Stop saying “Moneyball.” Stop it. You don’t know what it means. If English fans insist on droning on and on about the blasphemous use of “soccer” to refer to the beautiful game (more on this ridiculousness some other time), then this American insists that you drop your improper use of “Moneyball.”
It does not refer to small clubs buying youth players and developing them until they can be sold on for a profit. In fact, the historical genesis of the term describes the model that replaced that system. It is true that Moneyball does reflect a focus on deriving value from player purchases.
It is also true that value can be derived from low-cost youth players that develop into valuable stars. However, Moneyball is primarily concerned with extracting success as measured by wins or points from each unit of currency spent. This is what true sports fans are looking for, looking for realities, such as powerplay.com.
In other words, it is about efficiency. And, considering the speculative nature of investing in youth players, most of whom will never amount to much, merely spending on youth is not necessarily efficient. The term Moneyball for those with little intellectual curiosity refers to a player acquisition system famously employed by Billy Beane, the general manager of the Oakland A’s baseball club in the US. It describes his efforts to compete with rivals who routinely spent far greater sums on player acquisition.
After another unsuccessful season in which his best players were sold to rivals, Beane concluded that the game was inherently unfair, and that the A’s could never compete on its rivals’ terms. He needed to discover a way to stretch his limited funds.
His revolutionary model was an offshoot of the statistical analyses of others that concluded that traditional baseball statistics used to evaluate players failed to meaningfully capture the contributions of individual players to win. An example would be helpful, but you heathens are unlikely to appreciate one derived from baseball, so I will postulate a fictional example from football.
Suppose statistical analysis suggested that dribbles over ten yards directly contributed to the success – that is that there was a strong correlation between progressive dribbles over ten yards and winning – but that passing accuracy had a very weak correlation to winning.
A Moneyball model would identify players that were high in such progressive dribbles, regardless of how low their passing accuracy is. After identifying this pool of players, success is merely a matter of finding inexpensive players that meet those criteria. This pool may consist of youth players from Spain or journeyman cast-offs from mid-level domestic clubs. The age, resale value, shirt sales, country of origin do not matter. That’s Moneyball for dummies.
Therefore, Moneyball is more about more precise player valuation in terms of contribution to winning than simply buying cheap players and then selling them on. It is about value but the key component of Moneyball is not profit maximization. Rather, it is success maximization. The trick is better identification of attributes that contribute to the only metric that matters – winning
Increasing revenues from money-sponging new stadia may do the same. But neither are examples of Moneyball because their metric for measuring success is profit, not wins. Moneyball is a (not the) model to spending money wisely.
Whether that money is derived from prudent investment in infrastructure or looting the natural resources of one’s country, Moneyball is about spending that money wisely. And, whether doing so requires better scouting, better youth recruitment, or a team of American statisticians, Tottenham could do with a model that measures the wisdom of expenditure in wins rather than profit.
For sure he backed Klopp.
Thank you!
Great guest blog from Hotspurious and it’s achieved precisely what I had hoped for 👊
There’s a little bit of talking at cross-purposes here. The fact is that players that are obviously contributing to increased success get sold for bigger prices than those that don’t. Gareth Bale didn’t get sold to Madrid for £89m or whatever it was because Daniel Levy bought him to sell for a profit. He got sold to Madrid because he helped Spurs get into the Champions League, then had a ridiculous season scoring and creating like rarely seen before. Mostly, he got noticed by Madrid courtesy of two exceptional performances against Maicon, skinning him ruthlessly again and again. Oh, and scoring a second half hatrick in the San Siro when the team were down to 10 men and already 4-0 down.
To be brutal, he got sold to Madrid for the same reason that Wayne Rooney got sold to Man Utd by Everton: their standards had outstripped those of the club and their ambition could only be fulfilled by moving to a bigger club.
I broadly agree with the tenet of this article though: identifying the relevant stats are what matters. I remember having a fairly pugilistic discussion several years ago about the difference between no-risk high-pass-completion players and those who tried 6 killer balls a match, succeeded with three of them and in effect helped create 2 goals. I opined that the latter would be on my team sheet in any game where playing for a draw was not on the agenda….
I noted many years ago that some teams earned more points with less overall possession than when they dominated possession, a very counter-intuitive concept proven in particular seasons by analysing the data from every match particular sides played. But if you think about it, dominating possession usually means an opposing team is ‘parking the bus’, whereas less possession may mean chances to hit decisively on the break. Sometimes scoring on the break is easier than scoring against parked double deckers….
Stats, however, are rarely useful in isolation. It may be the case for some teams that, the more crosses they deliver beyond the first defender, the more points per game they achieve. That’s fine if the crosses reach a team-mate and that implies that team-mates position themselves favourably to enhance scoring. Anyone watching the Leeds-Newcastle highlights yesterday will have seen that Leeds’ crossing wasn’t much use because the attackers were always in the wrong places in the box. You won’t get your wife pregnant if you are always wearing a condom and all that….
I would rather measure ‘successful crosses reaching a team-mate in a scoring position’ rather than ‘crosses delivered’. If 80% of the crosses either hit the first defender, end up near the far corner flag or don’t end up within 5 yards of an attacking team mate, they don’t really add to proceedings. So I’d want right/left wing backs with high crossing accuracy stats.
It’s the same with strikers. The number of shots per game is far less relevant than the likelihood that those shots will lead to a goal. Shots within 8 yards of goal in between the sticks have a far higher likelihood of scoring than 25 yard screamers. One-on-one shots inside the penalty area are far more likely to score than shots with four defenders covering the goal line. The best stats for strikers is not goals/game, it’s goals scored vs expected goals scored given the nature of the chances the striker has enjoyed. Scoring 20 goals in open play is far more impressive than scoring 30, with 18 penalties amongst that lot, isn’t it??
The reality is that clubs will keep meaningful metrics very close to their chests. The only ones they will discuss publicly will be the ones that everybody knows about, so have no competitive advantage in keeping quiet about….
I waved the white flag after a paragraph.
He brought the idea to the Red sox. He actually wanted to hire Beane but he turned it down. Then Henry took his model, then within 2 or 3 seasons the Red sox won the world series.
So more than likely he’s carried and created a similar model at Liverpool. But he also backed Klopp with signings he needed.
First, excellent work! You have absolutely defined the meaning of Moneyball as well as the term’s overuse and misuse.
Second, as an American, I can tell you the term Moneyball has now largely been replaced here by the more generic term “Analytics.” Reliance on advanced statistical measures has mostly ruined baseball and is on the verge of destroying basketball as intuitive in-game and lineup decisions have been nearly 100% replaced by strict adherance to statistics. But few people understand the biases inherent in statistical analysis and what we have now is an incoherent mess where logic and common sense is most often suspended.
Third, I would hope Analytics doesn’t infest European football but it’s probably too late as I imagine the US hedge fun run or owned teams like City, Liverpool, United, Arse and now Villa and Newcastle, to name a few, are already employing statistical analysis and have been for many years discretely.
Fourth, Spurs as usual seem to be years behind current trends and to my knowledge have no Analytics department, which may accidentally be a good thing. My suggestion to improve recruitment is simple. Fire the entire scouting department and hire detectives to unearth whom Southampton, Leicester, Dortmund, Monaco and Ajax are targeting then try to gazump their offers with improved contracts.
Apologies for the long-winded post!
Can honestly say I’ve never seen or heard the term. Certainly never seen it used in this blog. 🤷🏻♂️
In years to come they will do a film about Levyball
Correct me if I’m wrong, but did John Henry not bring this model to Liverpool?
You could have just posted a link to the movie ‘Moneyball’ 😉
TLDR