Bhagwan, I've thought once more about sprints with the CP and JT algorithyms.

Using the method there were some high liability close-calls in sprints so out of curiousity I've reversed the method, and the results are surprising. The rules I used are:

1) Maximum distance of 1,250 metres.
2) Lay the highest-points selection with the CP and JT algorithyms.
3) The top selection must not have a total of 60 points or more.
4) No Black Type races.
5) No NTD races.

For rule 3 there were 5 selections at 60 or more points - 2 won and the other three placed.

For rule 4 I don't feel comfortable laying the best horse in top-class races and also often ridden by one of the best jockeys.

Over the past three days there have been 43 races for 4 accidents with a total liability of $8.60 on NSWTAB.

This is an amazing success strike rate as the vast majority of the selections started at single figures.