A very interesting book - but not because of its analyses.
In fact, it;s the little factoids that give it life. For example, there's the story of a player who was substituted afetr 6 minutes (and on other ocasions came off very early one - yet who survived longer at the club than his manager did.
For a book that focuses so much on The Numbers Game (the clue's in the title, guys and gals) that's a big failing. There are a number of reasons for this, but to highlight just three:
1) The charts are badly designed - or badly reproduced.
Some look as though a spider's crawled across them, and the reader could interpret them in any way they fancy. Although some highlight teams (for example) in the chart, most don't. So if the outlier on the right is - say - Arsenal, you're left in the dark as to who the lefthand outlier is. And that could be just as informative.
2) A number of stats seem to be interlinked but are treated as separate ones.
It's as though research on people with 4 limbs had shown that the vast majority had two arms. Further research also showed that the majority had 2 legs. But, forgive me, if you have 4 limbs and 2 of them are arms, then ipso facto the others will be legs.
One example comes when looking at the fact that big wage bills are almost entirely down to the top teams. The implication is that money wins. However, winners (in England at least) also receive huge amounts of money, allowing them to pay the big wages and atract better players. the latter 'chicken and egg' point is to some extent glossed over.
3) The third problem is one outside the control of the writers - timing.
There'a an interesting analysis of the importance of managers, with emphasis on people such as Ferguson, Guardiola, Mourinho and to a lesser extenet Stein. Unfortunately, since the book was written Sir Alex has retired, Guardiola has moved to Bayern and Mourinho has moved back to Chelsea. These changes would have given the authors an excellent chance to look at comparative stats when 'heroes' move.
Although these events took place too late to be covered in the book, however, a reasonably solid stab at it would have been possible. Take for example, the opportunity to assess the effect of Shankly's retirement (which thanks to his replacements actually saw, if anything, an improvement) and Stein's retirement from Celtic (which saw a comparatively major decline in the club's fortunes).
Maybe I'm missing or misunderstanding some of the analyses, which are after all pretty complex, but I don't think so.
There are also a couple of other points that I should bring up. When revealing that the way to strengthen a team is to improve the weakest link, it should be remembered that athough it's big news for managers, the rest of the world has known for centuries that a chain is only as strong as its weakest link.
Finally, the authors decide when comparing forwards that Messi is too good and should therefore be ignored for the purpose of calculation. WHAT?
An interesting book, which certainly does challenge, in my view very successfully, a number of preconceptions you might have about football. First of all, on the role of luck in the game (probably around 50% of the distribution of league tables at the end of a season can be set down to what you'd expect from chance results), and related to this the fact that games are hard to predict (the favourite wins much less often than in other sports). The role of substitutes (they make a difference when you are losing and you should probably use them much earlier than most managers do use them). The role of managers (a bit more important that some other studies have suggested - but it's still hard to quantify). And perhaps most interesting of all, some insight into the techniques of different managers - with teams managed by Tony Pulis specialising in keeping the ball out of play (a very special variant of possession football) and with Wigan under Roberto Martinez specialising in long-range shots and free kicks and totally ignoring corners. Other findings - such as that winning teams tend to find the right blend between attack and defence - are perhaps more in line with received wisdom.
My reservations: the style seemed to me a bit long-winded (with the authors wanting always to build up to their punchlines/surprising findings a bit too much); the chapter on 'predictions' I could do without (I'm not sure how falsifiable most of them are!); and the central thesis - that perhaps there is no one 'right' or 'best' way to play football, because the game evolves and styles of attack and defence evolve - is perhaps underplayed.Of course perhaps it isn't the central thesis of the book - the authors do seem to think that while Stoke or Wigan might defeat (some of) the numbers through their style of play, neither will ever be wining the Premiership...
on 30 August 2015
I read this a few weeks ago and remain in two minds. In certain parts the book really got me thinking and it is very refreshing to see evidence based analysis of the game challenging the seemingly more plausible anecdotal stories that dominate the game.
However, there was a nagging doubt throughout that some of the analysis was not thought through properly. Comparing football to other sports is all well and good, but they failed to consider fully that there are 3 common results in football (win, draw or lose) as opposed to two (win or lose) in other sports they were comparing too, particularly American sports.
That said, sections about when to make substitutions and the comparison of the worth of scoring vs. conceding (who knew that scoring two scores is worth the same as a clean sheat, hence why some teams focus on defence so much?) were done very well.
That all said, the book tried to make this all sound new, when in reality you suspect there is much similar work happening at top clubs. But it would be refreshing to see a further emergence of this style of analysis in the television pundrity of the game, rather than simply over discussing refereeing decisions.
on 20 May 2014
If the authors of this book (or the publishers - I don't know who's ultimately responsible) had just been content to leave it with its main title then I would have walked away from it not terribly impressed, but not particularly bothered by it either; as a statistics-heavy summary of general trends in football there's relatively little wrong with it, aside from being a bit dry. But no, they also feel the need to lure in the unsuspecting (which, it turns out, includes me) with the strident assertion that "Everything You Know About Football Is Wrong". Others will just have to take my word for it that I went in with an open mind, fully prepared to believe that there are plenty of traditionalist myths in football just ripe for debunking. As it turns out, however, Chris Anderson and David Sally's earth-shattering, paradigm-smashing conclusions are mainly as follows:
- Scoring goals is good; conceding goals is bad.
- Defence is just as important as attack.
- Paolo Maldini was really good.
- 3 goals is usually enough to win the game, unless your defence is rubbish.
- Teams who pass well do better than teams who don't.
- Keeping the ball is good; giving the ball away is bad.
- More shots on target lead to more goals.
- Wealthier teams tend to be more successful.
- Smaller/weaker teams do better when they play to their own strengths instead of their opponents'.
Well consider my mind blown. The further I got into this book the more I found myself wondering what its target audience was supposed to be - I assume that only football fans would be interested in the first place, but anyone who's ever watched a game of football will know that all the authors are doing is dressing up blindingly obvious facts and trends with a lot of numbers which are almost completely in line with common sense and general knowledge.
It was only about halfway through that I bothered to check the author bio at the front of the book, and discovered, to my complete lack of surprise, that the authors are statisticians who consult for football clubs. I realise that almost no non-fiction book is written without some kind of agenda or bias, but the degree to which these gentlemen manipulate the data to make their chosen profession sound utterly indispensable, when they're actually telling nobody involved in football anything they haven't known since they were four years old, would be funny if it weren't so irritating.
On the rare occasion that the authors' line of argument does seem to deviate significantly from conventional wisdom, their propensity for misinterpreting or flat-out ignoring facts that don't fit their theories has to imply either genuine ignorance (unlikely, given the amount of time they've evidently spent poring over the data), or a willingness to wilfully mislead. A few examples of this (for anyone still reading):
- They declare that, "As we have seen, underestimating the role of the defence has been a characteristic of football ever since its first codification" (p. 175). The only evidence they offer to support this is that strikers tend to cost more in wages and transfer fees than defenders and goalkeepers, and that Alex Ferguson sold Jaap Stam because he mistakenly thought he wasn't tackling enough (and not because of the controversial autobiography). In the first instance, I think the authors are giving football fans far too little credit; yes, fans love to see goals and yes, it's attacking players who tend to be the superstars with the salaries to match (albeit with plenty of notable exceptions), but the money side of it has a great deal to do with the fact that exciting and/or effective attackers sell more season tickets and more jerseys than defenders do, and the simple fact that fans find goals and goal scorers sexier doesn't mean they don't understand the importance of a good defence to their team's success. As for the second thing, I'll assume, for argument's sake, that it is true, and that Ferguson made a blunder not because of Stam's book but because he though that the reduction in tackles made by Stam was indicative of a player getting lazy and/or old, when it was actually a positive sign (because, like Maldini, he was using excellent vision and positional sense to shut down the opposition's attempts to attack without having to make a tackle); if that's the case, then it seems as though Ferguson acted based on statistics taken without proper context or interpretation, and his error in judgement is thus a neat metaphor for this book as a whole.
- There's a whole chapter given over to the relative values of goals scored first, second, third, etc., the idea being that first and especially second goals are far more important than the extras added onto the end of a thrashing; this is then converted into an "exchange rate" with regard to the points that these goals typically end up winning for the team. This, by itself, makes sense enough, I suppose, but their narrow interpretation of the resulting numbers leads them to declare that one of the top teams should have bought Darren Bent (rather than Torres or Andy Carroll) in the January 2011 transfer window, because his goals over the past couple of years had been so valuable to Sunderland; while it's undoubtedly true that Torres and Carroll both ended up offering a poor return on their investment in terms of production, they conveniently overlook both the obvious mitigating factor that Bent's goals were proportionately more valuable because he was one of the better players at a weaker team and the weight of history - Bent had already had a shot at a top club, and his Tottenham career was, while not catastrophic, hardly the stuff that football legends are made of (18 goals in 60 league matches).
- Finally, it turns out that Roberto Martinez, late of Wigan, now of Everton, is quite keen on his stats, thus making him the perfect poster boy for Anderson and Sally's manifesto; he did, after all, keep Wigan up time and time again despite the extent to which Wigan's financial limitations stacked the odds against them every year, the club surviving for seven years before finally going down last year. Except that anyone with a functioning memory could tell you that they were still managed by Paul Jewell from 2005 to 2007 and by Steve Bruce from 2007 to 2009, and a moment's research will also tell you that it was these two managers who guided them to their best finishes (10th in 2005-06 and 11th in 2008-09) - and the salary details handily provided in the book show that Martinez had no less money to play with than his predecessors. In fact, if only by default, Martinez was the only one of Wigan's three Premier League managers to get them relegated - and the authors conspicuously fail to mention whether or not Jewell and Bruce were big fans of statistics and statisticians. Don't get me wrong, I think Martinez is a good manager, as reflected by his FA Cup win with Wigan and Everton's strong performance this season, but this omission of his predecessors' contributions is really quite outrageous given his alleged status as "one of the heroes of this book" (p. 7) and a brilliant exponent of "guerrilla football" (p. 182).
These are just the most egregious examples that I can call to mind from a book packed cover to cover with information that is either desperately obvious, largely irrelevant, badly misinterpreted (whether deliberately or otherwise), or flat out wrong. I apologise for ending up writing a bit of an essay here when my disdain for this book could probably have been summed up far more concisely, but given the laziness and negligence of which I feel the authors are guilty it seemed important to provide actual examples rather than sweeping generalisations.
on 12 December 2015
Was pleased when my girlfriend bought me this book, because I'm vaguely interested in football and teach stats to psychologists. While an entertaining read, anyone with any fundamental formal training in statistics will find fault with many of the analyses presented. In only the first couple of chapters, I spotted at least three errors: one was banal in that numbers in the text did not tally to summed probabilities in the corresponding table; another misleadingly suggested that the correlation between number of corners and goals scored was "essentially zero" based on average goals scored per number of corners (an example of the ecological fallacy: conclusions from aggregated data over different numbers of games are incorrectly used to infer the game-by-game relationship), and the authors overplay the role of chance in football compared to other sports by neglecting to emphasise the high proportion of draws in football (e.g., the probability of the team taking the most shots winning the game is .49, but that's not as bad as the authors pretend because 1 in 4 games ends in a draw).
When you've spotted that many errors so quickly, trust in the authors breaks down and reading this book became an exercise in checking whether each of the analyses was solid. While that is a useful exercise in its own right, this is a Penguin book that is aimed at a wide audience, many of whom may not be able to apply critical skills to the text. That is concerning, and I'll be on the hunt for a better book.
on 21 June 2013
I loved the whole Moneyball story from the US, and this book does have plenty of interesting insights.
It's time that football used reason and evidence. I think 'gut' feelings in football are as unreliable as in much of the rest of business, sport and life. (We see what we want to see / confirmation bias, etc., etc..)
But... as another reviewer has said, there's a nagging feeling that some of the correlations are the result of other uncontrolled variables.
The authors do make it clear that there isn't one 'best' formula for all teams; that managers must play to their team's strengths. But that feels like an excuse for (for example) Wigan being relegated in spite of the book praising Martinez and his methods there.
Having said all that I enjoyed the book, in particular the analysis of why winning corners isn't much cause for celebration.
So - it's not quite 5 stars. If you think that a bit of intelligence and 'stats', could improve your team - I recommend it.