Hi Jeff. You have mentioned synergism on several occasions. This would be adequate. Separate the distances, don't worry about class at this point.
I usually track the median energy expenditure at all distances of a track
I handicap. The most stable numbers usually come from tracks with the longest meets. For instance, Calder, Philadelphia are a couple off the top.
I prefer Calder pricewise. handicap about 50 races at each distance.
Some distances will have a lot fewer instances because they are created by the Racing Secretary to accomodate certain barns. The distances that have
only a few instances can be eliminated from your model, or profile.
Form the 50 or so races, eliminate the extremely hi's and low's and all like
them. The extremely hi and lo i speak of, refers to the factors you are
modeling. This info is produced from checking the winners of the races and
making note of repeating factors. A predictive profile needs about 15 matches.
This having been said, a model, or profile will be relatively accurate in
3 of a nine race card. At some tracks maybe 4 races out of nine or more.
This is due to the "Matchup" in any given race. The more balanced the
race , the more predictive. Balance meaning the collective running styles.
stick to betting 2 horses per race for awhile. No exotics for now.
Hope this helps Jeff.
BobC
