Thanks George. Having an idea is one thing but despite the early success the question is always what will the long term results be? Is the selection process sound and above all is it likely to be profitable in the long term? What size bank is required? All these and other questions can only truly be answered after a period of time that encompasses the different periods of the year, however statistical analysis of the historic results can provide a reasonable indication. So I keep meticulous records on a spreadsheet. This enables me to study and analyse the data and results to identify trends or potential areas of improvement.
The fact that the selection process is 95% mathematical in nature means that in theory there is little subjective bias i.e. I’m not relying on my own ability or knowledge to identify the selections, which in theory should mean that the future results are more likely to replicate the past. Statistically this means that having reached a 300 selection threshold the future Strike Rate should be plus or minus 3% as long as the selection process is constant. Add in the fact that recording the results at BFSP invariably means that improved results can be obtained by obtaining better early odds or Betfair prices (I keep records of both early prices and industry SP) and that 22% of the winners were Early Priced double figure winners, then the question becomes: after 300 selections would the process still be profitable with a Strike Rate of 3% less?
Over the past 12 days the process has identified 571 base selections. I then reduce these with a filter to produce the 300 published selections. The original 571 selections (from 271 races) have produced 96 winners. That’s a 16.81% SR. The average early price of all these has been 14/1. These winners have lost 20 points at industry SP, but at Early prices we see a 135 point profit (23.64% ROI) and at BFSP a 117 point gain excluding betfair commission. The average early price of the winners equates to 7.35/1. If the strike rate of the base selections reduces by 3% to 13.81% I would then have 79 winners. At the average winning early odds of 6.35/1 this would generate a return of 580 points making a net gain at early prices of 9 points. So still profitable despite a 3% reduction in strike rate at early prices.
Looking at the filtered published selections. 57 winners from 300 selections (from 234 races) equals a 19% Strike rate. At SP we have attained a 53.62 points profit. ( This figure alone is highly unusual – a 18% ROI at industry SP!) At early prices a gain of 154 points (51.33% ROI) and at BFSP a profit of 136 points (45.33% ROI). The average early price of the winning selections is 6.96/1. If a 3% reduction in the winning SR occurred, this would mean a 16% SR and 48 winners. At the average winning early price of 6.96/1 we would then see a return of 382 points for a 82 point profit – a 27.33% ROI at early prices.
So on balance I’m extremely pleased, but also conscious in this game that results can change quickly. The numbers to date however are very positive and promising. The process isn’t price dependent meaning that value propositions are definitely being identified. The process finds them. Also the Strike rate being in excess of 15% means that in terms of a betting bank a 150 point bank should be sufficient enough to cope with any adverse bad losing patch. We all like and seek big priced winners, but my experience with the sinister type threads leads me to personally preferring a steady stream of winners at all prices with the odd big winner thrown in. I’ll review the data again at the end of November.