Day 1 – Thursday 4th
18 qualifying races, 11 wins, 7 losers. Using £10 level stakes the day ended £8.82 down.
Day 2 – Friday 5th
20 qualifying races, 12 wins, 6 losers. £5.38 profit
Overall 38 races, 25 wins (66% strike rate) at an average odds of 1.54, and a loss of £6.03.
I was in an Instant Message conversation with Michael, the author, on Friday. Two things became apparent:
Firstly Michael does not play all the races but tends to stop when sufficiently ahead. I will monitor the results and may recommend a staking pattern which is something along the lines of stake 2% of bank, stop for the day when you have made 6%. Not there yet. The two days so far have both been up and then been pulled back to more or less break-even. Day 2, for example, started with 5 consecutive winners! The point is Michael appears to be making good money by using this method himself, but slightly different as advertised in the manual.
Secondly, Michael suggested that my bot is being slightly too aggressive in selecting a horse, especially in the short sprints. Michael has made a couple of suggestions on how to turn this aggressiveness down to more closely mirror human behaviour. I will implement these this morning.