I'm sort of game
I started with this 'system' based on a few books I have read and feedback from various sources. At the beginning of the spring carnival. I put a whopping $100 into 2 online accounts (to spread my bets based on who was paying the best odds). At the end of last weekend I had $152. Given the amount of time I've put in, this equates to about an hourly rate of 1c an hour :wink:

Seriously, it has not been worth the effort from a financial perspective - but I am enjoying the challenge.

So, without giving you the weightings for each category, below are most of the statistics I collect on the history of each horse running in a race that interests me :

1) Race class (difficulty)
2) Loosing Margin (in lengths)
3) Course
4) Barrier
5) Jockey
6) Running 'issues' (checks,blocks,etc)
7) Weight (carried & limit)

Each one adds up to a 'score' for the horse. You need to do this for the past few runs for each horse to get an idea of the horses form (improving or not).

For the race you are interested in you take the score for each horse and modify it based on:

1) Ratio of number of wins to races at this distance.
2) Ration of number of wins to races at this track
3) Penalties for 1st run back from a spell (weighted by age)
4) Penalties for previous declining form or bonus for previous improving form
5) Jockey
6) Barrier
7) Weight carried & weight limit
8) Ratio of total wins to races
9) Race class difference from previous runs(is the horse going up or down in class for this race)

From this I end with a final score (in lengths) for each horse and from this I frame a book. If the TAB odds are better than the ones for my winning horses, then I bet on them. I don't just pick one horse for each race, that's too hard. I try and place a number of smaller bets around my top 3-5 horses (depending on score), so that if ANY of the bets come in, I covered all others for that race. For example, if you look at my tips for the past two weekends, I've had a few donkeys in there. Fortunately, a few have paid well (Just Devine/La Zoffany quinella, Sir Talaq/Cousin Alf quinella both last week &
Let's Planet Win the week before).

Simple, right?

There are (quite) a number of faults/limitations to this 'system' that I have uncovered over time and am trying to eliminate. Most of my work has been in automating the above so I don't spend hours doing it manually.

I'm sure you can poke a few in it yourself. However, in this limited period of testing, it seems to be working reasonably well.

The above system only HELPS me work out where there is possibly a value bet. It's based on quite a number of assumptions, such as knowing when to ignore a previous race when calculating a horses history.

I'll let you know how I'm going (if I'm still going) in a year from now.

(now I wait for the waves of laughter of how I waste my time & money...)



[ This Message was edited by: mac on 2002-01-09 16:26 ]