Quote:
Originally Posted by o_crunk
As an aside, my belief is their most valuable asset is not so much the handicapping aspect. Everybody and their mother can come up with reasonably accurate pricing from just the handicapping data. It's not that hard a task. The hard thing is spitting out the bets in relation to the public and computing it all as all the pools come to a close. Basically there's 2 types of data - the data about the horses (handicapping) and the data about the betting (tote). By far the biggest upside for teams is dealing with the betting data - the general public is very poor at understanding it and the teams can deploy much more robust and speedy calculations than Joe Blow can. The handicapping data is not as important as many think. Having good handicapping data is like a quarter of the battle.
|
You are probably right.
When I was at DRF, I was brought in on a meeting with some Ivy League statistics/computer experts. These guys said they built an automated self learning model that was beating the horses consistently around the country. They didn't want to give up that model, but they thought DRF could use aspects of it to create metrics from the betting pattern data that could then be used to help with marketing DRFBets to improve handle and market share. That's what they were trying to sell. I asked a LOT of questions. From what I could gather, they were mostly monitoring pool data and betting patterns. Their own handicapping knowledge was not what I would call extensive even though it was part of the model.
