For someone subscribing to the theory that the more standard deviations away from the consensus the Vegas line is the higher the probability it's the Vegas line that's wrong...
Would you expect an increased utility to the gambler if someone were to only use the predictive systems from Todd's page tbeck.freeshell.org/fb/predncaa.html
and exclude all the retrodictive ones when getting the predicted average and it's Standard deviation?
If you answered no, would it just be because you think the smaller sample size alone would hurt the accuracy of the estimation or do you see other value the retrodictive systems add?
Would you expect an increased utility to the gambler if someone were to only use the predictive systems from Todd's page tbeck.freeshell.org/fb/predncaa.html
and exclude all the retrodictive ones when getting the predicted average and it's Standard deviation?
If you answered no, would it just be because you think the smaller sample size alone would hurt the accuracy of the estimation or do you see other value the retrodictive systems add?
