The Bill James of Soccer
November 6, 2011 4:42 PM Subscribe
Who is the Bill James of soccer? Looking for good soccer statistics analysis that is beyond what others are doing and is considered groundbreaking.
Hardly current, but you might be interested in the system Valeriy Lobanovskiy used at Dynamo Kiev. I seem to recall Jonathan Wilson mentioning in Inverting the Pyramid somebody in Britain (who's probably rather famous, so I'm probably embarrassing myself not knowing who) trying to tally statistics in the early 20th century.
To be honest, I've always assumed soccer doesn't lend itself well to sabermetrics-like analysis because it's not discrete like baseball is. The fad in soccer analysis right now seems to be stuff like Zonal Marking.
posted by hoyland at 5:00 PM on November 6, 2011 [1 favorite]
To be honest, I've always assumed soccer doesn't lend itself well to sabermetrics-like analysis because it's not discrete like baseball is. The fad in soccer analysis right now seems to be stuff like Zonal Marking.
posted by hoyland at 5:00 PM on November 6, 2011 [1 favorite]
I'm sentimentally with hoyland, like Zonal Marking a lot, and will point to Jonathan Wilson's weekly column for the Guardian.
That said: Simon Kupar's recent FT piece on the quant revolution, and clubs' attempts to refine data collection and statistical models, points to the direction the game is going. According to Kuper, a lot of the newest data and models are proprietary, so you're not as likely to find third-party online discussion.
posted by holgate at 5:05 PM on November 6, 2011 [1 favorite]
That said: Simon Kupar's recent FT piece on the quant revolution, and clubs' attempts to refine data collection and statistical models, points to the direction the game is going. According to Kuper, a lot of the newest data and models are proprietary, so you're not as likely to find third-party online discussion.
posted by holgate at 5:05 PM on November 6, 2011 [1 favorite]
The new ownership/management of Liverpool is allegedly using statistics/etc to pick up/get rid of players. (That said, I'm not convinced it's not a combination of statistics and darts.)
That said, holgate recommends Wilson's weekly column, but I'd also suggest tweeting him this question.
posted by inigo2 at 5:41 PM on November 6, 2011
That said, holgate recommends Wilson's weekly column, but I'd also suggest tweeting him this question.
posted by inigo2 at 5:41 PM on November 6, 2011
hoyland: who's probably rather famous, so I'm probably embarrassing myself not knowing who
You're probably thinking of Charles Reep, and it's a name that's not terribly embarrassing to forget. He's been forgotten by most. That said, he seems like an interesting figure. I suspect Marcelo Bielsa, the fairly radical coach of Athletic Bilbao (he also coached Chile at the last World Cup), has studied Reep.
To echo hoyland, Jonathan Wilson is the English language gold standard in statistical analysis, but there are quite a few others, such as Michael Cox of Zonal Marking.
Baseball is more amenable to statistics than soccer. It is higher scoring for one. In soccer a team can completely dominate every single stat (e.g. possession, shots on goal, corners) but be beaten by a team that only manages to get on chance, which they manage to capitalize on. Freak results are fairly common (e.g. Blackburn Rovers beat Arsenal 4-3 earlier this season despite only managing two shots on goal, Arsenal scored two own goals). Not that stats are meaningless, but some things are somewhat hard to quantify.
Soccer is a more chaotic system than baseball (here's a diagram of a move that led to a goal scored by Spain against Scotland last month). You can break baseball down into fairly discrete units, which is harder with soccer. Sometimes all 22 players are affecting play at the same time, and usually it's at least 10-15.
posted by Kattullus at 6:11 PM on November 6, 2011 [1 favorite]
You're probably thinking of Charles Reep, and it's a name that's not terribly embarrassing to forget. He's been forgotten by most. That said, he seems like an interesting figure. I suspect Marcelo Bielsa, the fairly radical coach of Athletic Bilbao (he also coached Chile at the last World Cup), has studied Reep.
To echo hoyland, Jonathan Wilson is the English language gold standard in statistical analysis, but there are quite a few others, such as Michael Cox of Zonal Marking.
Baseball is more amenable to statistics than soccer. It is higher scoring for one. In soccer a team can completely dominate every single stat (e.g. possession, shots on goal, corners) but be beaten by a team that only manages to get on chance, which they manage to capitalize on. Freak results are fairly common (e.g. Blackburn Rovers beat Arsenal 4-3 earlier this season despite only managing two shots on goal, Arsenal scored two own goals). Not that stats are meaningless, but some things are somewhat hard to quantify.
Soccer is a more chaotic system than baseball (here's a diagram of a move that led to a goal scored by Spain against Scotland last month). You can break baseball down into fairly discrete units, which is harder with soccer. Sometimes all 22 players are affecting play at the same time, and usually it's at least 10-15.
posted by Kattullus at 6:11 PM on November 6, 2011 [1 favorite]
Oh, just to be clear, I'm not saying stats are useless in soccer, they can tell you quite a number of important things, but some things are really hard to quantify (e.g. how do you quantify how a defender denies the attacking team space). This is something humans can grasp quite easily, but turning it into a measurable stat is quite tricky.
posted by Kattullus at 6:22 PM on November 6, 2011
posted by Kattullus at 6:22 PM on November 6, 2011
some things are really hard to quantify (e.g. how do you quantify how a defender denies the attacking team space).
That's where Kuper's piece is interesting, pointing to a move towards second- and third-order data that vindicates much of what our guts told us when Opta first started pushing out the kind of stats that American commentators seem to adore:
posted by holgate at 6:55 PM on November 6, 2011
That's where Kuper's piece is interesting, pointing to a move towards second- and third-order data that vindicates much of what our guts told us when Opta first started pushing out the kind of stats that American commentators seem to adore:
Tackles seemed a poor indicator too. There was the awkward issue of the great Italian defender Paolo Maldini. “He made one tackle every two games,” Forde noted ruefully. Maldini positioned himself so well that he didn’t need to tackle. That rather argued against judging defenders on their number of tackles, the way Ferguson had when he sold Stam.So I think the question's based on an entirely valid premise: we're at the point where, like James, the people at the top end of the game are looking for new data sources and derivatives of existing data that have statistically significant correlations with performance. However, I don't think we're at the point where those sources and indexes have entered the public sphere, and I think it'll take a good while before they do so.
posted by holgate at 6:55 PM on November 6, 2011
Doesn't Wilson pretty fairly savage Reep's "analysis" in inverting the pyramid.
Also collecting data is one thing, using that data to show statistically valid non obvious relationships is quite another.
posted by JPD at 6:59 PM on November 6, 2011
Also collecting data is one thing, using that data to show statistically valid non obvious relationships is quite another.
posted by JPD at 6:59 PM on November 6, 2011
It's worth noting that on the Guardian Football Weekly they asked Michael Cox about whether teams were interested in what he was doing. The answer seemed to be not really.
What Opta collect in stats and sell to teams and commentators is deeper.
You can occasionally find the preparation sheets that players in the major European teams get. They also apparently have people who look in detail at how other teams typically play. In Phillip Lahm's recently published autobiography he said that the preparation for Germany (one of the best organised national teams) was poor compared to his club team Bayern Muenchen. So the big European club teams are presumably doing sabrenomics type stuff.
There are some teams that beat the system in that they perform better than the money they spend indicates they should. In the book Soccernomics by Simon Kuper Arsenal and Olympic Lyon and said to be some of those teams.
posted by sien at 7:29 PM on November 6, 2011
What Opta collect in stats and sell to teams and commentators is deeper.
You can occasionally find the preparation sheets that players in the major European teams get. They also apparently have people who look in detail at how other teams typically play. In Phillip Lahm's recently published autobiography he said that the preparation for Germany (one of the best organised national teams) was poor compared to his club team Bayern Muenchen. So the big European club teams are presumably doing sabrenomics type stuff.
There are some teams that beat the system in that they perform better than the money they spend indicates they should. In the book Soccernomics by Simon Kuper Arsenal and Olympic Lyon and said to be some of those teams.
posted by sien at 7:29 PM on November 6, 2011
Ooph... I was finishing a huge comment and accidentally hit ctrl+r instead of ctrl+t...
Anyway... it's so late it's early in my part of the world, so I'll do a quick run through of the main points.
Reep is considered severely outmoded now, but people seem to have been taking a second look at his work lately. Wilson, for instance, has been kinder to him in recent years, this article on possession is a good example.
Kuper's article in the Financial Times is a hype-fest and many of its examples don't really tell you much. Wenger had been following Henry since Henry's teens as Henry had been a youth player at Monaco while Wenger was the manager there. Makélélé wasn't sold by Real Madrid because they hadn't done any statistical analysis (Pérez' quote makes it clear they had), he was sold because he didn't fit Real's style of play. He fit Chelsea's style perfectly, however. And Carroll is starting to look perilously like a flop (though the matter is far from settled and I hope he will turn around). Oh, and Lobanovskiy was a great coach, a true wonder, but he never did manage to win any of the three most important trophies, the World Cup, the European Championship or the European Cup (now Champions League). Winning the European Cup Winners' Cup (now Europa League) was no small achievement, but it's not one of the big ones.
Finally, there are counter-examples that indicate other paths. The most powerful is total football, the footballing philosophy developed in Holland in the 60s and 70s, most famously at Ajax but also at Feyenoord and some other clubs. This was in great part due to a new management culture that favored democratic principles. Having the players all contribute their qualitative insights increase the understanding between players and gave the managers perspectives that helped them change the way they saw the game. This was not stat driven at all and took place at around the same time Lobanovskiy was doing his thing in the Soviet Union. Dynamo Kiyv may have become a regional powerhouse and a strong contender on the European scene, but Ajax won the European Cup three years running (and Feynoord once). Before the emergence of total football the Netherlands were far from a soccer powerhouse, if anything they were something of a backwater.
Oh, and one last note... philosophy is a third order of management that not many managers actually operate on. Tactics is how you arrange your players on the pitch and the instructions you give them. Strategy is how you marshall the resources at your disposal over the course of a season. Philosophy is usually about achieving a certain kind of style and making sure that the entire organization of your team (from youth coaches to chefs to fans) works towards developing the team towards the playing style you favor. There are some major schools of soccer philosophy, total football being the most famous and most successful, and it has a number of variants (e.g. tiki taka). Long ball is the style Reep argued for, Bielsa has developed vertical football and in Italy there's long been an emphasis on defensive football (catenaccio was a particular style that is now very outmoded, but it led to various more subtle and advanced philosophies).
posted by Kattullus at 9:31 PM on November 6, 2011 [1 favorite]
Anyway... it's so late it's early in my part of the world, so I'll do a quick run through of the main points.
Reep is considered severely outmoded now, but people seem to have been taking a second look at his work lately. Wilson, for instance, has been kinder to him in recent years, this article on possession is a good example.
Kuper's article in the Financial Times is a hype-fest and many of its examples don't really tell you much. Wenger had been following Henry since Henry's teens as Henry had been a youth player at Monaco while Wenger was the manager there. Makélélé wasn't sold by Real Madrid because they hadn't done any statistical analysis (Pérez' quote makes it clear they had), he was sold because he didn't fit Real's style of play. He fit Chelsea's style perfectly, however. And Carroll is starting to look perilously like a flop (though the matter is far from settled and I hope he will turn around). Oh, and Lobanovskiy was a great coach, a true wonder, but he never did manage to win any of the three most important trophies, the World Cup, the European Championship or the European Cup (now Champions League). Winning the European Cup Winners' Cup (now Europa League) was no small achievement, but it's not one of the big ones.
Finally, there are counter-examples that indicate other paths. The most powerful is total football, the footballing philosophy developed in Holland in the 60s and 70s, most famously at Ajax but also at Feyenoord and some other clubs. This was in great part due to a new management culture that favored democratic principles. Having the players all contribute their qualitative insights increase the understanding between players and gave the managers perspectives that helped them change the way they saw the game. This was not stat driven at all and took place at around the same time Lobanovskiy was doing his thing in the Soviet Union. Dynamo Kiyv may have become a regional powerhouse and a strong contender on the European scene, but Ajax won the European Cup three years running (and Feynoord once). Before the emergence of total football the Netherlands were far from a soccer powerhouse, if anything they were something of a backwater.
Oh, and one last note... philosophy is a third order of management that not many managers actually operate on. Tactics is how you arrange your players on the pitch and the instructions you give them. Strategy is how you marshall the resources at your disposal over the course of a season. Philosophy is usually about achieving a certain kind of style and making sure that the entire organization of your team (from youth coaches to chefs to fans) works towards developing the team towards the playing style you favor. There are some major schools of soccer philosophy, total football being the most famous and most successful, and it has a number of variants (e.g. tiki taka). Long ball is the style Reep argued for, Bielsa has developed vertical football and in Italy there's long been an emphasis on defensive football (catenaccio was a particular style that is now very outmoded, but it led to various more subtle and advanced philosophies).
posted by Kattullus at 9:31 PM on November 6, 2011 [1 favorite]
I definately agree with Katullus re: the hype machine surrounding statistical analysis. Even basketball which suffers from similar issues of continuity and dependence on multiple players at once, but does have a decent volume of historic data to use in analysis has yet to be cracked by sabre folks.
And as for "the good models" being private, well there are tons of econometricians and statisticians who are also die-hard sports fans - especially basketball given the big role it plays on University campuses in the US. If it was an easily solvable problem they'd be punching out papers for fun left and right.
posted by JPD at 6:16 AM on November 7, 2011
And as for "the good models" being private, well there are tons of econometricians and statisticians who are also die-hard sports fans - especially basketball given the big role it plays on University campuses in the US. If it was an easily solvable problem they'd be punching out papers for fun left and right.
posted by JPD at 6:16 AM on November 7, 2011
A Beautiful Numbers Game, Soccer By The Numbers and On Football are three metrics-focused football blogs. There's a bit more MLS than I would like, but they could be what you're looking for. You can also check out their blog rolls for other similar blogs. I can't vouch for how useful/ rigorous their investigations are, but they're definitely intriguing.
posted by WalterMitty at 1:28 AM on November 11, 2011 [2 favorites]
posted by WalterMitty at 1:28 AM on November 11, 2011 [2 favorites]
This thread is closed to new comments.
posted by milkrate at 4:58 PM on November 6, 2011 [1 favorite]