Bris Prime Power Rating

[karlskorner]

Bris Prime Power Rating had Monachos at 156.0 and Point Given at 154.9. The rest weren't even near.

Looks like Fortune Hunter and I are the only 2 board members that bothered to read the print outs.

Karl

[Rick Ransom]

Be careful about playing those top Prime Power ratings all the time. You'll probably find out that the 2nd and 3rd rated do better overall (ROI wise). Their ratings suffer the same problem most do. They include too many dependent factors which cause the top rated horse to have a high win percentage but a bad ROI. Most handicappers have the same problem. When you try to go over about 29-30% winners you start hurting your ROI. Check it out. Nevertheless, congratulations on your winner. You got a good price on this one!

[Que]

Rick,

Interesting observation about ROI and win percentages. Last month I was experimenting with a rating system that give me a class rating for each of the horse's previous races. The method merely looked at the horse's finish position and the class/conditions of the race. The rating was computed by multiplying a weight assigned by the finish position to a weight assigned to the class/conditions. A genetic algorithm was used to derive the optimum weights, based on the results of over a 100 races at the same class/conditions. The interesting thing is that I optimized the weights once for win percentage, then again by ROI. The result was the weights for the class/conditions remained virtually the same in both studies, but the weights assigned to the finish positions varied significantly. Specifically, when optimized for win percentage the better the finish the higher the weight, but when optimized for ROI the worse the finish position the higher weight. For example, when optimized for ROI, the system would gave a higher class rating to horse that finished 8th in a NW1 than it would to a horse that finished second in a NW1--totally opposite of what you would expect. Now I not saying to bet horse's with the worst finishes at similar class/conditions, because in this study the class ratings were then added to a pace rating to arrive at the predicted order of finish--but I do believe the study indicates that when analyzing class you may need to use a contrarian approach in order to optimize your ROI (vice your win percentage). Anyway, excellent observation and it coincides with my limited research.

regards, Que.

[Rick Ransom]

I think it might be because where the connections enter the horse tells you something about how good they think he is, and they're usually not idiots. Now the horse may do better or worse than what they thought, which affects the public opinion more, but it's only one race. There is so much "noise" in horse racing performances that you have to have to give some weight to other things like this.

Sometime I'd like to study what variables are good at predicting actual odds. What little I have looked at suggests that about 80% of the variance in actual odds is explained by morning line odds. Jockey performances might add a little to that. So much for "smart money". I think if you could find a good set of "dumb money" variables, you would be better off using this to find your overlays than actual odds.

[Que]

Rick,

I think that's what Glenny's program does, although I haven't had a chance to reasearch it myself. You can sometimes find Glenny in the War room. Later....

[Rick Ransom]

A few years back I had the problem mentioned before with my 2nd rated doing better than 1st rated. I found that I did best if I played the one with the lowest average earnings. Highest morning line odds also worked, but lowered my win percentage too much. Since then I have improved my method and don't need to use that anymore. Morning line odds are quite good at some tracks (I play Southern California and New York) and it was pretty tough to come up with factors that are not already included in it. I'd like to see morning line odds eliminated and see how smart the public really is without any outside input. The local track here, Turf Paradise, used to have some pretty lousy morning line odds and was easy to beat despite the high takeout.