A while back I posted my perl script for football tipping , then later updated it with some new features . Subsequent web site redeigns by the AFL meant that the HTML parser has since changed, and I also added a feature to predict the winning margin for the Footy Park match (something that one of the comps I'm in requires). Finally the script grew a subroutine to help correct typo's in the team names on the expert tips page. The final version is available here.

So, after a promising start, how did the script fare? After the 22 minor rounds of the season, my perl script tipped 120 winners (avg. 5.45 per week), of a maximum 176. Past season's suggest that an average of about 5 will win it for you, but in one comp I finished 5th while the winner scored 123 (avg. 5.59). Although the final results aren't in yet, I expect to be further off the pace in our work comp due to some excellent tipping by the leaders.

If nothing else, it proves that 143 lines of Perl will produce a better result than a coin (when I tipped with a coin a few years back I got 90 for the season) . It also proves that a small Perl program that knows nothing about the game is likely to tip better than most footy fans :)