You know something has entered the realms of popular culture when everybody speaks about it in the same breath as Hollywood blockbusters, their tax bills or smartphones. Well, OK… Monte Carlo analysis hasn’t quite made it that far, but it has cropped up in connection with women’s tennis championships, and a number of other sports as well. Carlton J. Chin (portfolio strategist and fund manager when he’s not analyzing sports events) applied Monte Carlo analysis to forecast the results of the 2013 US Open Tennis and the Women’s Singles in particular. So what were his predictions – and, more to the point, was he right?
The method behind the Monte Carlo madness
Chin asserts that sports are often good candidates for Monte Carlo Analysis because they are marked by specific events: in tennis, such events are, for instance, holding or breaking serve. He used the ability of certain players to hold or break serve drawing on statistics from the rest of the year. Then he used a Monte Carlo analysis in a simulation of thousands of games between these players. His forecasts were that Serena Williams had a 62.3 per cent chance of winning, followed by Victoria Azarenka (16.2 per cent) and Li Na (10.8 per cent). In general, his predictions held good, barring some US Open position upsets like Flavia Pennetta (0.5 per cent) beating her fellow Italian Robert Vinci (6.4 per cent) in the quarter finals.
Monte Carlo analysis in football
Football (soccer in this case) also makes for an interesting proving ground. Football is played over a season with teams matched against each other within each league or division. Monte Carlo simulation allows individual matches to be modeled and the results of those matches to be aggregated for a forecast over a complete season. Besides being the subject of more esoteric writing on sequences of football results, Monte Carlo methods also feature in articles on football written with the betting public in mind.
Other Monte Carlo mainstream applications
Monte Carlo analysis can be used in almost any situation involving complexity and/or multiple decision criteria. But for examples that directly concern large numbers of people, sports and finance may be the two biggest areas of application. Monte Carlo sports predictions are easy to check; financial predictions less so, especially when they concern people’s retirement planning. As the immediate impact of such analyses grows, Monte Carlo methods are also being adapted accordingly. Models of the game of Go (more complicated than chess once you get into it) may use existing knowledge to make ‘realistic’ simulations, rather than random ones. Some financial analysts question the use of normal distributions as the probability distribution function of variables in financial planning, and suggest replacement for instance by lognormal distributions.
Make your own Monte Carlo analysis with Analytica
But why bother about what is or isn’t mainstream, when what’s really important is what interests you? In the tutorial manual for Analytica, you’ll find a range of more ‘popular’ sample models that have already been made. They include decision support for where to have a party (partying is mainstream, right?), for when to leave home to go catch a flight to figuring out which parking space to select. Or, of course, you can use Analytica yourself for the Monte Carlo analysis that you want.
If you’d like to know how Analytica, the modeling software from Lumina, can help you to easily make a powerful Monte Carlo Analysis for any situation, then try a free evaluation of Analytica to see what it can do for you.