For the sake of something to do, and because of the immense sophistication of the industry-leading football traders at work, I am reading, at the most cursory and useless level, a few books on financial mathematics.
Quantitative sports analysis really kicked off with the advent of the personal computer. The Computer Group, adhoc pioneers of statistical sports analysis, were able to crunch the data, model the variables and produce some sharp numbers to bet the handicap and total points markets in American sports in late 1970s Vegas. They were able to attack less sophisticated competition. The sports bookmakers at the time produced prices which were more reflective of public supply and demand. Common knowledge is one thing but it comes from news stories, heuristics and gut feeling. Mathematical analysis allows for colder, clearer thinking and the Computer Group beat the NFL and NBA lines pretty comprehensively.
Baseball always resisted analysis. A sport which has a huge culture of statistics (eg a Society for American Baseball Research, no less) has in the main defied the sports quants, although the numbers produced by the bookies look pretty damn accurate before each game.
Cricket is also very hard to model mathematically. For one thing there aren't enough games from which to extract meaningful data. Off the top of my head I think there are around 500 international cricket matches per year. There are also many variables, such as the state of the wicket or the affect of the weather on ball movement, which determine the course of the game but which defy linear mathematical modelling since they are (perhaps) impossible to quantify.
If you could marry Black-Scholes with Duckworth-Lewis you might make a fortune. The problem of induction would more than likely get you though. The future ain't what it used to be.
Subscribe to:
Post Comments (Atom)
1 comment:
Ugh. Black-Scholes. It's basically my 9 to 5 at the moment.
Post a Comment