Tuesday, September 18, 2007

Last Time Longshots

I love statistics and although I recognize that they can be misleading I think they can also be used to great effect when handicapping races. Especially championship events like the Triple Crown and Breeders Cup races. I believe statistics to be more relevant for those races because of the one constant: Trainer intent. You would think that trainers simply enter horses in races in order to win but in reality trainers often have many varying unstated goals for their charges. Sometimes they just want them to gain experience or fitness and maybe try them on a new surface to wake them up. Many handicappers put money down on horses who were not entered with winning in mind. But on championship days that all changes. The Breeders Cup is the culmination of the season. Trainers are there to win and predictable patterns often emerge.

Over the next few weeks I'll try to highlight some relevant Breeders Cup statistics that will hopefully help you to identify some winners and avoid some losers. Nothing is ironclad in this sport but personally I like to be on the side of the trend. I think there is more money to be found following the rules rather than banking on exceptions.

All stats quoted include only data from 1996-2006 unless otherwise stated.

One statistic I mentioned briefly last year prior to the Breeders Cup is the negative angle of horse running in the Breeders Cup after being cold on the board in their last start. Or more specifically be wary of any horse who was 10/1 or more in their final BC prep. Of course true to form Street Sense and Thor's Echo both came out and made a mockery of the numbers but I still think the premise holds true.

Quality is often reflected on the tote board. Horses that are long odds in the preps usually have to run above themselves to convince their connections that they belong in the Breeders Cup. However regression after a career top is quite common. This is an attempt to identify horses who are susceptible to regression. Or to identify the horses that simply have no chance at all.

The one exception to this angle is horses who came out of the Arc de Triomphe. The reason that race gets an exception is because its a championship race in its own right. The quality and size of the field is such that very good horses are routinely overlooked in the betting. The odds are not really an accurate reflection of the horses quality.

The numbers against horses who were 10/1 or more in their last start are fairly solid. They have a cumulative record of 137-3-6-9. Street Sense, Thor's Echo and Adoration were the only horses who defied the angle. Even with those bombers win prices you would still have lost $131 (-47.81%) by betting on these horses. The bottom line is that you're better off missing those winners and staying clear of all those losers. Pretty much all of those horses will be 10/1 or more again in the Breeders Cup itself but don't fall into the trap of thinking of them as value. The numbers say they're very poor value.

Be especially bearish on horses who win their last prep at 10/1 or more. They are 37-0-2-3 in the Breeders Cup since 1996. Horses like River Keen, Kelly's Landing, Came Home and Stormello were all prime examples of this negative play.


Spiderman said...

Interesting statistic, but I wouldn't use it as an absolute criteria for "tossing" a contender. The premise of the 10-1 or greater odds-in-previous-race theory is that public money is right.

In BC races, I look for a sharp last race, no matter the odds. A win is obvious, but more likely it is preferred that the horse ran within 2 lengths of the winner. Any sort of trouble or running in a manner that is a departure from usual racing style, will enhance the odds.

Kennedy said...

You're right in that nothing is absolute. Your own discretion must always be applied. Statistics merely give us many ways to look at the races differently than we might have done.

As for the premise that public money is right that's one I'll stick with. Public money is remarkably accurate and statistical reviews of large sample sizes tend to show that a 3-1 shot wins at the same percentage as a 3-1 shot could statistically be expected to. I was actaully shocked when I found out how accurate the public is. Overlays and underlays can definitely be found in particular scenario's but all in all the public gets it just about right.

Spiderman said...

Come BC day, I'll have a runner to announce that upsets the 10-1-last-out statistical theory. Stay tuned.

Kennedy said...

I think you're missing the point. The point is not that this stat can never be breached. It already has been, twice last year in fact. There are many good candidates who were not well backed in their last. I myself like Wicked Style very much for the Juvenile.

This is simply something to help you think. To more fully consider every factor before making a decision.

I hope it has done that.