bh's post of the Dahlman article refreshed my memory of this player, a math genius who reportedly cleared $2.8 million at the GP meet alone last yr using what essentially was an algorithmic dutching system which he applied to st bets, exactas & tris. One of the possible explanations for the 15% drop in handle at GP this yr would be this guy's absence or a serious reduction in his action. I son't see GP every day, but when I do it has seemed that there are not as many situations where horses' odds seemingly change dramatically during the race.
Unlike Dahlman, my theory has been that one might be better off betting against the so-called mystery man, as his system seemed to depress the favs' odds and as often as not seemed to increase the odds of horses in the 5-1 & higher category. While virtually all of the press coverage related to his betting strategy, it was entirely dependent on his odds line, which was supposedly generated by some unidentified computer program. At one pt I thought I could reverse engineer the basics of that program based on observation, but that proved impossible. Perhaps one of you computer wizards has had better luck figuring that out or someone knows if the mystery man is still lurking out there.