I have two betting approaches I use at the moment based on ratings that are highly profitable. The following is an explanation of why (my reasoning at least) they are profitable in line with what you and others have suggested either regards to being contrarian.

I have ratings which I put together myself which are based on combining a few different ratings for a 'ratings of ratings' if you will plus some other factors i calculate myself for a final rating. It's pretty much a consensus rating of which you'd expect is overbet and not useful however using 3 filters the top rater is quite profitable:

- I only bet if my own speed rating is in a certain range (not contrarian just found this to be profitable)
- I only bet fields of 8 to 14 runners (not contrarian just improves the results/strike rate)
- final filters is my contrarian filter and all i'll say is it has to do with the horses last win. That should possibly be a big enough hint in itself, but it has been the filter thats tipped me in the right direction.

That's all i'll say for now, but it definitely pays to be contrarian. You don't necessarily have to be contrarian to the actual ratings, but possibly just contrarian with 1 or more filters once you have those ratings.

I honestly feel like I can make a consistent profit over the long-term these days, even though its only early for me. If I managed to get a bot in place that automated my selections and only bet overlays then i'd be even further infront but i'm not at that point yet.