Based on Maria's perceived main factors I have looked at our racing for the last three days using the CP, TA, CRS & DIST neural algorithyms. I used CP hoping it would somehow compensate for Progeny as we do not apply (IMHO) much importance on this when calculating our ratings or markets, and there is no neural category to reflect a sire's/mare's progeny.

I also looked at Unitab's prices using a cut-off price of $6.90 as this would more or less equate to $9.00 on betfair. My method was to look at the bottom-third of the neural rankings using the above-named four categories. So a selection would have to be under $7.00 on Unitab and also be in the bottom-third of the neural rankings. The only filter I applied was to look at races where every starter had a minimum of 2 previous race starts.

The results are quite promising:
Tuesday - 3 selections for no accidents.
Wednesday - 11 selections for one accident paying $3.90 on Unitab.
Thursday - 5 selections for no accidents.

So over the past three days there have been 19 selections for 1 accident of $3.90.

The drawback with this method is having to be at one's computer for every qualifying race. If we had the same volumes in betting as England does, we could more or less tell by midday which horses would be selections.