In the first article of this series (The Horse of June 1995, page 21), I went on at some length about how much one could expect to predict performance based on the examination and evaluation of conformation. I tried to make it clear that, despite the many claims advanced, there is really not much accurate predicting of performance possible.

It is perfectly obvious, without any mechanical folderol, that a Shetland pony is not going to run as fast as a Thoroughbred. Why? If for no other reason, the short-legged Shetland has a much shorter stride length than the long-legged Thoroughbred. That such obvious differences are of limited predictive value may be shown by an experience I had many years ago.

In a Standardbred racing stable there was a black filly by Billy Direct that was a marvel of compact, powerful-looking pacer. She trained like a champion until about 2:18 or 2:20 for the mile. At that level she hung up and simply would not or could not go any faster. Watching her carefully, it was apparent that the frequency increased, but she didn’t go any faster–the stride length did not increase. The effect was that she was working herself to death–legs just flying back and forth but going no faster. There was no lameness or any other apparent problem. Why? We’ll never know; perhaps she just didn’t want to go any faster and was increasing the frequency just to convince us she was doing the best she could.

In any case there was nothing that conformation could tell us about her lack of stride length. She was as big as her father, with better conformation (as one judges such things), and he was no mean pacing horse.

This gloomy picture is mitigated by the fact that conformation and other physical factors can be used to predict whether certain types of lamenes or singularly poor performance are probable. Let’s begin with one of the most common lamenesses of racing horses–carpal arth