Round 1 Tips and Results

So, here are my round 1 predictions:

Predicted Odds	Available Odds
Richmond 7.69 vs Carlton 1.15	3.00 vs 1.45
Geelong 1.33 vs Essendon 4.00	1.18 vs 4.50
Melbourne 7.69 vs Hawthorn 1.15	4.50 vs 1.18
Sydney 4.55 vs St. Kilda 1.28	2.60 vs 1.45
Brisbane 1.20 vs West Coast 5.88	1.30 vs 3.30
Port Adelaide 1.20 vs North Melb. 5.88	1.40 vs 2.80
Western Bulldogs 2.00 vs Collingwood 2.00	1.60 vs 2.25
Adelaide 1.28 vs Fremantle 4.55	1.80 vs 1.95

Basically, the first column shows what my program predicts for the odds for each match. The second column contains the odds available from a betting agency.

When deciding which bets to place, its about choosing the teams with predicted odds that are lower than the available odds. So, from the table above, we place bets on the teams in bold.

The results of the weekend, placing $100 on each applicable match, would have given the following wins/losses:

Team	Outcome	Winnings
Carlton	W	$45
Essendon	L	-$100
Hawthorn	W	$18
St. Kilda	W	$45
Brisbane	W	$30
Port Adelaide	W	$40
Collingwood	W	$125
Adelaide	L	-$100
Total ($800 outlay)		$103 profit
Percentage Return		13%

Well, that's actually a pretty good result... a 13% return on investment. I'll also keep a tally of tips predicted for the season. That's 6 for this week.

Now, before you say anything, it might look suspicious that I have posted the predictions and results after the round has completed. Well, lets just say I wanted to make sure it wasn't a complete failure before I even started this thing.

However, I promise to post my predictions for next round BEFORE the games start.

Story so far...

For my first post, here is the story behind this blog and how it came about.
In 2003 I was between IT jobs and thought of a little project in order to learn VB.NET. I came up with the idea of using statistics from past AFL footy games in order to predict the winners in future matches. I didn't know if it would work, but I gave it a go anyway.
Over the next few years I built up the statistical database and worked out different ways of analysing the data. After a couple of trial runs in footy tipping competitions, I had very limited success - not quite what I was looking for.
I put it aside for a couple of years, and the recently began looking at it again after talking to a friend (thanks Ian) who was creating his own sports betting application. I started to take another path where it is not necessary to predict the winner every time, but to be able to come up with odds that are closer to results compared with what is offered by betting agencies. If this can be done, then in the long run I can use sports betting to come out ahead.
And so this is where I'm at. This season will be my first full trial run of how my system goes. I won't actually be placing any bets - I'm not that confident yet! However, I will be documenting my predictions, and showing what could be done this season if it was used for real.