Andrew Leigh

Today the Herald Sun features an extract from my book 'The Economics of Just About Everything', exploring the link between money and team performance in the AFL. Read on...

Who's the Moneyball star of the AFL? Herald Sun, 7 August 2014

Moneyball tells the true story of how Oakland A's general manager Billy Beane (played by Brad Pitt) put together a great baseball team on a shoestring budget. The movie has two messages: first, money helps win games. Second, some sports teams spend their money more wisely than others. 

But how does Moneyball apply to the AFL? Money allows teams to buy better players (up to the salary cap), to hire better coaches and to buy experts like physiotherapists, masseurs and even statisticians. But do some AFL teams spend their cash more wisely than others?

To see the effect of money on team outcomes, I added up the total amount each AFL club spent on its football programs (player salaries and other team expenses). Then, I sum the amount teams spent over the five-year period 2008–2012, and then look at how this compared with the number of games the team won over the same period (I excluded Gold Coast and the GWS Giants since they were not part of the competition for the full five years).

The first thing you notice is that there is generally a positive relationship between spending and wins. In the period 2008 to 2012, there were seven teams that spent less than $90 million: Western Bulldogs, North Melbourne, Richmond, Port Adelaide, Melbourne, Adelaide and Brisbane. All of them won fewer than 60 games in the five-year period, an average of fewer than 12 games in a 22-game season. At the other end of the spectrum, the biggest spending team was Collingwood, whose $103 million spend saw them win a total of 80 AFL games. On average, teams win one more game for every additional $1.1 million they spend.

But the second thing you notice is that while some teams— like Sydney and Carlton—get about as many wins as their spending would predict, others win a whole lot more or fewer. Money accounts for about one-fifth of the variation between teams, but that leaves four-fifths to be explained by other factors. Geelong and Fremantle both spent about $62 million, but Geelong won 90 matches, while Fremantle won 47 games. St Kilda and Essendon both spent about $90 million, but St Kilda won 71 games, while Essendon won only 47 games.

Relative to their spending, the AFL teams that did best were Geelong (which won 28 more games than their expenditure would predict), St Kilda and the Western Bulldogs (both with 16 more games than their spending would predict). Those who under-performed their expenditure were Fremantle (who won 16 fewer games than their spending would predict), West Coast (18 fewer) and Melbourne (23 fewer).

Economists like to talk about ‘productivity’, meaning how much you get out for what you put in. If you can sew twenty shirt buttons each hour, and I can only sew ten, we’d say you’re twice as productive as I am. In these terms, the AFL’s most productive team is Geelong (who also happened to pick up two premierships in this period), and the least productive team is Melbourne (who won the wooden spoon twice in this period).

One possible reason for Geelong’s success is the skill of its statistics- loving recruiter Stephen Wells, who has identified talent in unusual places, such as Mark Blicavs (an 800-metre runner with little background in the game when he was drafted in 2012) and James Podsiadly (drafted as a mature-aged rookie in 2010).

In Moneyball terms, Geelong are the Oakland A’s of the AFL.

But what about if we look across the globe? One way to do this is to see how well a country’s national income predicts the number of medals they won at the 2012 Olympics. The top two countries on the medal tally are the world’s two biggest economies: the United States (104 medals) and China (88 medals). At the other end, there are 190-odd nations with a national income below US$1.5 trillion.  All of them won fewer than 40 Olympic medals.

The effect of money on Olympic outcomes is significant. For every additional $125 billion in GDP, a country wins on average one more medal.  To put this into perspective, $125 billion is about the value of the economic growth that Australia experienced between the Global Financial Crisis and the 2012 Games. During this period, Britain’s economy failed to grow. If Australia’s had also been stagnant, we could have expected one fewer medal in the 2012 games. But growth or no growth, it’s clear that Australia does significantly better at the Olympics than our economy would predict. For example, a nation with our size economy would have been predicted to garner 16 medals, less than half the 35 we actually won.

In the case of AFL, I was looking at actual sport spending figures by clubs. Here, the comparison is national income, so it omits the differences in how large a slice of the pie goes to sports. Yet while the AFL analysis only managed to explain one-fifth of the variation in results, the Olympic analysis succeeds in explaining more than two-thirds of the variation in medal counts. That’s partly because richer nations are larger (and therefore have a bigger talent pool to choose their team from), but also because – unlike AFL – there’s no attempt at equalization in the Olympics. 

This is an edited extract from 'The Economics of Just About Everything', Allen & Unwin, 2014