By SB Tang
Michael Lewis’s Moneyball is, courtesy of a critically-acclaimed film adaptation starring Brad Pitt, the sports writing flavour of the month. Lewis, granted fly-on-the-wall access to the Oakland A’s Major League Baseball team during their 2002 season, tells the story of how Billy Beane, their general manager, employs non-traditional statistical methodology to achieve on-field success out of all proportion to the A’s miniscule budget. It is the tale of how David abandons his conventional sling-shot for regression analyses carried out by Ivy League eggheads and, in so doing, is at last able to compete with the cashed-up Goliath that was the late George Steinbrenner’s New York Yankees.
I hesitated for a long time to read Moneyball because, firstly, I have absolutely no interest in, or understanding of, baseball (it turned out that neither were necessary to comprehend and enjoy the book) and, secondly, as someone who studied economics at high school and university, I am acutely aware of the manifold dangers in mechanically applying statistical methodology to non-physical (that is, socio-economic) data — the fundamental difficulty in distinguishing between correlation and causality, the ease with which results can be manipulated by tweaking one variable in an equation or including/excluding data from a sample set and the inherent difficulties in detecting, much less correcting such manipulation. Accordingly, I am sceptical of any claims as to the revolutionary benefits of the application of such methodology to sport or, indeed, any socio-economic study.
There are, as Benjamin Disraeli noted, “three kinds of lies: lies, damned lies, and statistics”. Of course, statistical methodology can be useful in any socio-economic study but it must be applied and interpreted carefully; not as a wholesale substitution for substantive analytical thought. At its best, statistical methodology can tell you what happened; but it cannot, in and of itself, tell you why something happened — that requires a theory which is something no computer programme can generate. A theory can only be devised by a human being with a deep understanding not only of the particular socio-economic subject matter the statistical methodology is being applied to but, ideally, the broader historical, cultural, political and philosophical context in which that subject matter is situated.
That being said, by the time I finished reading Moneyball, I could readily accept Beane and Lewis’s low-threshold thesis that the empirical rigour and statistical methodology (christened “sabermetrics” by its founder, Bill James, after SABR, the acronym for the Society for American Baseball Research) embraced by Beane’s Oakland A’s, although by no means perfect, is at least better than the irrational impressionism and superstition which it replaced.
Lewis, a high-flying trader at Salomon Brothers (the Lehman Brothers of the shoulder-pads generation) in its mid-80s heyday, possesses not only a deep understanding of the statistical methodology which underpins derivatives and structured products, but the rare narrative verve necessary to explain the reasoning behind such dry subject matter and how it can be applied to baseball in a lucid and engaging manner.
All this is well and good but if Lewis stopped here, Moneyball would remain nothing more than a well-written economics journal article. What infuses the otherwise skeletal tale with a beating heart is Lewis’s interweaving of the exposition of Beane’s application of sabermetrics to baseball management with an intimate character study of Beane and the human beneficiaries of his empirical approach.
This interweaving reveals two profound ironies which stay with the reader long after “hits” and “on-base percentages” have faded from memory.
The first is that, as a player, Beane himself was the very embodiment of the traditional impressionistic aestheticism which he rejects as a manager — anointed as a future Hall of Famer by the baseball Powers That Be whilst still a high school student on the basis of traditional aesthetic virtues such as wheels, a hose and, most absurdly of all, a “Good Face”, and drafted in the first round by the New York Mets. Beane never fulfils the establishment’s prophecies about him. He does play Major League Baseball but rather than retiring a legend at 37, he retires a journeyman at 27 when he should be at the very peak of his powers as a professional ballplayer, frustrated by his own inability to convert his undeniable talent into statistical results (ie runs) at the highest level.
Lewis poses but does not conclusively answer the question of whether Beane’s subsequent rejection of traditional impressionistic aestheticism as a general manager in favour of heretical sabermetrics is an existential reaction against his own fate as a player or the mere logical upshot of his rational and professional approach to baseball management.
The second irony revolves around the individual beneficiaries of Beane’s application of sabermetrics. We meet: Chad Bradford, the underarm pitcher from Byram, Mississippi who barely made his high school team and honed his craft on the backyard pitcher’s mound his paralysed father built for him with his bare hands; Jeremy Brown, the fat college catcher who can’t run; and Scott Hatteberg, the one-time Red Sox catcher who, courtesy of a nerve injury, can no longer throw a ball.
Lewis takes the time to get to know these people and paints their portraits with pathos and warmth. They are decent, hardworking men whose lives are changed for the better as a result of Beane’s application of sabermetrics. Lewis lets us feel their dreams and fears as they toil away in the obscurity of the minor leagues in Canada (Bradford) or college in Alabama (Brown) and player after player with inferior statistical performance but superior traditional aesthetics is promoted ahead of them; hoping, praying that someday, someone from a major league team will look past their atypical aesthetics to see their excellent statistical results and reward them with a shot at their dream. He captures their excitement and gratitude at finally being given the first or second chance they dreamed of but feared that they’d never get.
The irony is this: Beane himself treats his players like nothing more than economic assets to be bought (when undervalued by the market) and sold (when the profit to be made from their sale exceeds their utility to the team). As Lewis explains: “Players who just a couple of months earlier he’d sworn by he dumped, without so much as a wave good-bye.” A man, in the famous words of one baseball fan, is treated like “a piece of fruit”, the flesh devoured and the peel discarded. Sabermetrics requires nothing less.
What, if anything, can Moneyball and the sabermetrics it champions tell us about cricket?
Broadly, Beane and Paul DePodesta, his Harvard economics graduate right-hand man and number-cruncher, employ two strategies — firstly, the application of non-traditional metrics to the existing data (for example, on-base percentage and pitches seen per plate appearance) and, secondly, the re-interpretation of existing, conventional metrics (for example, walks and home runs).
I expected to see the former stratagem in a book about statistical innovation authored by a former Wall Street trader. The application of such a data-intensive stratagem to cricket is beyond the scope of this article. In any event, it appears that the English Professional Cricketers’ Association, the Australian Cricketers’ Association and the England team have already started to do so.
As Ed Cowan explained, the ACA’s Most Valuable Player Award now awards points not only for “runs, wickets and catches but also for achieving such benchmarks as run- or economy rates, hundreds, volume of maidens, and even playing or captaining in winning sides.”
The England team’s statistician Nathan Leamon, hired by England coach Andy Flower after he read Moneyball, has utilised sabermetrics-style analysis to discover weaknesses in the batting technique of Sachin Tendulkar (struggles to score through the off-side early in his innings) and Mike Hussey (struggles against full balls pitched on middle-and-leg). These two contrasting examples serve as a reminder that the knowledge gained from sabermetrics, although undeniably useful, can only take a team so far — England’s sabermetrics-derived knowledge enabled them to dominate Tendulkar but it did not save them from being dominated by Hussey.
The latter stratagem was surprising if only because it appears, at first glance, to be so obvious as to go without saying. In any big time professional sport, it seems implausible that the management and coaching staff would ignore, misinterpret or fail to fully appreciate well-established, freely available conventional metrics.
But Beane and DePodesta realise that that is precisely what traditionalists are guilty of — they evaluate a young player on the basis of “what he looks like, or what he might become” rather than “what he has done”, a player’s actual performances (as measured by statistics) are treated as less important than his possession of the five “tools”.
Beane’s high school batting average “collapse[s] from over .500 in his junior year to just over .300 in his senior year” and the scouts chasing him neither know nor care.
Bradford statistically dominates Double-A ball and Triple-A ball and, for good measure, impresses in a two month stint with the White Sox, “finish[ing] with an earned run average of 3.23”, but, until Beane comes along, he still can’t find a permanent spot on a major league roster because of his weird underarm throwing motion and speed-gun-slow fastball.
Brown, despite “own[ing] the Alabama record books” and possessing “numbers … better than anyone’s in minor league baseball” fails to even make Baseball America’s list of the top 25 amateur catchers.
This dichotomy between statistical results achieved at the second-highest level of the game and what could be termed aesthetics or technique as seen through the eyes of the establishment, permeates cricket as much as it does baseball. Nowhere is it better exemplified than in the contrasting plights of Phil Hughes and Shaun Marsh.
Phil Hughes is 23 years of age and possesses a first-class average of 49.40, with 17 centuries from 64 matches. Shaun Marsh is 28 years of age and possesses a first-class average of 39.53, with 7 centuries from 65 matches. To put this into perspective, when Marsh was chosen to make his Test debut at the age of 28 against Sri Lanka a few months ago, he’d only scored 6 first-class centuries (he scored his 7th on his Test debut); at the time Hughes was dropped from the Test XI in middle of the 2009 Ashes series, he’d already scored 10 first-class centuries.
Yet, at the dawn of a new era in Australian cricket, it is Hughes who finds himself under pressure to retain his place in the Test XI whereas Marsh, on the basis of three Tests, seems certain to reclaim his place in the side once he recovers from a back injury.
At the crease, Hughes looks fidgety, hyperactive and streaky. Marsh is an idyllic island of composure and serenity.
Hughes, as his detractors point out, scores predominantly through the off-side (they fail to mention that this did not stop Sourav Ganguly and Herschelle Gibbs from having long, successful Test careers) and his footwork veers from unorthodox (scurrying away to leg to make room to slash the ball through the off-side) to non-existent. Marsh scores on both sides of the wicket, off both the front and back foot, and his footwork is beautifully orthodox.
Hughes hits a lot of balls in the air, particularly through the gully, point and cover regions. Marsh plays perfectly-timed classical strokes, keeping the ball on the ground wherever possible — even when he is scoring 115 off 58 balls in a T20 match.
Hughes is the son of a Macksville banana farmer. Marsh is the son of a former Australian Test batman and coach.
From the day he set foot in the Test arena, Hughes has had to endure a constant barrage of barbs from traditionalists aimed at his home-spun technique. On his maiden Ashes tour in 2009, The Times went so far as to label him “technically incompetent”. In the lead-up to Australia’s 2011 Test series against South Africa, his technique was ridiculed at length by Dominic Cork and Rob Key on SkySports. Cork and Key repeatedly questioned why Hughes was being picked ahead of other batsmen.
The answer is: sheer weight of first-class runs over a period of years in the face of inferior returns from his competitors.
The only alternative candidate Cork and Key mentioned by name was Callum Ferguson. Age: 27. First-class average: 34.63, with 6 centuries from 60 first-class matches. Statistically, not even on the same plane of existence as Hughes; but, like Marsh, Ferguson looks good at the crease.
What would a sabermetrician have to say about all this?
Firstly, it is important to distinguish between the market for major league baseballers and the market for Australian Test cricketers. Major League Baseball is a competitive market with 30 participant teams (albeit with different budget constraints) competing to sign, trade and draft players where the value of players fluctuates constantly. Cricket Australia is akin to a vertically integrated monopolist — Cricket Australia and its subsidiaries (the six state cricket associations) unilaterally determine the value of each and every Australian professional cricketer by awarding state and national contracts and selecting players for domestic and international sides.
In a Major League Baseball-style competitive market in which Beane was operating, a player like Hughes with his unconventional technique but outstanding first-class record, would be undervalued whereas a player such as Marsh with his classical technique but mediocre Shield record, would be overvalued. So Hughes would be relatively cheaper than Marsh, that is, Hughes would offer more bang per buck even if his absolute level of performance at the top-level turned out to be lower than Marsh’s. That in itself would be sufficient justification for Beane, operating with a limited budget, to acquire Hughes over Marsh. But that motivation clearly does not apply to Cricket Australia which is only interested in the absolute level of a player’s performance at the top level and can pick any Australian player it wishes.
Upon examining this first-class statistical data, a sabermetrician would not argue that it conclusively proves that Hughes will be a better Test batsman than Marsh or Ferguson. What the sabermetrician would argue is that the data suggests that Hughes deserves a fair go at the highest level. Until now, Hughes’s bad look has denied him that — he was dropped in the middle of the 2009 Ashes series just three Tests after scoring twin centuries in a Test at Kingsmead against a South African attack led by Dale Steyn, and he only returned to the Test XI on a full-time basis after Simon Katich was controversially axed from the Cricket Australia contract list in June.
What Hughes will do with such an opportunity is inherently and unavoidably uncertain. In Moneyball parlance, Hughes could go the way of Chad Bradford and Scott Hatteberg, who went on to enjoy long and successful major league careers, or he could just as easily go the way of Jeremy Brown, who ended up only playing 5 major league games. What is clear is that, empirically, Hughes has earned a fair go.
 Michael Lewis, Moneyball: The Art of Winning an Unfair Game (2004) 7.
 Ibid 200.
 Ibid 38 (emphasis in original).
 Ibid 9.
 Ibid 233.
 Ibid 33.
 Ibid 36 (emphasis in original).
 Ibid 40.