[Ted Craven]

I am grateful for Doc's comments in his latest letter and hope that I can live up to the praise for my programming efforts. I am compelled to clarify a few points:


1. I did not select the races or handicap them. That was done by Rich Pizzicara and Bill Varone, who helped me assemble the study I sent to Doc. All Doc's praise for that effort belongs to them, not to me.

2. The study I sent, and on which Doc comments, was specifically cherry-picked to include mutuels exceeding $9.00, per his request. Though the small study shows you can still get good longshots, I don't feel it should be compared to a cycle of races with a normal distribution of mutuel ranges. Further, the win percentage for a Validator version corrected for TrackMaster's 2006 changes (and by definition, Spec160 and RDSS), shows an even better rate than the classic Val3c. This win percent (by BL/BL tier), is born up by other users who have reported about their experience so far with Spec160. For example today, one user's brief report was that Spec160's top 3 BL/BL win rate is 81%; top 4: 92%, on his last 27 races. I would not be surprised to see the average win mutuel, though, notably lower during a normal distribution of races, including low paying favourites, some of which will beat us, some of which we'll hopefully pass (but don't forget exotic pools!)

3. The success of the BL/BL ranking (and surely other program readouts), is heavily dependent on proper contender selection (and pretender exclusion), and Richie and Bill V have worked very hard on those skills. It makes perfect sense to me that a thorough Match Up analysis of a race is the indispensable prelude to getting the right contenders (at least for win), and thus getting anywhere near the kind of win percentages reported by the programs. Jim Bradshaw and his students are giving lots of valuable help here in teaching this aspect of the Methodology.

Ted