I to have been surprised at some of the races put up. I think it shows that there are many more ways and more classes of race that can be analysed profitably. The specialist is possibly more profitable because his number of bets are limited and has a knowledge of the same horses of the same class run in the same races over and again.
I don’t follow horses and have no horse data base so every contest is a new event and I prefer to work with numbers , accumatting evidence for the numbers unfortunately we need a few runs to reflect the true position. Some of the comments on that thread are price less - owners stats - the fact this trainer runs horses into the ground before passing them on - are just two examples - you are not going to get such info easily from many other sources.
mick has pointed out that the winner is found daily and distillation of the selections gives a narrow band of probables.
So tip up and then consider what others have posted in order refine the process .