Can't work out how they do it, I have the number of races and added up the horse's winning and it agrees with the API I get automatically.

As far as I'm concerned this is not exactly a winning factor, so I don't use it, but still, it's intriguing what they do and why?

Although if I were to use it, I'd be inclined to use the last couple of years'
of earning at most, 12 months maybe a bit short, life time earnings may be utterly meaningless.
Just my opinion.

TBH, my main concern is the consistency of the data. Have been looking at something on Uselessbettor's tester, which uses R&S data but only goes back to August. It would seem misleading, therefore, to look back at earlier data from other sources with regards to "API".

Hopefully, however R&S do calculate their API's, they have always been consistent; though I've even got some qualms about that now, given the peculiarities of the figures.

When I've done research in the past, I generally found that career API was an overbet factor. But that can be just as useful to know as the reverse! Twelve month figures I do believe have a slightly higher impact on strike rates, but I don't have any hard evidence to hand on that.