|
JCapper Message Board JCapper 101
|
| Home |
Register
Log In |
| By | How to find class changes within... | |
| busseb 8/3/2009 5:34:27 PM | Been sitting here trying to figure out how to flag class changes when a horse is moving from a C5kN2L to a C5kN3L or maybe a C5kN1Y when the purse stays the same. Same with moving from a C5kN4L to a C5kN2Y. At MNR and WOX, the 2 tracks I play, there is a significant class change within the same claiming price. Any ideas? ElPaso ~Edited by: busseb on: 8/3/2009 at: 5:34:27 PM~
| |
| jeff 8/3/2009 9:20:41 PM | What you are asking for is purposely not in the program. Very early on (sometime in the late 1990's if memory serves) I did significant R&D into the exact type of thing you are asking about: horses fitting / not fitting / moving to and from various types of listed race conditions... N2L, N3L, N4L, N2Y, N3Y, N4Y, etc. I decided not to pursue it based on my R&D. ***This type of thing was very much already factored into the odds. ***An analysis of my best performing CPace and Form based UDMs at the time revealed that most if not all of the overlay winners driving the profit for my UDMs were horses that were race condition failures or misfits. For example: A horse with 2 lifetime wins entered in a N4L race today. ***During the time when I was actually studying listed race conditions, it wasn't hard to find several races each week where a comparison of the listed conditions and the records of the entrants didn't make sense. Keep in mind I was only following one circuit at a time closely back then. For example, the listed conditions might say N2L. But a glance at the lifetime records of the entrants might show 3 or 4 of them to have 2 lifetime wins. IOW, one or more horses didn't appear to qualify without an interpretation of whatever followed words like "other than" or "since." Once you applied "other than" or "since" then how some of them got into the race made sense. However... When horses from the original race would come back to run a few weeks later, the "other than" and "since" parts were almost always ignored in the symbols designating that past race. For example, the original race might have been listed as an N2L other than - and a careful analysis of the records of the horses entered would show that 3 or 4 of them had 2 lifetime wins. But fast forward to today's race which might be listed as N3L. A look at the symbol for the original race would only show N2L - which might very well lead one to believe a class change had taken place - when in fact a class change had not taken place at all. Even today it's still a common and misleading occurance. I recently had a conversation with someone considering JCapper who uses a competitor's program that uses HDW data files. We got around to talking about the fields in the HDW files that deal with the listed race conditions. I happen to know a little something about them because I'm working on a file import module that will soon make HDW files available to those JCapper users who want to use HDW as a data source. He wanted to know if the Bris/TSN fields were any more accurate than the HDW fields - which he thought were inaccurate for the very same reasons that I just described in the above paragraph. I explained that it's not a Bris, TSN, or HDW issue... but that Equibase has to use SOMETHING as a starting point - Looks to me like it's standard operating procedure to just perpetuate that race as forever being N2L despite the makeup of the field - some N2L races might actually have 5 starters in them with 2 lifetime wins who managed to get in because of the words "other than" or "since." Other N2L races might have not have any. Instead of going down the road of interpreting race condition symbols - I found far more bang for my buck (and I'm convinced an easier path to success) evaluating Class using factors such as: Class Rating - based on par for the level. Race Strength - based on speed figures earned by race winners each horse has previously faced. CMI - which is proprietary. Keep in mind that how the horses in today's race stack up vs. each other in terms of Class is more inportant roi-wise than what you will get from class shift evaluations. IOW class shift is already reflected in the odds to a greater degree than differences in class among the entrants of today's race. But class shift does have its place. And of course class shift is often worth adding as a factor constraint to UDMs. JCapper gives the user/player the ability to flag changes in class (class shift) from one race to the next using factors like purse value, claiming price, class descriptor, state bred vs open, Par vs Race Strength, and even Weight Shift which oddly enough can be related to changes in class. It isn't that I'm dead set against using listed race conditions like N2L to make class evaluations - it's that I've found better bang for my buck spending development hours in other more promising areas. I hope I managed to explain that in a way that makes sense - without sounding too harsh. -jp . ~Edited by: jeff on: 8/3/2009 at: 9:20:41 PM~
| |
| busseb 8/4/2009 10:07:07 AM | Not harsh at all. All I want to do is make the best UDM's possible and if I'm looking at non-significant factors, I appreciate the nudge in the right direction. Wrong and broke isn't the choice I'm trying to make. Thanks!! ElPaso
| |
| jeff 8/4/2009 4:24:00 PM | ElPaso, Check the private area of the site later tonight - or maybe sometime tomorrow. There's a thread about what to do when a UDM goes south. I'm trying to cut a video illustrating a simple yet viable approach to modeling and exploiting synthetic surfaces. I plan to post the video in that thread. Very much worth a look if you ask me. < G > -jp .
|
| Reply |
|
Copyright © 2018 JCapper Software back to the JCapper Message Board www.JCapper.com |