³ÉÈË¿ìÊÖ

Please. No more misunderstanding the entire concept of Moneyball. It's not simply about use of statistics.
The ethos is about innovation and gaining the slightest edge on your competition.
Billy Beane saw Sabermetrics as a way to gain that edge. Now other baseball teams use the same techniques to evaluate player performance, that advantage has been lost.
It could be argued that the trend for sports psychologists a few years ago gave some athletes and teams a similar advantage.
Similarly, British Cycling has always pushed the boundaries of what was legal concerning modifications to their equipment in an attempt to trim hundredths of seconds from their times.
It's all about finding the next new way to gain the advantage. Athletes and teams have done this for years. Sometimes legally, sometimes not but, only going into this century did someone put a nametag on the concept.

A few other quick points need clarified.
By most peoples standards the Red Sox didn't have a terrible 2011 season. Missing the playoffs by a single win is sickening but, not terrible.
On the contrary, Oakland haven't shown significantly more success during Beane's tenure.
3 playoff appearances since 2002 when sabermetrics were adopted and none since 2006 (and no winning records). 3 playoffs in the 10 previous years, 3 in the 10 years before that and 5 during the 1970's. Keeping in mind there were less playoff spots available prior to 1981. Not a great advert for Moneyball.
I can only hope the Moneyball movie shows Billy Beane for being a general manager that has showed bravery to make unfashionable decisions and, more importantly, has remained loyal to the Oakland Athletics and his family.

Would have been much improved with a mention of Bill James, whose work inspired Billy Beane to build his machine to see what it would do. (Sorry.)
Not so sure about essentially team games (baseball being essentially a collection of individuals) where character and heart are so important, but anyone betting on golf would be wise to look very closely at the statistical work available now from the PGA Tour and overlay that with knowledge of players, greens, schedules etc.
Many of the stat's surprise even the players as they analyze relative strengths and weaknesses in their game.

knucklonian,
The film is a little more respectful of reality, more contextual, than this blog might suggest. But agree that Slater is trying to convey applicability to other sports.

Comments about 2011 Red Sox spot on, but it does lead one to question Henry and the value of trying to transport Bill James and Epstein's work and try to make it relevant to LFC. He'd be better off devoting his talents trying to squeeze all the revenue he can out of Anfield, in the same way he did at Fenway.

#1: Sounds just like Formula 1, where every team every year tries to push the limits of the rules to gain the slightest advantage. We're talking a tenth of a second over three miles of circuit, but if it works they keep it and look smug while everyone else tries to catch up.

I wonder how long it'll be before some statistician tries to quantify team spirit and good centre-back partnerships and the psychological impact of the crowd's chanting and whatever else. I suspect not long.

There are obvious stats that are overrated, when people like Beane and Jose Mourinho have found the real reasons players/team have succeed or failed. I know this isn't player related but team possessions stats seem to be given too much importance.

i.e. England v Spain recently, if you're constantly passing backwards you are not going to create chances but everyone says "look at the possession stats".

Alan Hansen has got a lot to answer for on MOTD saying "possession is king"

Possession is king, in as much as you cannot score if you don't have the ball. However, the number of times that teams with superior possession lose shows it is indeed an overrated statistic. Perhaps distinguishing forward, sideways and backward passes would provide some fresh insights. Another point that I think is misinterpreted is goalscoring. Obviously, if a player is sticking away 20+ goals a season this is 'a good thing', but such key players can often be marked out of the game, and then they move out of the area, in search of the ball, and often become providers. I bring this up because a number of famed strikers have been the target of intense criticism for their low scoring rate, when they have simultaneously been setting up plenty of goals for their colleagues. As a team game, it's not all about who scores, but that the team scores more than the opposition.

Havent we already had a blog on this??

Interesting work Matt.
Strange to think that football has been relatively slow on the uptake given the slight margins involved and the relative success of 'manufactured' managers like Mourinho, Big Sam and Tony Pulis.

Possession is Queen
Scoreline is King ;-)

"Playing the percentages" an old fashioned term and proof that moneyballing has existed in English football for decades.

Also it's a well known theory that managers like to keep possession as you lose more energy running after players trying to win it back than taking two touches to pass it around.

Also doesn't guardiola promote a technique whereby players have no more than 3 touches of the ball based on statistical analysis of moves breaking down when players took more?

Also doesn't guardiola instruct players to press the ball for 3 seconds immediately after losing it as statistically this is the best time to win it back?

Baseball is essentially an individual sport rather than a team sport. If I’m pitching, or batting, it doesn’t matter what my team mates are doing as it doesn’t affect my performance. The chances of getting a hit are dependent just on pitcher and batter (more or less).

In football, this is not true. If I’m a forward it depends a lot what my team mates are doing to influence my chances of scoring. If I play alongside the best central defender in the world, he would look rubbish as our partnership would be very poor (all due to my fault).

Therefore, in baseball it’s simple to say if you put a good pitcher in another team, they will pitch as well as they did previously. But it doesn’t stand that if you put a good footballer in another team they will perform the same way.

Similarly, in baseball the tactic is the same in every game. In football, a defender for Stoke plays entirely differently to one for Man Utd. Players perform different roles in different sides, so can’t be as easily converted.

Football isn’t about getting the best performers, but getting the right balance and having players who compliment each other, regardless of their ability. This is why we frequently see the lesser ability teams beating the better sides, simply because their quality as a side is stronger.

This method will work, to some extent in football, but the great interdependence and complexity means there will be far, far less certainty in what’s achieved.

I can’t help but think despite all the work that would go on behind the scenes to make this system work, it wouldn’t be as effective as having the autocratic rule of a wiley old manager who knows the game. And if you have one of those, you’re more interested in gut feelings than statistical outputs.

The statistical analysis of football is mindblowing.
tackles, headers, passes, crosses, shots
number attempted, succeded, % succeded, % succeded in key moment
defender, def mid, att mid, attacker
wide player, central player
lone attacker, joint attacker
wide player who crosses, wide player who dribbles
age of player, number of injuries
possesion holding defensive tactics, possesion holding attacking tactics
possesion concedeing defensive tactics, possesion coceding attacking tacics (counter)
442, 451, 433, 343, 532
number of yellow/red cards
that's just off the top of my head and I don't really know anything about foorball other then I played championship manager 2001/2 for 15 seasons

#12:
Ah ha, Didn't see the instructions to open every link. Assumed you wrote the blog yourself. To describe Bill James as an "obseesive amateur" is somewhat misleading - he was a professional in most senses of the word when Billy Beane was still in the minors. I have much of his material to prove it. Plus he was hired by the Red Sox as a consultant to assist Henry and Epstein.
Still think Henry is barking up the wrong tree; the point about Fenway is very apropos.

actually, the management computer game is a good comparison. As it reduces football purely to statistics, but the flaws become very clear when playing. The main difference being the greater lack of variability in performance in the games. As in life there are infinite things that can affect how you play, in a game you will only do what you are programed to and so tend to have a much greater correlation of performance to ability

Thanks for the blog Matt.

One or two points, especially one of your replies............
I'm not surprised that football has been relatively slow to embrace this stuff at all. A resistance to change/innovation is the default position for the sport going back decades. Things have only ever moved forward when somebody has been brave enough to try something and done well. Success persuades!

Matt, not everything in life needs to change. Maybe the question, these other sports need to ask and research is, why is football so poular worldwide, while traditional US based sports, truly lack worldwide appeal?

Stats in any form, whether it be business or sport, are used to eliminate the chance aspect when making a decision. Yes we all want to make the correct decision but in many parts of life the chance aspect is what makes life interesting. Remove it totally, you will end up with sterility. Sport need to be about chance.

Stats have a place but if player choice is based totally on stats, for each position, you will have exactly the same type of players. If the actual game is stats driven, you will eventually end up with every game being played in the same manner because the players will dictate that, by coaches having to play to their strengths.

I am not against stats but only when used in moderation.

The majority of people hear or read something and they buy into it because it is natural to only look at the positives. That is why so many new ideas do not go to plan. When you read of a new idea the first thought should be, what is wrong with it and do I really need it.

Yes football has been slow to adopt new ideas but have you thought, that is exactly why it has a unique appeal. The topic of technology in football, is a good example, yes football has been slow to adopt it but it may be that, they want to get it right first time before adopting it.

I'm glad the book and statistical decision making are finally getting the attention they deserve in football.

I don't however, buy the argument that football is too different a sport for these techniques to be applied. Yes, it's important to identify the structural differences in the sport and its regulation, i.e. the absence of a salary cap, or football's fluidity. But these factors simply inform the application of the techniques rather than prohibit them.

Just as with baseball (as highlighted in the book), there were naysayers, who said 'we've got offensive statistics, but you can't do the same for defense - such as whether an outfield player (on defense) should have caught a ball.' Well, with time, these people were proven wrong, and more sophisticated techniques developed to produce usable statistics from the data, such as player positioning and speed to produce the needed answers. Similarly, someone will develop the same stats for football, particularly as computing power is more readily accessible, and improved network analysis techniques become easier to execute. Perhaps measuring a football player's time in space, or how much a striker drags defenders toward him when not in possession will give meaning to data that simply isn't properly measured now. Combine that with more conventional data on chance creation, etc. and someone will crack the football code. Let's just hope it's Arsenal!

Baseball has had a wealth of stats for over 100 years. Football has had prozone, etc for 10 ? This data opens up the game for analysis. There may not be a knockout stat in football, but that underestimates what moneyball says. It references others aspects of the game, such as pitchers being discarded as poor because they don't get strikeouts, their pitching action is odd or a player lacks is deemed to lack character. It talks about how vague fielding stats are. It says low payroll teams can compete by identifying aspects of the game undervalued by the open market. Wimbledon did something similar in the 80's with the long ball game. Stoke have done it by valuing the set piece.

Moneyball opened my eyes to how naively I watched football. The british mainstream media carries little in depth analysis of the game as this is perceived as boring. Therefore fans prefer to make vague judgements of players based on perceived effort rather than technical skill or actual fitness. This is exactly the culture of baseball scouting that money ball attacked.

Also, lots of people are missing the point about moneyball - the analysis identified which parts of baseball the open market undervalued and then applied that to stats on players rising through the US college system. Beane built a team from the bottom up - he didn't have the money to buy ready made pro players and the US system is different to ours. He had to select them as amateurs and develop them.

A moneyball analysis of football may be limited because the youth development of football is so different and detailed stats about youth football just aren't recorded. So, even if you find the undervalued stats you can only apply them to kids in your own youth development system.

A moneyball example might be that far fewer children born at the end of the academic year become professional footballers as they can be physically dominated out of games in their early developing years. Prozone would not identify that as football stats are mainly limited to the score. (Also perhaps it is one of the reasons Messi would not appear out of the British coaching system).

Snodot @19

Combine that with more conventional data on chance creation, etc. and someone will crack the football code. Let's just hope it's Arsenal!

--------------------------------------------------------------------------------

You have just summed up the reason why stats should not be the number one tool for coaches football.

Once you crack a code there is nothing left to aim for. Football is a great game because the it has the element of chance, it is unpredictable. The use of stats is designed purely to eliminate the chance element and remove the unpredictable nature of a player or team.

Is that what you really want from a game.

The beauty of football is that we all have something to talk about, a poor decision by a referee, flash of brilliance, a great save. Great saves can't be made if the goalkeeper is perfectly positioned because he too, has analysed the opposition, who can only play one way by design.

Dogstar @20
Moneyball opened my eyes to how naively I watched football. The british mainstream media carries little in depth analysis of the game as this is perceived as boring. Therefore fans prefer to make vague judgements of players based on perceived effort rather than technical skill or actual fitness. This is exactly the culture of baseball scouting that money ball attacked.

--------------------------------------------------------------------------------

Contrary to your belief there are many fans in the UK who have far more knowledge than you think.

I wonder if you can tall me the actual point of Moneyball because we seem to be floating around value of players, technique, physical ability and everything else under the sun.

Your comment regarding Messi, 'not appearing' as you put it, is wrong because you do not know if he would have or not.

You are confusing analysis and stats, we have always analysed in the UK and used some stats, it is not new. The difference is we place more reliance on what our eyes tell us.

#20: DogStar,
You may also be missing the point. Moneyball is not primarily about players coming out of College, it is about putting a statistical spin of attributes of established pros like Scott Hattieburg and Chad Bradford which were undervalued.
The other point (which both the film AND Slater's blog conveniently gloss over) about the Oakland A's teams of the Moneyball era is that very little of their success would have been possible without customarily good years from All Stars such as Eric Chavez and Tejada, plus a fine rotation.

@21
Stats haven't precluded randomness in baseball, nor will they in football. I never enjoyed baseball to start, but it hasn't become any more boring because of the spread of Moneyball techniques. In fact, as the statistical playing field levels, tactics become far more important, making the game more interesting in my view.

Statistics don't provide certainties, they provide probabilities. Just because Van Persie is averaging 1 goal per game, doesn't mean he will score exactly one goal in his next game. Furthermore, we will always have something interesting to talk about because of the randomness element - on any given saturday, a different element of the complex game may prevail and create a new picture. You can enjoy looking at the night sky, but you don't stop pushing the bounds of astrophysics for fear of figuring out the universe - the opposite is true, the more you know, the more fascinating the interplay can become.

You highlight something else that's interesting - referees. I think in fact that the FA ought to publish stats on refs. Open them up to scrutiny, admit their fallibility, and then measure their performance. It's far more honest and useful to accept that they make mistakes, monitor those mistakes, and attempt to correct them or reward the best performers with big matches, than to silence managers for stating the obvious. Why are there no referees from London? Who gives the most red cards per game? Who is biased to Man U? Who is the most influenced by the home crowd? These are all interesting questions that can be enhanced with statistics and make the game better, while at the same time protecting the referees. You can only correct a mistake when you're aware of it, and you'll get a break when people start from the point of acceptance of a certain degree of error.

Snodot
@24

I know what you are trying to relate but astrophysics and football in the same sentence, please.

Do you know why more people speak about football than any other sport on this planet. It has a vast number of issues, talking points.

Referees are a prime example, they make mistakes or interpret the game differently from others, people speak of them and certainly create debate. In a perfect world referees would not make mistakes.

Now will it make the actual game better, who knows? I doubt it will make a better game of football, although their may be a different result. One thing is a certainty is that it will reduce the topic of conversation surrounding referees.

Reducing the topics of conversation in anything in life, erodes the interest. Now follow that through the purpose of stats in football and in the majority of cases the conclusion will be the same.

Football is the world's number one sport because of it's unpredictability in many parts of the game.

Statistics are always used to support a debate, primarily because those using stats will present them to bolster their case. There is always an alternative view point to every stat, it only means what it does to the reader.

What I like about the concepts in Moneyball and the book it's based on is that it shows there are roles on all winning teams for players of all abilities - an individual only needs to be able to contribute one very significant stat in the team endeavour and it's the job of the non-playing staff to properly utilise certain combinations at key moments.

Most professional sportsmen in team sports hover around the average line and I'd say there is only a top 5% of geniuses we salivate over, while the others keep the canvass clean so those 5% can paint.

Using myself as an example, I have had a decent football and basketball career. Nothing flash, but still got paid for quite some years. Yet a football at my feet may as well be oval, and there have always been 16-year-olds around with easily more natural ability than me, but my OBP? I kept the opposition's star player's performance below his average stats. I was lucky to touch the ball for more than 15 seconds per game in either sport. It's that ol' possession argument again - just kept the ball out of their hands...or away from their feet.

You don't have to be extremely athletic to find elite sport. Just look at the body types in games like Gridiron for instance. Since the William's story in tennis encouraged parents and kids to strive for riches in indivdual sports, my advice to sporting ambitious kids with low abilty is "if you have no talent to be one of the top 5%, then stand next to greatness and have your one contribution game pulled up to something useful" - that'll be team sports for you, useless. (I jest!)

listening to the Beane interview on 5 live, it almost had a 'those that can't do,,,,,,,teach' feel about it. One thing I will say, he is a salesman first and foremost.

Like all salesmen, selling an idea, the effects are never seen immediately. Yes they can be successful or as they are in most cases bad.

I read the book Moneyball and one of the more interesting points that I gleaned from it was the pre-conceived conceptions that scouts had about players and the emphasis that they placed on physical attributes. This is definetly something that applies to football, we have all heard of players who have been told that they are not tall enough, big enough or fast enough to play at the top level.

18.At 16:08 25th Nov 2011, Londoner in exile returns wrote:

Maybe the question, these other sports need to ask and research is, why is football so poular worldwide, while traditional US based sports, truly lack worldwide appeal?
------------------------------------------------------------------------------
I am sure statistical analysis has its place in football particularly in areas such as fitness but maybe it is a more spontaneous game than other sports and hence the Moneyball approach may have less relevance?

I sometimes think modern players have fantastic power, pace and technique but lack the more instinctive flair and touch than typified past players like Best and Liam Brady.

Now and again a player like Messi comes along to remind us of what instinctive skills are all about but all too rare it seems.