jeebus1083
12-24-2006, 02:06 PM
The last few months, I have been using a formula with the BRIS speed points where you add the whole field up and then divide an individual horse's speed points by the total amount of the field. Example: if the speed points of every horse in a field added up to 30, a horse that scores an 8 speed point (most likely to want or be on the lead), the horse would have a 26.7% or roughly 5/2 chance of being on the lead.
Recently, I started breaking down Prime Power #s into a 0 through 8 ranking system, taking the difference between the high number and the low number and dividing it by 9. Example: if the #1 ranked horse has a 120 Prime Power rating and the lowest has a 100 Prime Power rating, I get the difference (20) and then divide it by 9 (2.22). The scale would look like this:
0: 100.0-102.2
1: 102.3-104.5
2: 104.6-106.8
3: 106.9-109.1
4: 109.2-111.4
5: 111.5-113.7
6: 113.8-116.0
7: 116.1-118.3
8: 118.4-120.6
Now, rather than ranking Prime Power by highest and lowest, now each Prime Power number has a point value attached. Like the speed points, I add the whole field's Prime Power ranks together and then divide each individual horse's number by the total field number for a Prime Power percentage. The next step is to add the Speed Point percentage and the Prime Power percentage together, divide by 2 for the average, then divide 100 by the average and subtract 1 for the odds.
My trouble with the method is this: it is time-consuming. If there is a scratch, my math has to be re-done to adjust for the scratches. With the Prime Power, I'm not sure if it's more accurate in the event of a scratch to re-rank the field (if the scratch happened to be the top or bottom Prime Power horse), or work with what I've got. The program that I want to create would allow automatic adjustment for a scratch. The big problem is that this method is not workable in maiden races where data is insufficient (not enough starts to warrant a rank). I understand that old-school handicapping is always the way to go, but this is strictly a value guide to examine horses that could be borderline contenders that don't necessarily look good on the surface. It forces me to give a 2nd look, and I can eliminate horses whose odds are too high on the value line (add the field's value odds and divide by field number, the median is the cut-off point. Any horse above the given median is subject to elimination unless old-school capping proves otherwise).
Recently, I started breaking down Prime Power #s into a 0 through 8 ranking system, taking the difference between the high number and the low number and dividing it by 9. Example: if the #1 ranked horse has a 120 Prime Power rating and the lowest has a 100 Prime Power rating, I get the difference (20) and then divide it by 9 (2.22). The scale would look like this:
0: 100.0-102.2
1: 102.3-104.5
2: 104.6-106.8
3: 106.9-109.1
4: 109.2-111.4
5: 111.5-113.7
6: 113.8-116.0
7: 116.1-118.3
8: 118.4-120.6
Now, rather than ranking Prime Power by highest and lowest, now each Prime Power number has a point value attached. Like the speed points, I add the whole field's Prime Power ranks together and then divide each individual horse's number by the total field number for a Prime Power percentage. The next step is to add the Speed Point percentage and the Prime Power percentage together, divide by 2 for the average, then divide 100 by the average and subtract 1 for the odds.
My trouble with the method is this: it is time-consuming. If there is a scratch, my math has to be re-done to adjust for the scratches. With the Prime Power, I'm not sure if it's more accurate in the event of a scratch to re-rank the field (if the scratch happened to be the top or bottom Prime Power horse), or work with what I've got. The program that I want to create would allow automatic adjustment for a scratch. The big problem is that this method is not workable in maiden races where data is insufficient (not enough starts to warrant a rank). I understand that old-school handicapping is always the way to go, but this is strictly a value guide to examine horses that could be borderline contenders that don't necessarily look good on the surface. It forces me to give a 2nd look, and I can eliminate horses whose odds are too high on the value line (add the field's value odds and divide by field number, the median is the cut-off point. Any horse above the given median is subject to elimination unless old-school capping proves otherwise).