Thanks for responding Crash. It isn't as much work as it sounds because most of my ideas don't get past the first sample. I only put hours into something that shows real promise. Some guys spend months or even years compiling their own databases. The main point of my post is that you don't need a huge database. You just need a good sampling technique.

My results tend to confirm things said by others on this forum. I haven't made any uniquie discoveries. For instance, I find outside barriers as good as inside barriers. Days since last start is a very good factor to include.

But I do try to find strategies that are unusual and other punters aren't doing. For fancy factors I turn to my collection of old copies of the Sportsman, especially the zip ratings. I also have a little system that is testing well from data from the Wizard speed ratings.

But as I say, most ideas crash, Crash. I gave up on one last night. Showed profit in the first few samples but after 600 races it was 20% LOT. Another one bites the dust.

In over a year of research I've come up with three systems I think have a solid basis. All have done OK but two are struggling at present. Hoping for a better day today.

Good luck to all

Hermes