Earlier this week in the Thoroughbred Daily News I wrote a letter to the editor in response to Bill Oppenheim's column. Bill pointed out that the Beyer speed figures for the Derby and Belmont were starting to get really low, that is the winners were generally getting slower. Bill's suggestion was that we should recognize that the market has selected for speed and change the distances of the races to suit the horses that are being bred for.
We have done some work on this from a genetic perspective. There are quite a few winners of Triple Crown races that are genetically Sprinter/Milers. Two in particular have gone on to stud and have in the main sired sprinters, which is what they were genetically. There is another with his first yearlings selling this year that we think based on his genetic markers is going to sire a lot more sprinters than breeders think he should. We have also had a look at the variants that we look for when testing horses for distance and found that the North American breeding population has the second highest proportion of "sprinters" in terms of their genetic markers. This ratio is behind Australia who are noted for their turf sprinters.
I decided to take a look at the times for the three North American classic races and see exactly what has gone on in terms of raw times. Here is a plot of the Kentucky Derby times since the distance was 10 furlongs (click on the image to expand it).
You can see from the trend line that there is a significant reduction in times with the period between 1970 and 1990 really being where the horses were running at their fastest.
Here is the Preakness Stakes since they started running it at 9 furlongs.
Again, you are looking at the fastest times being posted in the 1980's and the trend line starting to creep back up in the 2000's.
Finally, here is the Belmont Stakes since it has been run at a mile and a half.
High Echelon and Secretariat are outliers for the Belmont, and really if you take a look at the standard deviation of the three races, especially the Derby, the horses that ran in these races weren't out to break time at all. If we take a look at the last 40 years of all three races, which excludes the outlier Secretariat (1973), it makes for interesting viewing.
It might be hard to exactly see, but in isolating the last 40 years of these races you can see that for the Kentucky Derby there has been a small but gradual rise in overall times, for the Preakness the rise in time has been a little more dramatic, whereas for the Belmont, the horses were actually getting a little quicker through to the early 90's and then they slowed up dramatically in the mid 2000's.
I made the hypothesis in my letter to the TDN that while we have seen that a high proportion of the breeding population in North America are sprinters, it wouldn't be possible to blame genetics alone for this. Breeders would have literally had to start breeding for sprinting speed en masse for such a dramatic effect on times to be taking place. What is more likely is that breeding for speed had been occurring for an awful lot longer, but that recent environmental changes had made this more obvious. My thoughts were that since they started testing for milkshakes in 2005 and banned steroids in 2008, these two events could have precipitated the reduction in times and speed figures. After all, milkshakes allow horses to buffer lactate and run further than they should, and steroids allow horses to be trained harder and recover faster than they should. Both of these would affect any running over distances further than a mile for horses not truly capable of running that far.
It is also possible that in the wake of Eight Belles, racetrack administrators have been instructed to slow the tracks down for the big races. Nobody wants to see a repeat of that event, especially in a widely televised event. I know that Jerry Brown of Thoro-Graph has been on to this for a decade or more. You can't discount the purposeful slowing of tracks as an additional possibility for a reduction in overall running times for Triple Crown winners.