I’ve been tinkering with the idea of creating a spreadsheet to systemise match tips for Australian rules football for a while now. This is because of anecdotal evidence that a reasonable systemised method of selecting winners is likely to out perform most football tippers because it removes any psychological bias.

So, in my spare time in the last month I’ve done a little statistical analysis of the past four years of match results and arrived at a few factors that seem to have the strongest correlation on who does (and doesn’t) win each match. As you imagine, these include who is the home team, relative ladder positions, and so on. I’ve come up with four slightly different systems to tip each week, with various parameters tuned to fit actual match results over the last four years.

Systems 1 and 3

Systems one and three use the same straight comparison of weighted relative ladder position and match type (home, away, interstate, neutral). System 1 includes a bias for some teams, predicting that certain teams will rise up the ladder in 2015, and certain teams will drop. System 3 does not.

Systems 2 and 4

Systems two and four use the same comparison used by the first two systems but apply a dead band selection. Basically this means if the comparison is sufficiently small then these systems select the team that has the highest weighted relative ladder position. Additionally system 2 uses the same bias for some teams that system 1 uses.

Tips for Round 1

Here’s the tips for round 1 (along with my own tips). Each week I’ll update the tips and compare the performance of each system.