if you read my previous post on this topic then you know how horse racing odds and payouts work. the more a horse is bet, the lower the odds, and vice versa.
now, how does one make money at horse racing? obviously this like all other forms of organized gambling in which the odds are stacked against the player and for the house. the takeout for horse racing, at least for win bets, is somewhere in the range of 15-20%, depending on the track. so this means that if you bet at random, you’ll steadily lose 15-20% of your bankroll until you have nothing. fun.
now of course, nobody bets at random, but unless you have some idea of what you are doing, you might as well be. there is no end of stuff written by and about people who think they can predict who will win horse races, using any number of methods. speed figures, pace, bounce, appearance, etc. you could probably read 1,000’s of books on the subject.
i say they’re all bullshit. seriously. in my opinion, any attempt to predict the future performance of a horse from their past performance is doomed to fail because there’s just not enough data for any particular horse. every track is different. every field of horses is different. i think anybody who claims they can do it is relying on the human mind’s ability to find patterns and draw connections, regardless of whether there is any true cause and effect.
obviously, this is just my opinion. there could well be a few people who are so knowledgable about horse racing and go to the track so many times that they can become very intuitive about outcomes. maybe.
it’s also made harder by the fact that the odds adjust based on who bets on what horse. so there is no easy money because if there are things one person sees in a past performance, more than likely other people will see it, too, and then the odds will drop.
so now that i’ve painted that bleak picture, let me introduce my take on this: horse racing is truly unpredictable. that’s why people still bet on long-shot horses; those horses sometimes win. and even prohibitive favorites don’t always win. but again, the odds on long-shots or favorites typically are good predictors of the true odds of those horses winning. now, that may be true in general, but is it always true?
i began thinking about the possibility that some races might be more unpredictable than others. if there were some races where the odds of the horses were not very close to the true probability of a horse winning, that might be an opportunity. if you knew when favorites would be less likely than their odds to win, or long-shots more likely to win than their odds, then you could tailor a betting strategy to take advantage of this.
i’ve followed horse racing, with varying degrees of dedication, since 1988 or so. i started by going to the free state raceway in laurel, maryland. it was a harness racing track that ran at night, and is now shopping mall. then i started going to the thoroughbred track in laurel and then to pimlico in baltimore. i actually didn’t take betting that seriously, it was more for the atmosphere and diversion. the track was a good place to go to see a cross-section of society. there were nicely-dressed folks there along with the working class people, all more or less peacefully coexisting (except for the knife fight i saw at pimlico once, but that’s another story).
it was also a good opportunity to reconnect with some of my college friends since they lived up and down the east coast. we could go various big races like the preakness in baltimore, the belmont stakes in new york, and follow the kentucky derby and the breeder’s cup races on TV or via simulcast. one thing i began to notice over the years is that there were some shocking upsets in some of these races. in the 1993 breeder’s cup classic, for instance, a horse called arcangues won at odds of 133-1. in the belmont stakes, lemon drop kid won in 1999 at odds of 29.75-1, commendable won in 2000 at 18.8-1, sarava won in 2002 at 70.25-1, birdstone won in 2004 at 36-1, da’ tara won in 2008 at 38.5-1, and ruler on ice won in 2011 at 24.75-1.
it’s true that long shots win in all types of races, but if you could predict in advance in which types of races they were more likely to win, then you could have an advantage. so i submit that races like the triple crown and breeder’s cup are like this. why would that be? for one thing, these are races where lots of non-experts bet – that means the normal expectation that the consensus of all the betters to be right in judging the likelihood of success will less likely to be true. also, these races bring together horse from all over the country, and the world, so that there is even less evidence to compare horses than usual (in which horses usually run at the same track and may have faced the same horses before). also, races like the kentucky derby and the belmont are run at distances that are not common, giving people even less evidence to use in predicting outcomes. for example, the belmont stakes is run at 1 1/2 miles, pretty much the only race on dirt at that distance. lastly, a lot of these races are run with very large fields; in some cases as many as 20 horses compete, which can only lead to more chaos and unpredictability.
so my scheme is simple. bet on all long-shot horses with odds over a certain threshold. i have no claim on being able to pick which long shot might win, so i’ll bet them all. based on over a decade of empirical observations, i’ve decided to limit myself to the kentucky derby, the belmont stakes, and all the breeder’s cup races. you might wonder why i don’t include the preakness in this. i’ve observed that a long shot rarely wins that race – probably due to it being at shorter distance and generally having a smaller field than the kentucky derby.
i started with $200 before this years kentucky derby. there were 10 horses with odds of 20-1. as we now know, none of those horses won, so i am out $20 so far. i’ll keep everyone posted as this progresses. the next race will be the belmont stakes on june 9th.
note (5/9/22) – i did not keep up with this as promised, and obviously the results of the 2022 kentucky derby are relevant to this discussion.