fixed relationships

[thomas]

TGJB
  You put up a recent post about not assuming fixed relationships between distances. I conjured up half a dozen possible interpretations as to what you might be trying to say, could you elaborate on this a bit ?

[TGJB]

One example takes place when the run-ups are variable, like when they changed them this meet at Belmont. Length of run-up affects the time of the race, and if your speed chart is based on ones that don\'t exist on a given day it will be wrong. Another example is track maintenance-- all parts of the track don\'t always get the same things done. Weather itself can also be a factor-- the stretch and backstretch at Belmont can be dramatically different, depending on wind and shade. And sometimes we don\'t know why things are different, but we know they are, by looking at how fast the horses run at different distances.

[thomas]

What mainly prompted my question TGJB was how you're approaching the new 7&½ distances. If you're not assuming fixed relationships (extrapolating between 7furlongs and a mile) can we assume you're 7&½ figures are being generated strictly on the histories of the horses?

[TGJB]

You do make the assumption as a starting point, and barring something unusual that particular one will hold up-- it\'s a much different situation than with 1 and 2 turn races, especially since there are longer one turn races at Belmont. But ultimately you make your final decisions based on the horses, and anyone who tells you otherewise is talking nonsense. Whether you do it off the horses in one race or the horses over a long period of time (creating an average), there is no other way to do it than by using the horses.