I think price still needs to be taken into account but not in the way most people think.

Using ratings you identify the top contenders. You don't even need ratings to do this but it gives you a different perspective then just the top X in the market.

Having identified your strike rate for your top contenders (could be 2,3,4,5,etc) that you feel is consistent enough you then need to work on identifying which races to attack.

Its no good having a 80% strike rate for the top 4 selections and then bet into market where you dutch the selections to 95%. In the long term you will likely lose. You want to identify the races where you still see value in those selections. For example getting a 70% dutched book over your selections will probably be a long term winning strategy providing you can record and confirm that your strike rate is not substantially lower in these sorts of races as it probably won't be 80% anymore and may drop to 75%. As long as it holds above the 70% then you are showing some long term edge in your method.