The main aim of my ‘Horses to Watch’ series is beating the market start price (SP) by finding value in the betting market the night before the race. I use a combination of statistical analysis and AI value-modelling to find a horses’ implied probability (% chance of winning the race based on stats) and compare that to current odds – looking for horses that are overvalued. I focus on the lesser-glamorous races (i.e Class 4s & 5s etc) as that’s where the value often is found – often because these races receive less betting money than the bigger races, so the market is less volatile. Whilst the aim of this game is not primarily profit, statistical variance suggests that if you can consistently find value in the markets (by beating the SP) you will be profitable.
Story of the month:
- 21 selections, 12 of which the SP was beaten (3 non-runners)
- Average odds of 9.55 (~17/2)
- On average, the early odds beat the market SP by 42% across the month
- 3 selections beat the SP by at least 100%
- LAW SUPREME was 21.00 into 10.00 (or 20/1 into 9/1), 110% value found (Placed)
- FOSTERED PHIL was 9.00 into 4.50 (or 8/1 into 7/2), 100% value found (Winner)
- PARLANDO was 13.00 into 3.50 (or 12/1 into 5/2), 271% value found (Winner)
- At 271% value found, PARLANDO was the single most value found for a winner by the model so far.
- 5 winners at 8/1 (7/2 SP), 6/1 (9/2 SP), 13/2 (5/1 SP), 12/1 (5/2 SP) and 7/2 (10/3 SP).
- 6 places at 13/2 (4/1 SP), 20/1 (9/1 SP), 9/1 (11/2 SP), 20/1 (10/1 SP), 11/2 (3/1 SP), 8/1 (13/2 SP).
- Profit from early odds was approximately 210% higher than profit from market SP, 19.9 units v 6.43 units profit.
