markfinn, when I was looking for 'converging' factors I used to compile rating is the style of Prof Frank George.
Professor Frank George
The Prof was a top number cruncher of his time (circa 1978). He began by trying for value bets but ended up in 1985 conceding that ‘viable’ bets were a better option. His selection method was to rate the critical filters in the descending order that he advised:
Class
Weight carried
Recent form 1 from the past two outings
Recent form 2 (and I think this is Dowst from 1936 the year of my birth) placed 50% starts with 30% wins.
Distance
The betting market
The going (only noted when it was obviously unsuited for a horse).
There is a lot of other checking to do once the ratings were compiled:
Who is the jockey/trainer?
How long since the horse last raced?
What is the likely starting price?
Is the going suitable?
Quote: ’Additional emphasis is placed on horses which have raced (and won) on the same track before. The track where SF and wins were achieved is also noted, and at what distance and in what going. The age of the horse should be noted in cae it is over or under the best age for the race, so weight-for-age considerations of the Timeform kind can be used’.