What area were you working in before you started your research in sports analytics, and how did you decide to focus on the betting markets?
I have mainly worked in higher education, with a short stint in the actuarial profession. While working with the Insurance and Pensions Supervisions Unit with the Maltese financial regulator (MFSA), I was involved in collating industry stats, checking the adequacy of solvency reserves and preparing for the new insurance regulation.
While watching a Euro 2012 match, Dr. Frank Neumann, a senior colleague was asking me questions about the betting markets. I couldn't answer most of them, but in a eureka moment I decided to test an idea revolving around applying financial techniques to the betting markets.
At that point I knew that my PhD would take a more sports betting focus rather than a traditional finance one. The current focus of my research is developing and testing a solvency framework for bookmakers.
What are the fundamental similarities in what you were doing then and what you do now?
In both financial and betting modelling; you decide a set of assumptions on which you build your model. These make or break the model, so a lot of care needs to be made with respect to choosing, testing and updating these.
The categories of experts working in both industries are also very similar. You find individuals who focus on the intricacies of a model to develop sophisticated models that provide an edge, others that focus on price movements, and more that focus on the quality of data.
Both fields are focused on measuring risk and finding value, but are there any aspects fundamentally different?
A bet is after all a binary-option and hence can be classified as a derivative. Furthermore, I consider insurance and betting as basically the same, except for one key feature – 'insurable interest'.
When purchasing insurance, one must have 'insurable interest'. The concept is the opposite for a bet. A bet is made to win an amount, while an insurance is made to place you in the same position should something occur.
When purchasing insurance, one must have 'insurable interest', that is have something to lose, if an outcome occurs. The concept is the opposite for a bet; a player of a particular team cannot purchase a bet on his team losing a match. A bet is made to win an amount while an insurance is made to place you in the same position should something occur.
That is, if I purchase a contract that pays me a sum on Queen's Elizabeth's death, I am betting. If Prince Philip does it, then he is purchasing insurance.
This leads to less scope for asymmetric information in sports and betting. When purchasing life insurance, I may not tell the insurer that I have a health condition, if not asked.
The insurer, on the other hand, may have sophisticated models that give them a better estimate of what the likelihood and severity of certain outcomes are. This leads to one party having better and/or more information. The extent of this difference is much lower in betting markets at the time of writing.
The three key items to predict in finance are the probability, size (severity) and timing. Take insuring against a catastrophe, for example. The extent of the damage may range from a few hundred to billions of dollars.
One would also be interested in the timing of such an event so as to have investments that can be liquidated accordingly. There is also significant uncertainty over claims development over several years following an event.
Yet, in betting, we are less interested in the severity and timing, since most bets are short term.
Do you think an actuary is more likely to become a profitable sports bettor and why?
Actuaries are experts in risk and are trained to ask questions and build models. I think this gives an edge over the traditional bettor or the modeler who strictly focuses on the maths.
Actuaries are experts in risk and are trained to ask questions and build models, which gives them an edge over the traditional bettor
Furthermore, actuaries are often involved in interpreting results from a range of sources and presenting them to a non-expert audience. This ability, essential in a consulting actuary, is also very useful when trying to work out what biases in the betting population can be usefully exploited in betting.
What advice would you give to an actuary aspiring to apply their knowledge in sports betting to make money?
It is truly enjoyable to apply actuarial knowledge in sports betting. If you read through my strategy articles, you might as well recognise the influence of an actuarial mind. As with any betting suggestion, treat it as a fun exercise!
The interviewee would like to thank Nick Foster for his helpful insight. Nick Foster is a qualified actuary and lecturer in actuarial sciences at the University of Leicester, where he is Programme Director for the BSc Mathematics and Actuarial Science. He writes at www.weknow0.co.uk, on a range of pensions and economics-related topics.