View Full Version : Hanicapping Magic - programming
Lou G
05-14-2005, 08:18 PM
After reading HM, I wrote a program to create the necessary numbers using bris single-file data and bris pace figs instead of raw times. I began using it in September, 2004 and have had very good results. I know there are others out here doing more or less the same thing and for them I have a question:
How are you determining "pressure" in your code?
This area has given me more fits than any other aspect of the approach. I've tried at least a dozen ways, some of which are promising, but have yet to find a solid answer. If anyone has a solution that they would be willing to share I would appreciate it and will gladly share my code and ideas.
Thanks,
Lou
Lou G
05-14-2005, 08:19 PM
hanDicapping...
mhrussell
05-14-2005, 09:59 PM
Lou-
Nice to see another person trying this alternative approach.
As far as pace pressure, I will refer you to some of the earlier "All Ways" newsletters (you can get to them via the BRISnet website) and also Randy Giles papers on the subject. For a numerical pace pressure calculation in my code, I use the following: for all the E and E/P style horses in a race, sum the Qurin speed points that are 5 or more. A total less than 19 leans toward the 'early horses' as having the advantage <less pressure>; 20-21 is ~neutral and >21 tends to be pressured and gives the advantage to the closing horses.
There are a number of ways to do this and I highly recommend you read the All Ways newsletters along with Randys published stuff.
Good luck to you.
best,
Matt
mhrussell
05-14-2005, 10:03 PM
Lou-
You may also want to check out the Handicapping Magic user board that Kingfin has set up on Yahoo:
http://sports.groups.yahoo.com/group/handicappingmagic/messages/
It looks like this will be another good forum for us 'magician's to exchange ideas in addition to PA.
best,
Matt
Lou G
05-15-2005, 10:08 AM
Thanks Matt - your method is intriguing. One thing I tried that worked pretty well is similar but not as detailed:
Count the "E" horses in the race. Multiply by 10. Add the number of "E/P" horses. The result gives some insight into the race structure e.g. a fig of 20 shows two front-runners who may be able to have the lead to themselves; a fig of 25 may indicate a pace battle in the second fraction.
I'll definitely code your approach and run it through my database.
Lou
headhawg
05-15-2005, 11:20 AM
Lou and Matt,
Do you find the Bris run style close enough for your formulas or do you adjust them? I sometimes wonder how Bris determines some of them as I don't see the styles for many horses in quite the same way that they do.
I always liked the concepts in HM but couldn't make them work for me. Good stuff here, though.
HH
BRIS running styles are worthless in many, many races. I never use them, I don't even use the ones in HTR. My advice, make your own calls - you will never regret it. Your doing this exercise wilkl in itself gived you insights into the horse's from cycle.
Caeat:
Unless you are using Sartin advanced programs, where running style is not the same thing we are talking about here.
From Walter Coffey, the running style designations:
Front runner - 1-2-3 at the fist call in two of the last three races
Closer - in the rear half of the field in two of the last three races
Middle runner - everything else
I have modified this somewhat, requiring the horse to be 1-2 at the first call in two of the last three representative races - this might mean crossing our several pacelines that do not apply to toda'y comditions.
F M C breaks the race down enough so that when I analyze the F horses, I differentiate between the need to lead types and ealry presser types.
I am also experimenting with redefining Quirin speed poinst as well. I don't like the idea that a horse running on the lead by open lengths gets rated the same "2" as a horse running third by 2 lengths. I am requiring a horse be no more than a length back in order to get points when running third. So 3 by 4 gets 0 points, and a horse running 3 by a half length gets 1 point
mhrussell
05-15-2005, 12:02 PM
Tom is right. The BRIS running style designations are not very accurate. I have found that ~20% of the time, they are just WAY wrong. But the BRIS Quirin speed point assginments are accurate.
Several months ago, CJ shared his running style determination code in a separate thread and I have incorporated that and it has been working very well (thanks again CJ!). I would recommend you code up his algorithm and use that in combination with the BRIS Q points.
Again, I found the "All Ways" newsletters and Randy Giles documents on this topic to be very helpful and informative.
Lou G
05-15-2005, 01:57 PM
HH - I've looked at the bris run style and found myself shaking my head a few times...
Matt and Tom - thanks for the tips, I'll research further.
Lou
Lou G
05-15-2005, 02:54 PM
Matt - I did a few searches and was unable to find the cj post you mentioned. If you get a chance, could you point me at it?
Thanks
P.S. - joined the yahoo board, looks interesting
mhrussell
05-15-2005, 04:07 PM
Lou-
GameTheory posted a thread called "Running styles" back on 2-12-2005.
His post #17 in that thread provides a general "english" version of CJ's code logic which can be used to program in any computer language.
Just do a search on "Running styles" and you will find the thread.
NoDayJob
05-15-2005, 08:33 PM
http://www.paceadvantage.com/forum/showthread.php?t=2148&referrerid=0
NDJ
From Walter Coffey, the running style designations:
Front runner - 1-2-3 at the fist call in two of the last three races
Closer - in the rear half of the field in two of the last three races
Middle runner - everything else
I have modified this somewhat, requiring the horse to be 1-2 at the first call in two of the last three representative races - this might mean crossing our several pacelines that do not apply to toda'y comditions.
F M C breaks the race down enough so that when I analyze the F horses, I differentiate between the need to lead types and ealry presser types.
I am also experimenting with redefining Quirin speed poinst as well. I don't like the idea that a horse running on the lead by open lengths gets rated the same "2" as a horse running third by 2 lengths. I am requiring a horse be no more than a length back in order to get points when running third. So 3 by 4 gets 0 points, and a horse running 3 by a half length gets 1 point
Tom,
Q PTS seem to be very generous with sprinters going to a route. Also I check the last race at todays distance and if the horse has a bad q pt race you might want to flage wither of these two.
Doug
vBulletin® v3.8.9, Copyright ©2000-2024, vBulletin Solutions, Inc.