TGJB –
A couple of questions related to the Derby figures (from one of the other 98%):
At the close of the Oaks broadcast, they were showing the tractors working on Churchill Downs and Jerry Bailey commented that he hoped the jockeys were paying attention, because Churchill was going to be a paved highway (or words to that effect) on Derby Day. In light of what happened to Eight Belles, I’m wondering how true that was. Do you calculate a variant that reflects only speed of the track surface, independent of the wind? If so, was the CD surface significantly faster for the Derby than it was before the rains came on Friday?
I realize that the Derby figs are pretty noncontroversial, but I do have a general question relating to methodology for which the Derby figs serve as a good example. You often talk about doing races based on “all the horses”, but it strikes me that many races don’t leave you much to work with. You have one or two horses running really well and a bunch running clunkers, and you are left trying to decide whether anyone in between paired their top and, if so, who. The Derby falls in that category. The Gotham is another race that comes to mind. In the Derby you have Eight Belles, and only Eight Belles, exactly pairing her top. (I don’t count Colonel John, pairing a cushion-track top on dirt.) It seems equally likely that she ran a little better or a little worse than her previous top (perhaps getting back to the “1.5” she had run in her last two races). That being the case, why did you opt to adjust the variant by a point, rather than just tying the Derby to the route that followed it?
