Sunday, 23 May 2010

Last Favourite

I'm not a horse racing man by any means, but some of you out there might be interested in this post.

Most serious bettors are aware of the favourite-longshot bias in horse-racing, a phenomenon that was first observed in 1949, and 60 years later continues to be seen.

I remember first hearing about this while I was studying my 'A' Level in Pure Mathematics with Statistics. My Pure Maths was crap, but I did rather enjoy the Statistics part. A pity I'd forgotten most of it by the time Betfair emerged, but the Internet's a wonderful thing.

A recently (2010) published paper from the National Bureau of Economic Research in the USA reviewed the results of no less than all 6,403,712 horse races run in the United States between 1992 and 2001 and some of the findings are very interesting.

The rate of return to betting on horses with odds of 100/1 or greater is about -61%, betting randomly yields average returns of -23%, while betting the favorite in every race yields losses of 'only' 5.5%.

They then also included races from Australia (2,725,000 starts) and the UK (a mere 350,000) which also showed the same favourite-longshot bias. The chart is included above, along with the findings of other reports.

I also remember being told at a tender age that the favourite in the last race was seldom value because it was typically over-bet. It seemed logical, with the explanation that punters were losing by that time, and were looking for a lower risk selection in the last race to recoup some of those losses. According to the report, this was first mentioned in a study made in 1956 and cited in other subsequent studies, but all these were based on small samples.

Not many people would consider 6.4 million a small sample, and it was interesting that this study did not show the last-race results to be statistically significant from other races. So there goes that theory.

I'm noticing a similar trend with baseball, but I haven't yet researched 6.4 million matches, (could be a while), and my sample is way too small for any conclusions.


skinnyfat said...

Very Interesting. I guess, as this is based on bookmaker's SP, starting point to a good strategy would be backing evens or under at betfair. Of course with a big bank and correct money management

dcjm said...

Nice post. There have been a few academic studies looking at whether there is a reverse favourite-longshot bias in baseball (i.e. underdogs are under-bet), but they disagreed about whether that was the case or not. Even so, they looked at bookmaker odds in America, and Betfair odds in the UK may be quite different.

Anonymous said...

Al Pacino in "Two for the money" (movie about NFL betting) claims, that underdogs usually win MNF game because bettors either chase their losses or have money to spend after sunday games...

Anonymous said...

I haven't bet on the Horses for years but isn't this just a reflection of the bookmakers covering themselves on both sides of an e/w bet. They offer shorter prices on win to cover the place side, something that is reflected in the Betfair prices at longer odds where the win part is normally well above the SP and the place below what the SP would pay.

Anonymous said...

This is outdated information and this is no longer apparent, but especially so on betting exchanges.