Analytics and the Premier League

MiracleOfO2704

not AWOL
SoSH Member
Jul 12, 2005
9,576
The Island
So in the aftermath of the Rodgers firing, a couple of people had pointed out that English pundits were quick to say that the new man at Liverpool would need to be in charge of transfers on his own. In response, Dlew had this thought:
 
 
DLew On Roids said:
This is an extremely UK-centric position, though. Having a sporting director or director of football is the standard in virtually every other major league. The trick is to get the manager and the sporting director on the same page philosophically, as has usually been the case at a place like Barcelona since Cruyff established its current structure.

It's fair to say, though, that the director of football structure is about as popular among British pundits as herpes. Just like zonal marking, when a club using it fails, the guy on TV will always point the finger at it.
 
I didn't give it too much thought until this afternoon, when I caught a good chunk of TalkSport's evening show, Kick-Off. On it, Stan Collymore blamed "Moneyball" for a lot of Liverpool's woes, in particular, that big on-field decisions were being left to (paraphrasing) "men in suits with business degrees getting on their laptops and playing FIFA 16." His thought was that, as much as Rodgers shared in the blame, the transfer committee was total rubbish and should be done away with, and the following changes be made:
 
-Henry/Werner are there only to sign checks.
-Klopp runs the show. He names his scouts and most of his staff, and they'll handle all team business on the pitch, with two exceptions:
   -Jamie Carragher is brought on as first team coach to bridge the gap between players and Klopp, and
   -Steven Gerrard is the Global Ambassador, sent to woo the big players to join Liverpool
 
My instinct is to dismiss a lot of this out of hand, as Collymore has, on numerous occasions, sounded like a stodgy ex-player longing for the good old days before the oil money came in. Then he made an observation that got my attention. He said that Moneyball was borne out of a completely different economic environment in a completely different sport in a completely different country. And by and large, he's right. Had it been just the last part, I'd have dismissed it as his usual wishcasting for the days before kit sponsors, but when you think about it, the clubs that win the big leagues of Europe on a fairly consistent basis aren't really looking to exploit market weakness. Their only concern is identifying a superstar player that will makes the team better, figuring out what their rivals' highest possible bid will be, and finding out if they can beat it. Just as different, as well, are the consequences to mistakes. The Athletics, for instance, whiffed badly this offseason, particularly in the Lawrie-for-Donaldson deal. As disastrous as it was for Oakland this year, they're still in MLB, and they'll get a top-5 pick in the amateur draft for their troubles. If Crystal Palace were to make that big a mistake with their team, their modest goal of maybe qualifying for the Europa League is reduced to fighting off relegation, losing PL money, and having to compete with strong Championship clubs for one of the few spots back to the PL.
 
So the big question is, for bigger clubs in England (the Mancs, Liverpool, Arsenal, Chelsea, Tottenham), can a Moneyball approach work, or do you just need to have the money to outbid everyone?
 

singaporesoxfan

Well-Known Member
Lifetime Member
SoSH Member
Jul 21, 2004
11,890
Washington, DC
It isn't inherently impossible to make a sporting director - coach combination work, since as DLew rightly pointed out it works in other football leagues. The open question for me is what the "short term vs long term" tradeoff looks like in a sport with relegation. Because of relegation, you obviously can't do in the EPL what the Astros or the 76ers have tried to do in their respective leagues, which is play poorly and rebuild through the draft. So you lose at least some of the benefit of having separate roles that look at different time horizons. However,  the dual structure still allows people to specialize in roles and skills that aren't necessarily the same: coaching tactics vs identifying talent to be brought in from elsewhere. Obviously it's easier for the sporting director and coach to be on the same page when they are the same person (hello Bill Belichick!) but that creates some tradeoffs that your team will have to accept or have to supplement appropriately.
 

Mighty Joe Young

The North remembers
SoSH Member
Sep 14, 2002
8,466
Halifax, Nova Scotia , Canada
Interesting topic. I didn't see the interview so it's rather difficult to really critique Collimore's comments,. But there seems to be some fundamental misconceptions here.

MoneyBall philosophy was simply about finding market inefficiencies .. That is players who were undervalued because GMs failed to understand the importance of stuff like OBP .. Or any other undervalued trait.

Liverpool, on the other hand seemed to be targeting young talented players who would expect to improve (Luis Suarez being Exhibit A) And conversely avoiding expensive , older players on the decline .. Of which examples abound. They did this because Henry couldn't compete with the Oil Sheiks .. In fact he stated this publicly when he bought the team.

So it has nothing to do with MoneyBall .. But that being said, to compete at the highest level consistently you pretty much have to spend accordingly .. In transfers AND in wages. And the relegation factor is huge .. you can't build a team by being bad.

As for using a MoneyBall philosophy in world football .. I'm not sure the existing state of statistical analysis is rigorous or detailed enough to garner any hidden insights. From what I've seen of it, they seem even cruder than hockey stats .. Which are still pretty crude.
Corsi be dammed.
 

Stanley Steamer

Member
SoSH Member
Jan 11, 2012
1,439
Rossland, BC
I think in the end you have to have the money, in order to ensure that any dip in form from your team doesn't become longstanding.
That said, a lot comes from having a good corps of players led by a tactically strong, usually younger coach.  Once a group like that is in place, wins enough to compete in Europe, and has at least adequate financial resources to back them up, then you have the ingredients to succeed.  It obviously didn't happen just recently for Liverpool, as losing Suarez and playing more matches brought them back down to average level.  But if they get Klopp, and he manages to get them back in Europe, then I think FSG will support them well.
And, just like Fenway, FSG are renovating and expanding Anfield.  That will lead to higher revenues going forward, strengthening their ability to compete.  I think they are savvy business owners, even if they are in a new environment.  The mere fact that they aren't hated by their own fans, like most American owners seem to be, means they must be doing something right.
I hope they get it right.  Having a strong Liverpool team is good for the BPL, and for Europe too.
 

JBJ_HOF

Member
SoSH Member
Apr 5, 2014
540
Stewart Downing was considered a huge anaylitical signing.
 
European soccer has player tracking and much much more analitical infomation than hockey.
 
The book Soccernomics has a bunch of cool info.
 

Mighty Joe Young

The North remembers
SoSH Member
Sep 14, 2002
8,466
Halifax, Nova Scotia , Canada
JBJ_HOF said:
Stewart Downing was considered a huge anaylitical signing.
 
European soccer has player tracking and much much more analitical infomation than hockey.
 
The book Soccernomics has a bunch of cool info.
Oh .. I know about that .. Have read Soccernomics .. It still strikes me as pretty crude stuff .. Probably akin to the early rumblings of sabrmetrics thirty years ago. In fact I think Bill James stuff was far more advanced in that he introduced the idea of synthesizing all the various bits of traditional stats into a single number representing a players offensive value (Runs Created/9)

I'm not sure you can even do that with soccer/football stats .. Much like hockey's statistical "revolution" it's too fluid a game to numerically quantify a players value.
 

Billy R Ford

douchebag q momfingerer
SoSH Member
Feb 10, 2010
876
Northeastern
A general point: inevitably when Moneyball comes up there are two different ways of making signings that are both conflated and described as the "Moneyball philosophy":
 
Moneyball philosophy #1: using advanced statistics to scout and sign players "right out of the computer"
 
Moneyball philosophy #2: exploiting market inefficiencies to get the most "bang for your buck" and build a winning team on a shoe-string budget
 
The Oakland A's of the early 2000s obviously did both; they used the advanced stats in order to uncover the market inefficiencies, which is why the two methods are treated as one and the same.
 
JBJ_HOF pointed out the perfect example of how they aren't however with Stewart Downing. Downing was one of the league leaders in chances created his last season at Villa and as such was a big "advanced stats" signing. He's also English and thus came with an over-inflated price tag; English players are the definition of a "market inefficiency," showing Liverpool weren't exploiting the market, but playing into it. Jordan Henderson is another good example- one of the league leaders in chances created in 2010-11 with Sunderland (advanced stats signing) while also a young English midfielder (very inefficient market). (And that doesn't make Henderson a "bad" signing - he was a great signing - just not an "efficient" signing).
 
To the point of the thread, I agree with BCs MJY: regarding philosophy #1, soccer stats are a little too rudimentary to give teams a huge edge right now. They can provide some insight, but aren't really at a point yet where they reveal that much more about the game than what a trained eye can see. Regarding philosophy #2, Liverpool simply haven't been doing that. What was "inefficient" about the markets for Firmino, Benteke, Clyne, Lallana, Lovren, or Balotelli?
 
Appropriately enough, to Collymore's point I think going forward a Director of Football can only help in identifying market inefficiencies, targeting the right players to exploit those inefficiencies, etc.
 

dirtynine

Member
SoSH Member
Dec 17, 2002
8,437
Philly
There's the interesting case of Midtjylland in Denmark – they're a small club having success leaving many decisions up to statistical analysis.  This is worlds away from big-money, high-stakes EPL stuff; the gap between the "big" and modest clubs in the Denmark pyramid are probably small enough that finding market inefficiencies can make a decent impact.  Not sure it would transfer.  But still, it's a cool approach and story. 
 

Morgan's Magic Snowplow

Member
SoSH Member
Jul 2, 2006
22,444
Philadelphia
Soccer analytics seem much more useful in making tactical decisions and coaching the players you have than in acquiring new players. Guardiola is a huge believer in tactical analytics and I know a lot of other younger coaches are as well.

The problem with using analytics for player acquisition is that context matters so much and it's very easy to read into the data what you want to see (ie, the Downing example). If you look at the clubs that have been most successful in the transfer market in recent years, and that have used this success to really compete with the richer clubs, their secret seems to be successful scouting more than anything else. I'm thinking of Juve, Atleti, BvB, etc.
 

Nick Kaufman

protector of human kind from spoilers
Lifetime Member
SoSH Member
Aug 2, 2003
13,447
A Lost Time
At its core, every team should try maximizing revenue and maximizing resources.
 
There are many problems and potential mines.
 
- As was mentioned, you not only have the issue of relegation, but the whole tanking strategy has no meaning when there's no draft.
 
- As was mentioned not only are stats crude, I think there is a problem of small sample size. In most leagues, there are only 30-40 games played per season and in some categories the numbers are miniscule. Like, you can be the leader for a season with 12 assists to goal for example. Who knows if that's illustrative of the player's talent or a statistical fluke? There's a ton of players who have wonderful seasons and then their performance just drops of the following season for no apparent reason.
 
- Even if teams find market inefficiencies, they really can't take full advantage of them because of the way player contracts work. Liverpool find Suarez for example, but they can't keep him because he wants to play elsewhere. And that happens with a lot of young players. As soon as they hit it big, no matter what the contract says, they can renegotiate it or demand a transfer. I don't think that happens in American sports as much, both because of the way contracts are structured and because the rules give players incentives to keep on being signed by their current teams.
 
- Last but not least, even if you do find a market inefficiency, sooner or later, the rest of the league imitates you and takes away the advantage.That's what happened with the Oakland As to a large extent I think. And I don't think there's a guarantee that if you come up with one sort of innovation, you will keep producing another type of innovation. I sense that a lot with coaches who come up with a specific way to play. Like let's say Arrigo Sacchi comes up with a fluid way to play 4-4-2. Soon enough, major clubs across Europe imitate his style of play. A bit of time passes by and Sacchi can't reproduce his success. And so on and so on.
 
All of the above factors over the long run tilts results in favor of the teams with most resources, especially since European leagues do virtually nothing to promote parity. Which means that to a large extent results in soccer competitions are a product of club resources. This btw is proof that overall, player evaluation with or without stats does work otherwise, results would be random with no correlation to how much money each club spends.
 

soxfan121

JAG
Lifetime Member
SoSH Member
Dec 22, 2002
23,043
dirtynine said:
There's the interesting case of Midtjylland in Denmark – they're a small club having success leaving many decisions up to statistical analysis.  This is worlds away from big-money, high-stakes EPL stuff; the gap between the "big" and modest clubs in the Denmark pyramid are probably small enough that finding market inefficiencies can make a decent impact.  Not sure it would transfer.  But still, it's a cool approach and story. 
 
Matthew Benham also owns Brentford FC (20th in the Championship at this date) and uses these methods there.