When I was young I got into actuarial studies. It was a great course to do, but unfortunately I couldn’t feed myself very well so I only lasted about six or nine months, and then went out and got jobs at the abattoirs and brickworks just to survive. It wasn’t until I was twenty-one that I was able to get a clerk’s licence.
I didn’t have much training in actuary, but I loved it and I loved a couple of the subjects. I think one of them was finite differences, which I still use today, and there was this big fat computing science book which I wish I would’ve stayed and absorbed all of!
Evolving your approach
I don’t tinkle with the algorithms too often, but something will come up or I’ll read something, and I’ll rewrite something and then test it and see whether it produces a higher profitability.
The co-efficiency of the different factors are in constant review. However I tend to prefer to leave things alone for a period. I don’t go in there and change things too often, because if you’re jumping at every whim, every weekend, then you would be doing yourself a massive disservice.
So I think once you’ve sort of set it up, you should let it run for quite a period, and only interfere when you think you’ve got something really valuable to test.
Usually that’s inspired from something that I read. I suppose profit’s always an indicator that you should be looking at something, but you know if you go in there and are re-running stuff reasonably regularly, then things are going to jump out at you.
There are trends in horse racing of course, the public are smarter in some areas, and they change in some other areas. My inputs are pretty much what everyone would use, it’s how you transform the inputs.
I’ve always had the view that horses at the top of the weights have got a much better chance than the public used to recognise. They’ve caught up now and it’s harder. But I’m not a big fan of people saying ‘this horse has dropped a lot in weight and therefore you should be emptying out on it today’.
I do think handicaps in Australia are way too conservative, and I’d like to see horses going up more for a win, and then coming down much more aggressively when they lose. Now the argument against that from the handicappers and racing authorities point of view is that ‘oh well, people play the game’.
Well it’s not going to take too long to work out what trainers are playing the game: they win then run tenth five times, and massively drop the weight.
So my horse might go up six points for a win, but if it runs tenth next start you won’t see it revised down. There are horses that are running around that run fifteen, twenty times unplaced, and if you look at their handicap ratings they’re just completely uncompetitive. I don’t think that’s reasonable to owners, but as an input there’s hardly one that’s more important than handicap.
They’re really important. Australia is currently just awash with information and data, and there’s a fair bit of misinterpretation of the data out there. But on the other hand the smart interpretation will reduce the positive prospect of making cash out of knowing sectionals.
Confidence and thinking long-term
If I looked back over a year I won a certain amount of money, I’ve probably won it on maybe twenty or thirty races. And pretty regularly it’s on something that looks like it’s got some sort of poisonous factor there that has turned the public off. It might be that it was first-up for a very long period, or it might be that it drew a very wide barrier.
So it’s the sum of the form that produces your final probability. I think a mistake that some punters make is that they’ll do the form and think something’s a $10 chance, and it’s $20. Now it’s easy to find out why it’s $20, it might be drawn barrier twenty four.
When it wins at $20 – and let’s say it wins one every seventeen – it’s a really valuable bet. It doesn’t win the rate of ten to one, but it might win significantly better than the twenty to one that the public’s giving you.
Yet when it loses and you put your money on it, you think ‘well I’m an idiot, I never back a winner when it comes off barrier twenty four’.
I think that punters, once they’ve done their probability, should stick with them and not try and find the reason why a horse is a lot longer in the public, because my experience is that’s where the money is.
I’ve never seen a set of data that doesn’t improve measurably if you add in the jockey factors. If you aren’t looking at the jockey factors I think that you’d want to take up a new career.
Click here for the podcast and a full transcript of the conversation.