Published on September 19th, 2012 | by Emma0
Teeing off a career in sports statistics
Bradley O’Bree, RMIT University, talks to Emma Bland about how he has combined two passions – sports and statistics – most recently through his AMSI vacation research project, Applications of Simulation and Bayesian Inference in Golf.
Your project on professional golf grew out of some previous work you have done at RMIT. What other sports have you looked at?
Last year I did some work that examined badminton serving strategies. In men’s badminton, for example, about 95% of the time they’ll do a low serve that just pops over the net, and then occasionally they’ll hit it up over their opponent’s head. What we wanted to do was optimise the timing of that, given that the receiver would be expecting a low serve. You want to throw your opponent off-balance to increase your chance of winning the point, so it is important to do that on the most critical points in the set. In badminton, and also in tennis, it’s often a good idea to avoid your non-standard serves early on so that your opponent doesn’t learn to watch out for them.
What are some of the general challenges in sports statistics?
We often have to make assumptions in sports statistics. We’re modelling human behaviour, and what’s more complex than someone’s brain? People have motivations for doing things and respond differently to scenarios. It’s too complex to know everything, so we have to generalise. What’s great is that you can still obtain a nice result.
Is it difficult to obtain data?
Gathering data is very time-consuming. Unless someone will hand you data for free, which they won’t normally do, you either have to pay for it or go out and get it yourself. You don’t realise how quick some sports are until you have to press buttons to record what’s going on! If you make a mistake or stop for a second, you’ve lost it.
It is good to get out and watch sports though, and this is one of the main reasons I’ve gone into sports statistics. I’ve always been strong in maths, and I’ve always loved sports; I’ll play anything – any weather, any conditions – and being able to merge them together is ideal. I was lucky that Associate Professor Anthony Bedford, my supervisor, was at RMIT because he’s a real figure in sports statistics. It’s a great place to get experience and my love for this sort of work has really developed there.