Thoroughbred Breakdowns and Preakness Predictions

The breakdown of Eight Belles in this year’s Kentucky Derby, just a few years after Barbaro‘s broken leg in the Preakness, has a lot of people worried about the safety and welfare of thoroughbreds.

Statistics on the frequency of horses breaking down are elusive. The closest thing to official statistics I could find comes from an Andrew Beyer column, in which he reports that in Uruguay, one in every 1,500 times a horse races it leads to a fatal accident.

He implies that the rate is higher in the United States, but doesn’t provide any hard numbers. (For purposes of comparison, the death rate for humans running the London marathon is 1 in 67,000.)

High death rates of horses have many causes: they are selectively bred for speed, running on dirt is hard on their legs, and they can’t talk so their handlers don’t know very well what they are feeling.

Nothing can be done about the selective breeding. The whole point is to win races.

A number of tracks have switched from dirt to synthetic surfaces. I don’t have good statistics, but I believe this has reduced breakdowns substantially.

Now, a researcher named Christine Ross is (indirectly) working on solving the problem about horses not being able to talk. Sure, horses can’t use words to say they are unfit, but if you believe this new research, a five minute, non-invasive test lets the horse’s heart speak up.

The test, which is featured in this month’s Equus magazine, measures a race horse’s heart rate variability. (Sorry, I can’t seem to find an ungated version of the article itself.)

Apparently, there are “signature” patterns of heart rate variability that indicate that a horse is in pain, fatigued, or otherwise heading for injury. If I understand the article, they examined 16 horses that were outwardly healthy; 13 of these horses had suspicious patterns of heart rate variability; 12 of these 13 horses sustained an injury or illness within the next 3 months.

On the other hand, none of the 3 horses whose heart rate variability looked fine had any problems.

These are small samples. Still, if these findings can be confirmed, it would be a remarkable case of creative, out-of-the-box thinking providing a simple solution to a problem which one would expect to be incredibly difficult to solve.

In comparison, picking the winner of the Preakness would seem to be a pretty easy task. Big Brown, who won the Kentucky Derby, is likely to be the overwhelming betting favorite.

According to our numbers, Big Brown is the only horse even remotely worth betting on to win in the Preakness. Our calculations suggest that a $2 bet on Big Brown has an expected value of $2. No other horse in the race has an expected value on a $2 win bet of more than $1.55. If I were to play an exacta, it would be Big Brown over Gayego.