Punt Musings

Well, so much for that little plan. Yesterday I thought I'd fart-arse around with One Note and then transfer my post into here. Really? Fucking useless. No such thing as a straight copy and paste between different platforms. It would take me almost as much time to recreate the post here as it did to write the bastard.

Lucky for me, my conclusions yesterday were inconclusive. So I had to go back to the drawing board, but this time with 2 weeks of data, or 172 races. So rather than deal with how I got to that point, let's just say that W2 was not the same as W1, and that we needed to reconsider our approach.

So here's the summary table of what we've got. The only real difference is I have introduced Provincial as a track grouping, replacing Suburban. Basically any non-metro race in a centre of greater than 50,000 population is classified as provincial.

Ok, so this is pretty compelling. As expected, the base shows a return of -8% on both win and place, and no improvement by getting rid of maidens, <3S or both. So what if we just look at metro, provincial or country?

Well Metro base is actually worse. This data says horses with <3S should be avoided in Metro, and quite possible maidens too. The data for provincial backs this up seriously. Country looks pretty bad all-round.

So from this data, what for W3? I think we're looking at a trial for Metro and Provincial only, win bets only, using my formula, but excluding maidens and <3S. This should reduce our workload considerably. Let's do it.