You have a selection system that performs best for races that contain 9 or 10 starters.
My selection system may perform best when there are 5 or 6 starters.

How you go about things is different from how I go about things.

Different strokes for different blokes.

What you need to do is try to conceive of any possible reason that the number of runners would influence your profit.
You may find that there is in fact no link at all ,regardless of your current winnings.
Your sample is so small , that it is in fact insignificant , regardless of the fact that you are doing very well.

Good Luck with it though.