It’s funny how things come out of the blue. Last week, I was reminded of months of slogging at a keyboard, writing a thesis about football referees, back in 2004 as part of my parting shot to university life.
First off I was at a talk in the Science Gallery hosted by Ger Gilroy of ballsdotie and Newstalk fame. The topic was Track and Field: Data and Sport, which was part of their Lifelogging exhibition exploring how all this data in the world today is being used by sports teams and players. The talk with sports journalist Simon Kuper and Tom Markham, who sells all his sports data to teams (and his company invented Championship Manager, which I also played through my Leaving Certificate).
Are Football Referees Biased?
Anyway, how did this relate to my thesis? Well back in 2004 I was finishing my BA in Business in Economics in Trinity College Dublin, and my economics Professor Brian Lucey suggested a topic on behavioural economics. Boring, I know. But he said the twist was the topic was football, so I’d get to spend hours analysing football stats. It sounded better than studying interest rates and macroeconomic theory, so I jumped into it.
The idea was to try prove if humans always make cold, rational, logical decisions or whether we’re influenced by other factors. My arena was the football field and the subject was the referee.
We’ve all heard of “home town referees” favouring the local team, or the famous Fergie Time at the end of games in Old Trafford, or the fact Man Utd seemed to get way more penalties than they gave away. Was it down to a level playing field or were referees being influenced by the hostile crowd or a screaming Scotman on the line?
How much added time?
The way I tested this theory of behavioural economics was by measuring the amount of added time a referee played at the end of the game. This time is at the referee’s discretion, although he has guidance on adding time for substitutes, yellow/red cards, penalties, etc. So the 4th official does estimate it, with say the 4 minute signed raised, however the referee could ultimately play 3 or even 5 minutes and no body could question him.
So what did I find? Well I compared a full season of data in the US MLS with the Italian Serie A. I expected Italian referees to be more biased, i.e. add more time if the home team was losing a tight game (e.g. AC Milan 0-1 down in the 89th minute against Juventus in a packed San Siro).
In fact it turned out the US referees were a little bit more biased, which was hard to explain, as the league is smaller, rivalries and crowds are less and there’s less media attention. However it was a valid thesis, I spent lots of times gathering data on subs per game, yellow cards per game and penalties per game.
All this was hard, especially as I had to manually transcribe an entire season of data from the Italian league websites. At least I learnt the Italian words rigore and carta gialla. Plus the best bit was getting to interview Jimmy Magee, Damien Richardson plus a top Irish international referee.
Next blog: how my Masters dissertation ended up at a conference in Kenya…