Sidebar

[rubicon55]

Originally Posted by pondman

I have data going back to Lotus 1-2-3 on a 286, starting in 1985. I moved it to R base in the 90s and then access. I found enough structure to be able to drop the relational side. I believe I'm beyond the stage of finding out what works.

At the same time I married into a California racing family. Fortunately, my wife doesn't train or have a track job, although she does have a small stable. She does get a number of mares from SA. And so over the years dinner coversations has included brawls about "your donkey can get to the half mile pole quick enough."

I've dropped all performance analysis and have started to focus on billionaire/ millionaires looking for a soft spot (because that's where I believe is the consistent money.) And the things they'll do to conceal their moves-- including low percentage or unkown jockeys, changes in distances, changes in surfaces, changes in tracks. Much of my play is based on past training habits of people such as Whittingham, Frankel, and now Baffert, and knowing when the high end is going to send their horses. Most of the clues won't be found in the past performances or speed ratings, especially at the high end. And therefore all of my computer usage is recording results and data. I don't use a computer to make selections. I believe putting a computer in charge will have marginal results at best.

Would your methodolgy involve tote board analysis - seems like your approach follows the money trail (i.e. early money or late money)?