I think Chrome's parameters for a winning system are a bit too rigid but he makes a couple of excellent points about keeping it simple and retro fitting.

It looks like a few people have very different veiws on retro fitting,to me it is looking at past results and adding rule after rule untill the POT is high enough to make you happy.

Trouble is that unless the added rules are very logical and have each been tested over hundreds of races all you are really doing is finding anomalies of the time that probably will not stand up when it comes to future racing.

An example would be to include a rule like must be in the top 5 weights to an angle you want to pursue,a short test could show this helps the POT a little but the reality of the situation could be that whilst the rule got rid of quite a few losers,during the short test period you performed those in the top 5 weights were running hot and those outside the top5 were going through a dry spell,when it comes to betting the system the situation corrects itself and your system goes bust.

So in the above example your intitial angle that showed a lot of promise was destroyed by your own over eagerness and greed when you applied a rule that was retro fitted.

Keeping it simple and putting say 80% of your systems fate in the hands of your inital angle is about the only way to make mechanical systems work.

Coming up with that inital angle is the hard part,get that right and yes you can make a profit from a mechanical system,dont let those black hats who are oblivious to their own mechanicaly influenced handicapping tell you otherwise.

Yes Crash that last bit was directed at you,your own words about how you are very strict about betting only on certain race types proves how oblivious you are to your own mechanicalism.