Bernie, my best systems were tested using nearly 6 years of old herald suns. one tends to become tired after a while, and with the odd one missing, the results weren't conclusive.

I now have a database which i believe to be accurate because all the horses that should be there with my systems, are there. If I can offer an insight into what I think is the best way to develop a system;

It needs to be logical with as few rules as possible.

There must be a nexus between all your rules !!!!! (This is my No. 1)

There is no point having rules that are similar. all this does is cut the bets down and doesn't do anything for the POT in a lot of cases.

Concentrate on doing the opposite of the herd without being too silly.

I have noticed that good form produces winners, BUT, there are quite a few horses in each race that would qualitfy as having good form.

Everyone says look outside the square and I believe it to be true for the simple reason you're going to get a better price.

A system needs to have a decent Win S/R or it's chances of having that muderous run of outs are very good.

I cannot make a 7 day LSW work, nor can I make a topweight system work ..... but I have several methods that show a really good POT. My systems allow only a few bets each month so to make wages I would need to put on far too much money for my liking. So it's a fun thing ATM. There are systems out there that work and show a decent POT, I'm sure of that.

A couple of strong filters which improve systems include "career starts of 8...21" courtesy of sarge1 and only betting "Metropolitan" races. The class of a horse (which I've poo-poo'd many times) is paramount to a successful system.

good luck