1. Save 40-80% on great soccer jerseys. Shop today at BigSoccer Shop!

My definition of Soccer Analyitics

Discussion in 'Statistics and Analysis' started by ENB Sports, Mar 5, 2012.

  1. ENB Sports

    ENB Sports Member

    Joined:
    Feb 5, 2007
    Via My Blog - http://enbsports.blogspot.com/2012/03/my-definition-of-soccer-analyitics.html

    The past weekend was the 2012 MIT Sloan Sports Analytic Conference. Due to the Moneyball phenomenon the conference has been very successful and I have hoped to attend have been unable to do so. During the conference a hour Is dedicated to Soccer. Originally soccer was part of the “other sports” section but there been enough interest so in 2012 it actually took place in the main conference room.

    During this conference the people who have an interest in soccer also plan a meet & greet before hand and many of the people who I read and follow on social networks such as twitter attend this and from what I heard the conversation during this get-to-together is better than the conference itself.

    So in general I don’t intend this article to be critical of anybody who currently is working on Soccer Analyitics but I do feel that there is a disconnect between effort behind soccer analytics and the game itself which is why in this article I want to give in my opinion a basic definition of Soccer Analytics and what I perceive its potential in the game.

    I’ve been working with soccer statistics for over 20 years and in the process I’ve tabulated millions of lines of data and currently cover 60 plus leagues, 1000 plus teams and close to 70,000 players. The reason I started this work is I was a huge North American stat geek who was getting bored by those sports so started following soccer. What I then realized that in comparison to all North American sports there was no statistics for soccer.

    I created a box score/template system to collect what I regarded as essential data using much of the same format used in a game with many similarities of play, that being Ice Hockey. I tested my system on local youth leagues and then started recording data for the English Premier League in 1992-1993 and the 1994 United States World Cup. Ever since I started there has been interest in the work and for the last 15 or so years I worked professionally as a Sports Statistician concentrating on Soccer.

    This being said Statistics in Soccer are still not common. Much of my own work is fairly exclusive to any work out there and I’m probably the only person in the World who is familiar with these 70,000 players J. An unlike North American sports the general public who follow the game have no encyclopedic knowledge of statistics. Even leagues like the English Premier League people only know who scored the most goals they don’t know who had the most assists or shots on target.

    So first up we’re dealing with a very naïve market place with no history or culture of statistics in comparison to baseball Bill James started his work 80 years after data he used started being collected. For example, there was more statistical data in relation to Baseball in 1928 than there is in the current New York Times. I experience this daily when people unfamiliar with my work come across it. I would argue that I produce the least complex type of work yet many readers are over whelmed by the amount of data and what it means.

    On the positive side it’s an open and overly underdeveloped market with a potential audience of billions to begin to get it. I feel soccer statistics alone have been poorly marketed as a commercial entity with misrepresentation in broadcast and media, limited amount of information in books, magazine and newspapers and no properly developed fantasy game or devices that would use statistics.

    Even more advantageous is this lack of general knowledge increases the effect of Soccer Analytics over any other competing sport if used in a real life scenario. The other reason why Soccer Analytics has so much great potential is the business of the game it’s self. The money in soccer worldwide is greater than all other sports combine and they use basically unregulated business system which instead of dealing purely with commodities through analytics you can actually generate real money. This season there has over $US 4,000,000,000 in player sales alone which with inflation has been the norm for over the past decade.

    In 2009 Manchester United sold Critiano Ronaldo to Real Madrid for close to $US 200,000,000 a few years before they bought him of Sporting Lisbon for around $US 40,000,000 that means Manchester United generated over $US 160,000,000. In comparison the Milwaukee Brewers lost their player Prince Fielder and in return they got the 27th pick in the draft.

    So now understanding the importance of Soccer Analytics we than need to ask ourselves what can we do with it. I think what we need to understand is the root of Bill James work and that being through his observation of the game he was confused why certain things were happening and start to write them down as questions and originally his books (which I read 90% of them starting from the age of 11) try to answer these questions.

    So let’s start asking questions about soccer.

    First up I would suggest ignore all game used analytics, I think the smartest thing Billy Beane has chosen to do is not to watch the game. In comparison if you owned a store you would not watch on a daily basis what people were buying instead after a set period of time you would analyze the data and create conclusion on how you are doing and what you need to improve.
    To me the most important questions in Soccer Analytics are:

    How to evaluate the transfer value/salary of a player?

    What is the difference of quality between leagues? (including as part of youth development)

    Although in the process of my work I’ve come up with hundreds of question in relation to Soccer Analytics. Examples include:

    Are away wins more valuable than home wins?
    At what age should a player be bought?
    At what age should a player be sold?
    Can a team develop a goal scoring style?
    Can you sell a player you picked up on a free transfer?
    Do corners lead to more goals on average than crosses?
    Do fouls lead to yellow and red cards?
    Does a club perform better after close or large loss? is there a difference?
    Does a team discipline hurts their chances for success?
    Does ball possession lead to success?
    Does having defender who can score benefit a team?
    Does having defender who create assists benefit a team?
    Does loaning a player out benefit his career?
    Does shots missing the net lead to goals?
    Does stadium have a factor on how an individual game is played?
    How beneficial is promotion to the development of a club?
    How big is the jump from division to division?
    How difficult is relegation to the development of a club?
    How many days between games till a team properly recovers?
    How much consistency is there in a player’s career?
    How much detriment is a red card?
    How much does the size of a goalkeeper matter?
    How much influence does a manager have on a player?
    How much influence does a manager have on how many goals are allowed?
    How much influence does a manager have on how many goals are scored?
    How much of a role does injuries play in a clubs failure?
    How much of a role does the referee tendencies have on a game?
    How much value is a draw over a loss in a playoff battle?
    How much value is a draw over a loss in a relegation battle?
    How to properly evaluate talent?
    How valuable is a defensive midfielder?
    How valuable is an assist in soccer?
    How valuable is save percentage in soccer?
    Is a specialist (long throw, free kick) on the pitch worth it?
    Is important that your central defenders are taller than other players on the pitch?
    Is it better to have one primary goal scorer or scoring shared amongst the team?
    Is it easier to score with your head or outside the box?
    Is it more beneficial to have a good offense or defense?
    Is it worth taking a direct free kick on net?
    Is playing offside trap worth the risk?
    Should a club change it's formation after receiving a red card?
    Should you never sell a player?
    Should you pay a transfer fee for a player?
    Should you recruit a player for a particular roles or just talent?
    What characteristics does a top class defender have?
    What constitutes a successful season?
    What is the best format for a fantasy soccer game?
    What is the home field advantage in soccer?
    What style of play raises a player transfer value the most?
    When does a player reach his prime?
    Which plays a bigger role the system or the players ability?
    Which position should you put the most resources into?
    Who should take corners/free kicks?
    Who should take penalties?

    I do feel we have the material and the process to answer all of these questions in a analyitical way and that would provide interesting reading but more importantly a benefit to a club and an influence on how the game is played.

    My emphasis is on my two core questions although I feel I incorporating a lot of the other questions in the process and if I find answers I will post my results. This being said the work takes a huge amount of time and effort, and at this point other the reward of accomplishment and smugness the return is underwhelming because like Bill James story the work primarily dismissed by everybody who has any connection to the game.

    Will this change? I assume it will but to be honest with you I’m very skeptical because the sports industry is a very egocentric and impractical business process where facts and stats get in the way of person thinking they know better. On the positive side unlike many things you can be judged by the success of your actions in sports so hopefully someone will be given the opportunity to change the game for the better.
     


  2. TouchMint

    TouchMint New Member

    Joined:
    Mar 26, 2012
    Location:
    Tempe, Az
    Club:
    AC Milan
    Pretty nice read I am also a pretty big stats guy but you have taken it to the next level for sure!
     
  3. SPA2TACU5

    SPA2TACU5 Member+

    Joined:
    Jul 27, 2001
    Question: why is this important to know?

    What is the difference of quality between leagues? (including as part of youth development)
     
  4. SPA2TACU5

    SPA2TACU5 Member+

    Joined:
    Jul 27, 2001
    I found the bold questions to be interesting. Useful, in some cases. But at least interesting.

    Are away wins more valuable than home wins? no, they're both worth 3 points
    At what age should a player be bought? cannot be answered
    At what age should a player be sold? cannot be answered
    Can a team develop a goal scoring style? yes
    Can you sell a player you picked up on a free transfer? yes
    Do corners lead to more goals on average than crosses?
    Do fouls lead to yellow and red cards? yes
    Does a club perform better after close or large loss? is there a difference?
    Does a team discipline hurts their chances for success? yes
    Does ball possession lead to success? yes (it's impossible to autonomically score goals without being in "possession")
    Does having defender who can score benefit a team? yes
    Does having defender who create assists benefit a team? yes
    Does loaning a player out benefit his career? yes, although impossible to answer
    Do shots missing the net lead to goals? yes
    Does stadium have a factor on how an individual game is played?

    How beneficial is promotion to the development of a club?
    How big is the jump from division to division?
    How difficult is relegation to the development of a club?
    How many days between games till a team properly recovers? I'm pretty sure this has been researched.
    How much consistency is there in a player’s career?
    How much detriment is a red card?
    How much does the size of a goalkeeper matter? probably impossible to answer

    How much influence does a manager have on a player?
    How much influence does a manager have on how many goals are allowed?
    How much influence does a manager have on how many goals are scored?


    How much of a role do injuries play in a clubs failure? define "failure"
    How much of a role does the referee tendencies have on a game? define "tendencies"
    How much value is a draw over a loss in a playoff battle? define "playoff battle"
    How much value is a draw over a loss in a relegation battle? define "relegation battle"

    How to properly evaluate talent?

    How valuable is a defensive midfielder? define "def midfielder"
    How valuable is an assist in soccer? (in general very valuable; an "assist" implies a goal was scored)
    How valuable is save percentage in soccer? define "save"

    Is a specialist (long throw, free kick) on the pitch worth it? yes
    Is important that your central defenders are taller than other players on the pitch? no
    Is it better to have one primary goal scorer (or scoring shared amongst the team)? no
    Is it easier to score with your head or outside the box? if you mean "with your head from inside the box" then head is the answer
    Is it more beneficial to have a good offense or defense? offense
    Is it worth taking a direct free kick on net? yes
    Is playing offside trap worth the risk? yes
    Should a club change it's formation after receiving a red card? isn't it inherently changed? 4-3-3 becomes 3-3-3 I would say that's a formation change
    Should you never sell a player? no
    Should you pay a transfer fee for a player? yes
    Should you recruit a player for a particular roles or just talent? just talent

    What characteristics does a top class defender have?

    What constitutes a successful season?
    What is the best format for a fantasy soccer game?
    What is the home field advantage in soccer? it's when the location of where the game is being played has positive influence on the performance of the home team - I guess the question is "how big is the home field advantage?"

    What style of play raises a player transfer value the most? offensive

    When does a player reach his prime?
    Which plays a bigger role the system or the players ability?
    Which position should you put the most resources into? The one with the most influence on the performance of the team.
    Who should take corners/free kicks?
    Who should take penalties?
     


  5. ENB Sports

    ENB Sports Member

    Joined:
    Feb 5, 2007
    In terms of "Moneyball" for soccer i feel the key is finding players via transfers or salaries that will give you the most bang for the buck. Questions such as is a 20 goal scorer in the MLS worth anything to an EPL Club?

    Understanding the difference of quality between leagues allows us to use statistical data and actually compare player performance at an comparable level.

    This was one of Bill James question in understanding baseball although more in a historical reference than league comparison (funny enough Bill James did little with Collage or Minor league ball) - ie: In 1990 he compared a 23 year old Roberto Alomar to Joe Morgan and to do this he took into account the average league statistics of Joe Morgan at 23 and compared to the current league stats of a 23 year old Roberto Alomar and from that he basically projected Roberto Alomar would have a HOF career.

    One of the project I been working on recently is tabulating statistics for NCAA Division Men's Soccer so far been doing it for four years but my hope is by say Superdraft 2020 i can use the players from that draft and statistically compare them with current MLS players and come up with an idea on who should be drafted where.

    Internationally you have cases like Dortmund's Kagawa who was bought from his Japan club for $300,000 and is now worth $25,000,000 on the transfer market. So finding a player who could give you that type of return could change the outcome of your team for many years to come.

    Boring to some but basically its the business of sport which I feel has become to true meaning of success in sports by manipulating the odds to give yourself or your team a greater advantage in winning over everybody else.

    -----

    When i wrote this thing through frustration i just came up with those questions off the top of my head without putting much thought to each :) Although to reiterate the point i find biggest problem in soccer analytic is people are looking for answers more than asking questions and I feel by asking questions you get a better idea of what you're looking for.
     
  6. SPA2TACU5

    SPA2TACU5 Member+

    Joined:
    Jul 27, 2001
    Can I disagree on some parts?

    Not really. Because soccer is a game of various playing styles. Leagues have different playing styles. Ie. Spanish players tend to be technical and thus perform well in a league where technical soccer is played.
    Even if you would find the EPL to be statistically stronger than the 'Primera Division', it would not mean the stars of EPL will do well when transferred to Spain.

    Isolating events, actions, any factors, is the most dangerous thing you can do in soccer - as an analyst.

    I don't think there is any way to compare how stats work in baseball to how soccer (-statistics should) works (unless you're looking at penalty kicks or kick offs). Even for the only specialist position in soccer, goalkeeper, this will not lead to any viable conclusions.
    Besides, I don't know how much baseball evolved over a period of 30 years (but taking an educated guess I would say hardly), but soccer has greatly changed from how it was played 30 years ago.

    I agree this could change the outcome of your team for many years.
    (That is, in case the club can actually sell him for $25,000,000. Which in Kagawa's case remains to be seen.)

    Agreed.

    Agreed. And there's nothing wrong with asking as many questions as you can come up with, as long as you know which ones are the important ones to answer.
     
  7. ENB Sports

    ENB Sports Member

    Joined:
    Feb 5, 2007
    From my own work and more detailed work I have seen i would disagree.

    I've actually have seen establish trends in soccer information that can allow you to come up statistical comparisons both historically and between leagues.

    Just one example is a referees eagerness to give Fouls and Cards directly changes the style of play and how a player plays.Through my own league comparison charts (i been doing these stats for close to 10 years here is the 2011 data) i've seen consistent data where I can almost predict the exact totals or range for years to follow.

    Final 2011 League Comparison Charts - Home Advantage - http://enbsports.blogspot.com/2012/02/final-2011-league-comparison-charts.html
    Final 2011 League Comparison Charts - Goals per Game - http://enbsports.blogspot.com/2012/02/final-2011-league-comparison-charts_26.html
    Final 2011 League Comparison Charts - Type of Goals Percentage - http://enbsports.blogspot.com/2012/02/final-2011-league-comparison-charts_9177.html
    Final 2011 League Comparison Charts - Assists and Shots on Goal - http://enbsports.blogspot.com/2012/02/final-2011-league-comparison-charts_9131.html
    Final 2011 League Comparison Charts - Disciplinary - http://enbsports.blogspot.com/2012/02/final-2011-league-comparison-charts_4762.html

    When it comes to baseball statistics my opinion is at least 90% of the time the player doesn't create his own data on purpose. You can sacrifice hit, you can in someways walk and strikeout but a player doesn't know he's going to hit a home run when he goes up to bat.

    So in that sense all statistics in sports are the same there not preset in stone but a historical reflection of what a player has done in the past and potentially what you can expect out of him in the future. Which you can use for your knowledge/advantage for all sports and quite honestly statistics are used the same way in many aspects outside of sport as well.

    The other thing is statistics in Baseball been around for 130 years analyzing of detail statistics and what they mean have been around for close to 30 years and only used in the past 10 years. When I started for English Premier League in 1992 there was only about 5 people in the World doing the same. Analyitics for soccer have only been around 10 years and an interest in soccer stats (an I would know :) ) is only starting now.

    So instead of comparing sports I would compare Work and Interest and would argue if statistics were prominent component of soccer since it's beginning based on it size we probably have far more stories of "Moneyball" for soccer than we do for Baseball.

    in terms of the Kagawa i agree although if an MLS/NASL/USL Pro team was able to produce players consistently worth $1,000,000 on the transfer market it would almost guarantee the long-term existence of that team even if the fan base disappeared.
     
  8. SPA2TACU5

    SPA2TACU5 Member+

    Joined:
    Jul 27, 2001
    You're gonna have to explain to me what you're showing me here and why.
    Sure these are league comparisons, but they're not a "comparison of quality of leagues".


    I agree.

    Maybe. Although I doubt it. I think people have been thinking about statistics in soccer for over the past 100 years, but it's taken a lot of time before people finally thought of ways to use them. Actually I believe hardly anyone's found a way to really dissect soccer.

    The thing with baseball is, it's very easy to dissect it. It's a turn based sport. With one repetitive event. The entire game of baseball is based on that single event. If the event wouldn't take place, a game of baseball would have no meaning. In soccer though, there is not one single event which defines the game of soccer. And I could go on and on as to why baseball is completely different.

    But they're only worth it when they're sold. So if you mean by "being worth $1,000,000 is being de facto sold for $1,000,000" then yes I agree.
    If you mean they could be sold for $1,000,000, I can come up with some examples of clubs consistently developing players that are worth $1,000,000 but never sell those players for $1,000,000 and those clubs have almost gone bankrupt.
     
  9. ENB Sports

    ENB Sports Member

    Joined:
    Feb 5, 2007
    In terms of your question regarding the value. Currently the MLS or any team in US is not actively selling players the main reason for this is the USA sees themselves as an individual entity oppose to being part of the global world of Football. So in the case of players like Ream, Edu, Altidore and others it's their agent who is actively trying to get their player a move to Europe and not the league itself.

    Which very much like my view on statistics is based generally in my opinion on niavety of the market place. No team in US puts the procedure of buying/selling players as part of their business model where in Europe/Elsewhere this is around 30% of their total business plan. So if you are looking at something to exploit in the US soccer market that would definitively be the first place i would look.

    In terms of a define stat in soccer there is Wins, Draws, Losses, Goals Scored and Goals Allowed and if you can control the outcome of those statistics you can control the outcome of the final result. Although that is kind of a cop out answer so I would ask the same questions about the other sports.

    Watch a collection of baseball games for a set period of time with out any recording of statistics other than how many runs each team scores and who scores them. When you do analyse it, what you'll get is what we currently get in soccer a statistical analyse of scores and goalscorers and individual observer opinion of what he thinks from watching those games.

    If Baseball or any North American sport was reported on like Soccer is reported on in Europe are perception of the game would be totally be different. As I said what we're dealing with here is less of a stats thing and more of a cultural thing on how we can perceive the game.

    Alternatively if we had complete statistics of every soccer player from the age of 12 (like we do in baseball) all passing records, tackling records, his involvement in goals scored and goals allowed and statistical analysis of his opposition we could make a judgement on how he will preform in the future as we already do for all North American sports.

    Truth is for all sports, observation and statistics are useful tools for analyzing a player but neither are perfect on their own.

    Yu Darvish is a great baseball prospect no matter how he's judged be it his 6'5 frame his four pitches and his 95 MPH fast ball or his stats in Japan last season 18-6 1.44 ERA and 276 K's in 232 IP but all we can say is he should be a good pitcher in the MLB not that he will be.

    So I'm not saying or would never say stats are it and if I was in charge of a team I would never sign a player just based on his stats but like all other North American sports i would use statistical output as a consideration of that player and what he is potentially worth.

    One of the things I think we need to understand when it comes to statistical analysis we are not looking for a final answer if it was that easy it would of been discovered a 100 years ago what we are looking for is additional information that might get us 5-10% closer to the answer we are looking for. Whatever the question may be.

    At this point you or others are not wrong by saying Soccer is a different sport than other sports and it can't be analysed like others because the truth is no one even know the answer to that question. Just like no one cared about Walks, OBA, WHIP and million other stats in baseball 10 years ago.

    So My opinion is the analysis is worth it because we don't know, and potentially knowing stuff that others don't is one of the reason for doing anything at all. Know that being said i do feel there is enough information out now that statistics can be used in the business of Soccer to assist a team in generating more revenue and potentially winning more games. We are just waiting for a situation like the Oakland A's for someone to actually take the risk and do it.
     
  10. SPA2TACU5

    SPA2TACU5 Member+

    Joined:
    Jul 27, 2001
    But those are isolated statistics, not actual 'events' or 'actions' like pitching and batting; they define what baseball is all about.

    Scoring a goal, or conceding a goal, does not define the game of soccer, nor does it define a player's performance. There is not one repetitive event in soccer, like pitch-strike, pitch-strike, etc. There is no event which defines the sport, in other words, no event which defines a player's performance.
     
  11. ENB Sports

    ENB Sports Member

    Joined:
    Feb 5, 2007
    In some ways i can understand your view but it also confuses me.

    We can judge player performance in soccer based off of goals, assists, shots, missed shots, passes, tackles, fouls, position on the pitch, how much the player runs just to name a few as I've seen work done analyzing on about 100 more categories.

    The object of baseball is scoring and preventing runs how you do that is up for debate but mostly based on the skill of the player, his opponent, and style of game. In terms of soccer this can be treated in the exact same way through scouting a players against specific competition and a a professional making a judgement on that player in statistics, numbers are calculated and then once again a professional makes a judgement on that player.

    The performance question is how well is that player going to perform in this situation and why. My argument has been that there is specific trends in soccer that you can potentially use to evaluate talent. When Bill James started working with baseball this was all he was trying to prove that players have a tendencies good or bad of performing in certain situation and if you take advantage of this it will improve your chances of success.

    This currently happens in soccer all the time. If Stoke is playing team who is poor at defending corners there is a good chance Stoke will start Peter Crouch as he is likely to score a goal with his head. In terms of the EPL I think OPTA/ProZone has almost taken statistics as far as it can go. Problem is there is 10,000 other leagues in the World with a 100,000 players potentially on the free market so the question is how can you analyse all those players like NCAA Division One Soccer that includes everybody but is also financially viable. Due to my own financial constraints and interests this is the aspect of Soccer Analytics that I concentrate on.

    I think the problem with most sports analytics is that the person working on a project thinks he/she can find some sort of ultimate solution. Sports doesn't work that way I can give you the top 25 baseball players based on any stat and can almost guarantee you that team would struggle to win 100 games.

    So the goal of sports is finding any way possible with stats or not to give your team a slight advantage knowing that once that advantage is noticed other teams will copy or the league will change the rules. So someone like an Alex Ferguson has 1000 ideas either planned or just who he is who gives him an advantage over his competitor

    Why I have difficulty answering specific questions is that's not the point of my work. The point of my work is to analyse every situation that happens and through statistical data try to find a solution that might give an advantage.

    If I say the MLS allows more goals per game by header than any other league so to take advantage of this a team in the MLS should hire two 6'5 strikers a free kick specialist and on every offensive move, cross in to the box and hope for a header. That doesn't mean anything until a team does it and scores a large amount goals because of it.

    Just to come back to your point so I'm assuming that you also feel statistics don't mean much in Hockey or Football or in Basketball (other than shooting baskets). In a sense of truly understanding the game statistically (like you can in someways with chess, poker or black jack) I agree with you and I don't think a statistical model can be created to say this is perfection and can not be beaten although i don't think that can happen in baseball either as there never been a game with 54 outs of all strikeouts or a game of unlimited Home Runs.

    My final point is i believe I know more about the game of soccer than most. I never played it a high level, I've hardly attended any games, and never took any coaching courses but the statistics has allowed me to track close 75,000 players who played the game professionally Worldwide and analyse their career through times of success and failure. So to ignore statistics completely limits a person ability to understand the game not only in soccer but all sports.
     
  12. ENB Sports

    ENB Sports Member

    Joined:
    Feb 5, 2007
    It will not let me edit my horrible grammar :)

    I try not to get into these long written online conversation because of my less than effective writing skills and that i don't have the time or energy to put thoughts down and write them in a understandable manor. The other thing I don't really like is its not a private conversation so I'm responding to you but in a way to all else who might read this.

    I also don't want to sound like I'm dismissing your view point as it is valid and one of the most common conversation when it comes discussing soccer statistics. So hopefully I don't sound defensive when it comes to soccer statistics or that I'm trying to convince others that I'm right and they're wrong.

    The point of original article was not aimed at people who have no interest or care about soccer statistics but instead people who have a passing interest in sports analytics and see soccer as this great opportunity. My belief is the game is 90% and analysis is 10% so you fully have to understand the game to even consider that you can change or improve things through analytics.
     
  13. SPA2TACU5

    SPA2TACU5 Member+

    Joined:
    Jul 27, 2001
    What's the part that confuses you?

    I agree.
    But what I see happening mostly in soccer statistics, is isolating those 'actions' and looking at absolutely numbers. Game 1: 50% tackle rate, 2 tackles; Game 2: 85%, etc. then claiming, player X is a good tackler, or player X had a good performance in Game 2, because his pass completion rate was 83%.

    Which is why Billy Beane took a bit of a different approach with his Sabermetrics for Soccer. Still far from 'perfect' though.

    How? In baseball it's inherent to the sport that the ball will be pitched over and over again, turn by turn, one team attacks, one team defends. In exactly the same fashion, that is, through exactly the same event.
    In soccer this is simply not the case at all. There are not turns, there are no events that have to take place. Teams can literally stare at the ball for 90 minutes. Players can have a great performance without touching the ball once.


    But the only examples you can give here are set plays. Because they, to a degree, resemble baseball's defining event - a set play. Corner kicks, free kicks, penalty kicks. Yes, when statistics show that player X, Y and Z on team A are bad in the air, and provided team A is playing those guys against you, it would seem logical to play your strong-in-the-air striker and get as many crosses in as possible. Then again, if the strong-in-the-air striker is otherwise useless, you might not even come near team A's penalty area.

    But it hardly takes statistics to diagnose these weaknesses, and come up with a "plan of attack" against them.

    I agree this is the way many teams are 'managed'. But it's old fashioned. Managers look at the table, "we scored 50 and allowed 60", so we need better goalscorers, and we need to improve the defense. If we sell the guy who scored 11 goals this season and replace by a guy who scored 16 goals elsewhere, then we'll have an improved squad next season.
    Of course this only works for players like Messi and C. Ronaldo. They will perform regardless of the opposition or their teammates.

    I am convinced there is a solution that is better than looking at isolated stats and better than subjective assessing players/teams.

    Then the stats you're looking at are seriously flawed. I can almost guarantee you that if I can pick 25 soccer players & their coach+his staff, they will win the league at the end of the season - this is pretty much what is going on at Real and Barça.


    Like Mourinho and Guardiola, Ferguson runs the 3rd richest FC in the world.

    "Real Madrid CF € 479.50, FC Barcelona € 450.70, Manchester United FC € 396.89" (source)

    I'm not really sure if their "ideas" give them an advantage over their competitors. I would say it's more about their transfer policy and the style they play. Which in Ferguson's case has been quite poor lately - luckily for him the entire EPL has been poor this season.

    Plus, in case of Guardiola, when the opponent has statistically shown to be weak in the air, are you going to play a tall attacker over Messi? Are you going to play long balls forward? I don't think so.

    Football, turn based, all about set plays, all about specialists.
    Basketball, turn based, all about set plays.
    Hockey has some resemblance to soccer. But there are a couple of reasons why hockey is still quite different.

    So all in all I think using statistics in those sports is a lot easier, and because it's easier, those statistics are being used in a more effective way than statistics are being currently used in soccer.

    But I don't think the objective in baseball is to make an unlimited number of HR's or strikeouts.

    Just to be clear, I am against meaningless statistics in soccer. I am pro statistics in soccer when done right.
     
  14. ENB Sports

    ENB Sports Member

    Joined:
    Feb 5, 2007
    I respect your position I think our perception of statistical analysis is totally different than mine. I guess from this conversation were finding that out.

    When I tabulate my statistics I produce them in a set order of plays from minute 1 to 90+ because of costs I only record stats at plays such as Goals, Shots, Fouls, so in average game through my stats there is about 80 plays i record per game. Where a company like OPTA or ProZone might record 2000 (every touch and pass)

    The final work is exactly like Baseball, Basketball, NFL, where the stats are totally based on possession by possession so in a game analysis you can break down the numbers any way you want including the exact same way they do in other sports. I would agree with you that individual game statistics pretty much mean nothing and personally i don't do any work with that data. What is important to me is consistent trend throughout that players career where from that you can start setting expectation for future performance.

    I guess it can as you say fall in the trap my team needs more goals so I replace the 10 goal scorer with the 16 goal scorer. This is actually done often because in most cases goals is the only stat recorded in soccer. I don't have any work but it might be interesting to see how successful doing this is.

    Using your Messi example this actually was a debating point for Barcelona when they had Ibrahimovic in 2010 it was unsuccessful because you had to pay and keep both players happy but I'm guessing if they could they would love to have that option this season as teams are preforming better against Barca current system of play which is not as successful as say last season. Now for Barca that might be losing 5 games a season and finishing second ecause of the amount of talent, resources and the environment they play in.

    I think were using Man Utd, Barcelona, EPL, because we both know those type of teams well. Personally i think the statistical analysis would be more effective in a weaker teams/league with budgetary constraints where slight advantages tend to mean more. Using your idea give me the management team/coach of my choice and I'll give you statistics and lets see who those better I don't that has been answered but at a level like MLS or England League Two I would certainly take that challenge.

    Sure i used the long ball method as a system because it's an easy situation to give an example about. But as shown in the questions i put up there is plenty of work done based on player skill set, athletic ability, playing formation, and opposition that would help a team use the best tactics possible against the opposition.

    I know Billy Beane attempted to use statistical analysis with the San Jose Earthquakes and to be honest I don't know much about what he took into account and what he's done but based on my MLS work I would say whatever he's tried to do there hasn't been any evidence that it made any difference. I applaud him for trying but I don't think he knows enough about soccer or the MLS to make difference and one of the reason is he doesn't have access to the information available to him as in Baseball.

    What confused me about your statement and I still don't fully understand is your perception of what sports statistics are. What is the "right" statistics in soccer or for that matter what are the right statistics in any sport.

    The reality of Moneyball is Oakland had the best pitching in Baseball in 2003 in Hudson, Zito, Mulder, Lilly, and Harden which the season before allowed 654 runs a game well they scored 800. They assumed their pitching would remain the same, which it did allowing 643 runs. So despite loosing two of their best hitters they used statistics to maintain the offense and win the pennant.

    So using a similar soccer scenario you have the best goalkeeping and defense in the league so you assume that will stay the same or get better but you need to replace the offense within a particular budget so you look at the available players and pick who you think will be an adequate replacement. My
    opinion being i rather have a statistical analysis of the candidates than just video tape and/or scouting reports.

    Now i guess the argument could be that the forward in soccer also plays a role in the defense on the team where in baseball it plays a lesser role. In someways I agree with that although statistically speaking this hasn't really been the case and soccer in general is a lot more positional based then maybe assumed. Although one of the problem in soccer is there is a lot more turnover and injuries than baseball so assuming you'll have a consistent lineup hardly ever is the case although again I would use statistical data and try to create depth that would take that into account
     
  15. SPA2TACU5

    SPA2TACU5 Member+

    Joined:
    Jul 27, 2001
    Very interesting.
    How do you deal with positioning. Do you track the exact movement of every player involved in those plays?

    Indeed, that's why I used Messi vs. tall striker (Zlatan, even though he's doesn't score a lot of headers).

    'In a weaker teams/league with budgetary constraints slight advantages tend to mean more'.
    Is that something that is backed up by statistics?

    That would be a lot of fun.

    I agree with some of this.
    Personally I think things like "formations" and "positions" are something of the past - I mean, they really never existed, but in the past at least it was easy to 'spot' them. So maybe we should talk about "roles", or just "tasks".
     
  16. ENB Sports

    ENB Sports Member

    Joined:
    Feb 5, 2007
    It's generally known within a set parameter statistics play a more important role because the risk of success/failure can be greater. Baseball example what makes Moneyball a story is the fact that Oakland had budget constraints, Statistics/Analytics have grown largely in Sports such as NFL, NBA and the NHL because of the Salary Cap system.

    A soccer example would be last year there was a case with Lincoln who were relegated from League Two to the Conference by 3 points during part of the season they had loanee youth goalkeeper from a Premiership side during this period the team 8-3-3 allowing 9 goals and the goalie Carson had 81% save percentage the rest of the season 5-5-22 with a 59% save percentage. I don't think at a top level one player can make such a huge difference.

    To answer your question about positioning the simple answer is yes and they've actually started using GPS to track every move a player makes on a pitch during the game. Actually some on Bigsoccer sent me a link to an American Company who are doing things with soccer analysis that you might enjoy - http://matchanalysis.com/mambo.htm (watch video). The problem is the cost I've met with both OPTA and ProZone and they say it's almost $200 per game to properly analyse a game which is even unrealistic for EPL but stupid for a League Two or College Soccer game.

    The point of my work and what you probably can see is to start simple and go from there. Like in your example of replacing the 10 goal scorer with the 16 goal scorer why can't we replace the 6 assist guy with the 12 assist guy or the 24 shots on target with the 50 shots on target guy, Now I think that is almost too much simplifying things and in my work there is more analysis then that, but that system is used in North American sports all the time.

    You response on formations is interesting as I'm noticing the opposite partly due greater analysis which explained in the book - Inverting the Pyramid: The History of Football Tactics. Basically the idea is you have specialists on the pitch like a defensive midfielder or a winger and they only play that role in a set formation. Examples of that is Toronto FC. Players such as Johnson and Plata wouldn't have a role in a normal 11 but within a 4-3-3 system can play the winger and be useful. Problem being 4-3-3 opens the counter attack on the wing so my opinion (backed up with some stat work) is you need at least 65% ball control to be competitive using that formation.

    Like other sports as they develop analytically, i see soccer becoming more a specialist sport where players will fit into a system or style as there isn't enough World Class players to go around. Stoke/Blackpool/Swansea are good examples of that over the past few years in the EPL although Bolton and Wimbledon were taking advantage of that over a decade ago. In countries such as Italy (Serie B), France (National), Spain (Segunda) this has been around at a high level for a number of years among the lesser clubs which unfortunately it means the soccer is horrible to watch.

    Like the chess analogy - between two equal players a player should always play for a stale mate and hopefully in the process find a way to gain victory because it is always easier to defend then attack.
     
  17. kenntomasch

    kenntomasch Member+

    Joined:
    Sep 2, 1999
    Location:
    El desierto
    Club:
    FC Tampa Bay Rowdies
    Country:
    United States
    And yet, there's not a very strong correlation between an advantage in possession and team success. At least not in our league.

    Obviously, you can't score without the ball (unless there's a major gaffe by your opponent) and your opponent can't score without it, but the duration that you possess the ball or possessing it a certain amount more than your opposition doesn't appear to correlate with winning.

    Which seems counter-intuitive, but a goal counts the same whether you take two minutes building toward it or score it in 15 seconds on a counter-attack.

    Possession seems to be (or at least feels) more important when you have the lead, because it forces your opponent to expend energy and each second closer to the referee blowing his whistle puts you that much closer to winning. But from an overall perspective, it may be that the ability to possess the ball - which is such a mantra of many coaches - either doesn't correlate that strongly with actually winning, or does so in a way we aren't quite understanding just yet.
     
  18. SPA2TACU5

    SPA2TACU5 Member+

    Joined:
    Jul 27, 2001

Share This Page