2015 was another successful year at AFLpredictions, marking our third consecutive year of profit since starting publication in 2013. This was the first year that player information and additional statistics (beyond points scored) were included in the model. The result was the most successful year to date, with a return of $3.15 on an initial investment of $1 (assuming full Kelly betting).
To illustrate the model’s performance, take for example someone investing $1000 and adopting quarter Kelly betting (that is, only betting the suggested fraction of a quarter of your total bank – which is suggested to reduce volatility). They would have been down $315 after the first two games of Round 6 (and back in profit by the end of the round), then in profit by $937 after Hawthorn beat Sydney in Round 16, and would have finished the year with a total profit of $788.
The fast turnaround on bets in the AFL means that when betting with $1000 at quarter Kelly, the total value of all bets laid in 2015 would have been $6390, giving a return on total investment of 12.3%. I would have liked to compare this number with the growth of the ASX 200, but an investment with AFLpredictions would have embarrassed the traders this year.
Biggest win: Brisbane Lions ($4.67) over Port Adelaide in Round 7. Profit of $0.73 on $1 ($183 if betting $1000 at quarter Kelly).
Biggest loss: Gold Coast over West Coast (suggested bet of $0.53 on West Coast). The game was a draw. Interestingly, we were also hurt by the other draw of 2015, where we had suggested a bet on St Kilda (at $3.78) over Geelong in Round 21.
We tipped 145 games correctly this year, which puts us third amongst The Age’s expert footy tipsters, and we finished the Home and Away season ahead of all the Herald Sun’s expert footy tipsters (I can’t find their leaderboard after the finals, so let me know if anyone comes across it).
Be sure to follow us next season for more profitable AFL predictions!