Moneyball and Cricket: Picking the Right Players

Do you have Moneyball fever? Non-American readers, let me explain: Once, there was a baseball team. It had little money. (Unlike in the IPL, where salary caps limit what teams can spend on talent, the MLB lets rich teams outbid for prize athletes.) So, the team’s manager uses statistical analysis and finds a new way to predict a player’s value. In doing so, he finds all sorts of hidden gems that carry the team to the top.

Why do I, a non-baseball fan, care? Two questions: 1) Are conventional ways of evaluating cricketers all wrong? 2) Is cricket ready for a similar statistics revolution?

1) I have long argued, for example, that good fielding is overrated. Once you cover the basic stuff — catch well, throw well, run well — I don’t think a good fielder adds that much. I’d rather have a good batsman with Munaf Patel energy than an average batsman with excellent fielding skills. But there’s a broader question at stake: do we know how to predict a good cricketer? For example, is a batsman who rotates strike often better than one who drops anchor and tires the bowlers out with a solid defense?  Is an economical bowler better than a strike one? Or take T20: would you rather have Jacques Kallis, or, say, 4 players who can hit 30 runs off 15 balls?

2) Can statistics really work in cricket? Baseball seems more one-dimensional; in a cricket line-up, you need a variety of characters. The openers have to be solid in defense; the lower-middle needs to be able to ramp up the pace, etc. Then again, I once had a math-minded professor who liked to try and predict what a batsman would do with each successive ball, and more often than not, he’d get it right. I’m sure the betting types are basing their values on some sort of modeling, yes? But has anyone read of a team that uses statistical analysis to try and a) value particular athletes; b) predict particular outcomes; and c) base strategy around the numbers?

UPDATE: Of course, the English are on it. Via The Old Batsman:

Ever since Lewis’s book, every sport has tried to find its version of Moneyball. Andy Flower found Nathan Leamon, a mathematician from Cambridge University who was also a qualified coach, and provided a well-funded black-ops stats department at the ECB for him to use [it’s easy to imagine A-Flo wrapping an arm around Nathan’s shoulders and telling him to ‘think the unthinkable…’]…

[Leamon’s] gone to town and then some. England’s enthusiasm for Hawkeye extends way beyond the DRS – they’ve used to it log and analyse every ball delivered in Test match cricket around the world in the last five years.

With access to such vast data they now run simulations of every Test match they play, taking into account venue, conditions, selection and pitch. Leamon reckons that such ‘games’, when he checks them against the actual matches, ‘are accurate to within four or five percent’.

Other work has been in breaking down pitches in areas for bowlers to aim at: Leamon claims England’s palpable success against Sachin Tendulkar was due in part to statistical analysis that showed Sachin made the bulk of his runs on the leg side until he reached fifty.


4 thoughts on “Moneyball and Cricket: Picking the Right Players

  1. Russ says:

    DB, a few points about advanced metrics.

    1. The whole point of moneyball is using advanced metrics to notice things that standard statistics didn’t represent well, and therefore misrepresented a player’s worth compared to a “true” measure of value: OBS is better than average; fielding is under-rated; long at-plates tire pitchers compared to short ones. This was possible largely because baseball is less one-dimensional than cricket. In cricket, by-and-large, averages and strike-rates are accurate measures of value.

    2. Value means a market-place for mid-range players; because cricket is international it generally includes the best players available. Within nations, the non-best players don’t have a comparable statistical database of scores that might allow improved selection; notwithstanding the possibility that we don’t understand the game at all.

    3. There is a garbage-in, garbage-out effect in cricket that is nearly impossible to overcome when comparing players. Unless you adjust for opposition and ground conditions the data isn’t accurate enough to matter; and there just isn’t enough data to allow accurate adjustments of this kind.

    4. The non-moneyball aspect of advanced metrics is the ability to analyse fine details: to get a sense of fielding value, probabilities against different types of pitches and pitchers. This is potentially very valuable in cricket, but only three teams can probably afford to gather that data; and only England seems to use it. The question then is how valuable is it to know a batsman’s average/strike-rate against certain types of bowlers or balls in certain areas, and how valuable is a good fielder. Unless you have a metric on that, you can’t know whether it is over-rated or not. A lot of sabermetrics is devoted to exploding myths about value too.

    5. For what its worth, some basic figures I ran suggest a very good fieldsman is worth anywhere between 1 and 5 runs per 100 runs scored, or 3-15 runs per innings. Low-balled, and who really cares, high-balled and fielding is extremely under-rated.

  2. David says:

    I think Russ is correct in his points as to why sabermetrics can not simply slot in to cricket. There are too many unmeasurable factors in cricket to get accurate and meaningful data on a player’s worth. It also seems that the only market in which player value is really an issue is domestic T20, with millions being spent at auctions, and yet paradoxically many teams seem motivated more by box office draws (usually older players) than those that represent the best cricketing value. It seems more fruitful to explore the role of data in coaching and match preparation, in which it seems that England is leading the way.

  3. David says:

    I am also awaiting a cricketing Moneyball movie, with Mike Atherton to be played by Brad Pitt, Ray Illingworth played by Nick Nolte, with Nicholas Cage as Darren Gough.

  4. […] finally watched Moneyball last night (I blogged about the movie here and here). While I agree with Russ that the movie’s message — statistics yields better […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: