way to better models.

  4. Chetman-UTD says:

    Don’t do it! Pleeeeaaaaasssse don’t American-ize the game with stat upon stat, upon stat. I can tell you that I have not watched an american game, anything!!!!, in almost 10 years because I got tired of the non-stop yakity yak of the broadcast booth, with stat after stat comparison of players and what they have done for the past 10 years! I fell in love with “THE BEAUTIFUL GAME” during the ’06 world cup. The first thing I noticed about the broadcast was the SILENT periods where the game explained itself. As an EPL fan from New Jersey USA, I can find all of the info I want on the internet. Again, please don’t American-ize the game. We have the MLS for that.

    • Moose9t9 says:

      I agree with you. The reason everyone loves this game is because you feel something when you watch it. Theres something non-robotic about it, you can tell its humans playing; making independent decisions, creating new dynamics for certain situations.

      My guess for the future is players will become almost robotic. There will be set plays (I know there are some now but it will be ridiculous in the future) to maximize results. The biggest effect? Less Messi’s, Xavi’s, Ozils, Wilshere’s, Guti’s and Berbatov’s. I understand I left many players off (because I personally think Rooney is overrated) but the names mentioned above are players that I find fascinating to watch.

      Rant makes sense in my head ;-)

  5. Dave C says:

    The problem with stats in soccer is that many kinds of event do not happen often enough to draw any statistical conclusion from. For example, in Basketball, rebounds happen dozens of times per game, and teams play a lot more games than soccer teams. So statistics on rebounds can reliably tell us a lot about a player’s prowess at grabbing rebounds. Likewise, a pitcher in baseball throws many pitches every night, and plays several games every week, so good statistical analysis can be made of his statistical ability.

    Soccer, on the other hand, involves far fewer games, and many events happen fairly infrequently (eg how many corner kicks does Ryan Giggs take in each game, or per week, or per season? Not a lot, statistically speaking). So simply due to the smaller population size, its harder to draw meaningful conclusions.

    Anyway, I’m no statistician, I’m just throwing in my half-informed two cents. I don’t mean to suggest that statistics can’t be applied to soccer.

    • R2Dad says:

      Good points. The higher-frequency events like completed passes are a valuable tool–more so than possession, which is always listed as one of the key metrics of a match. I anticipate that valuable metrics will be developed for backs who concede fewer corners, backs who get skinned by attacking players, percentage of chances a player successfully dribbles past an opponent.

    • Kyle Austin says:

      That’s absolutely right. Everyone in attendance discussed the issue with not having enough points (goals) to tie values to. For instance, think of all the scoreless draws that you cannot connect values to.

  6. Thomas says:

    I think it’s interesting to look at soccer from a statistical point of view like this. And I think you’re going to continue to see a lot more analytics involved in transfer decisions, etc., as the gap between the rich and the poor continues to grow.

    Many of you have pointed out what problems are posed by bringing a purely statistical point of view to the game. Sure you can look at passes completed, etc. for a player like Xabi Alonso, but if you look at it from a purely statistical stand point, he doesn’t nessecarily contribute goals and assists.

    A lot of what makes a player great, especially in soccer, are intangible qualities.

    Take a player like Makelele. How can you statistically quanitfy what he does?

    I think often players are given merit (rightly) based purely on how they affect the game. Often this isn’t statistically mesaurable. A fancy backheel, or cheeky flick that results in a goal looks the same as a 2 yard pass on a stat sheet.

    It simply doesn’t do the sport justice.

    But that doesn’t mean there isn’t room, and tremendous upside for it.

  7. ish says:

    stats have always been a part of football, they just havent been used as much in terms of quantitive measurments. we all know xavi is a great passer but stats back that up. dribbles completed, crosses completed from areas in the pitch. distance covered, delta, top acceleration, top decceleration, times in a match players hits 90% max top speed/acceleration. chances created.

    These are all stats based on technical, physical nature of the game. The big difference is the mental aspect. Yes xavi has a wonderful first touch and pass but his ability to read the field is what makes him deadly, lampard can time his runs, inzaghi can read play and move accordingly. Stuff like this is harder to quantify and needs proper scouting. In general unlike the stop start nature of nfl and basebal, football is played more uniformly, nba is as well but due to the relatively small size of the court compared to a football pitch the tactics are much more simplified. essentially it is more akin to futsal then true football.
    of course the model needs to take into account the opposition, your team, formation, role on the field and mutiple other factors which makes the analysis very very difficult, something scouts have trouble doing as well.

  8. ddtigers says:

    It just kills me how I hear “everyone” hates stats but when I listen UK based podcasts on football(about 8 a week), all I hear is stats. There is nothing wrong with using stats as a measure. We just know in soccer there is more involved, the intangible factors. I love the stats in all the sports I follow. The “Americanization” of football, the added stats, will not change the game or effect how fans look at the game.

  9. mosdef says:

    On the sporting side of it at the top level ,Stats have been used in the game for a while. Its on the reporting side that I do see stats being made more readily available by the media houses, reporters and the game analysts. From Labonovsky in the 70′s when he acquired a computer, to Total football and now Tiki-taka, stats or a scientific approach were always used to tweak systems. With the sporting world becoming smaller and smaller and especially with sports science becoming more prominent, stats are pretty much used for everything, but you watch the bundesliga on Tv and you realise its the SKY, EPL style of reporting that is changing more towards “Americanization”.

    • Chetman-UTD says:

      All I was trying to say is keep them off the broadcast. Let the game explain itself. I don’t know, but I just find a Tim McCarver-esque (no offense Tim) type of announcer a big TURN-OFF. Which is what I did to the volume on the t.v. before giving up totally on American sport. I kind of like the opportunity to hear the fans team chants in between comments by the broadcasters . Leave it be!

  10. Keith says:

    Moneyball was once overly praised but now it’s way under valued. In baseball, hockey, and football (soccer) the hardest thing to do is score. Baseball statisticians realized that they key stat that lead to an increase of runs was the on base percentage. From there it’s easy to see how they developed more and more stats to find out the true value of a player when luck is removed (as much as possible). The problem with hockey and football is that all the players are interconnected so it’s harder to credit a pass completion. It could be an amazing pass or the receiving player might have made an amazing run to catch a pass. I think all sports WAY over value easy basic statistics like goals, assists and tackles (or saves for goalies). Someone like Darren Bent might score a ton but if the goal is a tap in then little credit is deserved. Maybe someone else in that exact role would do even better. But maybe Bent is doing something amazing like predicting where a rebound will go. But because we look at the “goal” statistic we ignore “getting open”. Football will never have a chance to be as statistically broken apart as baseball.

    • word says:

      Also, it doesn’t matter how great a pass you made if the receiver fails to control it (in both footballs, bball, and hockey).

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>