|
JCapper Message Board JCapper 101
|
| Home |
Register
Log In |
| By | Basic Handicapping | |||||||
| Caveat 4/1/2013 8:55:15 AM | Hi Guys... How do you determine what posts are bad in this situation..Do you go by wins,roi,pct,or impact values?
Thxs Mike
| |||||||
| AndrewH 4/1/2013 8:59:05 AM | I'd personally go by impact seeing how the factor is post position. If the factor was a factor that I was using to find value then I'd go by ROI but in this case you are trying to handicap instead of find value. I'd also like to hear what others have to say. AJ
| |||||||
| jeff 4/1/2013 12:46:04 PM | Win rate, impact value, and roi... those are just measurements programmed into an app like the Data Window used to report attributes found in a data sample. How the player interprets a sample - that can be everything. Whenever I look at a data sample - I find myself constantly asking the following question: "How does this data sample compare to the big picture?" More often than not - I find that the act of contrasting the individual sample against the big picture - enables me to interpret the individual sample with better clarity. Win rate, impact value, and roi?... Each of those measurements helps me get there. Here are some big picture rail position results pulled from the Data Window... First up - all dirt races no matter what the track code or distance over the past 365 days:
Next up - all turf races no matter what the track code or distance over the past 365 days:
Using each of the available metrics, win rate, impact value, and roi - when you compare your sample to the above big picture samples, what do you see in terms of rail position? I see a dead rail. -jp . ~Edited by: jeff on: 4/1/2013 at: 12:46:04 PM~
| |||||||
| Caveat 4/2/2013 7:40:50 AM | Thxs Andrew , I was see it the same way Jeff , I don't get your response here... The statement that the fastest time between 2 points is the shortest distance traveled....so its obvious that post 1 will be the fastest, post 2 is the second fastest, post 3 is the third fastest, etc, etc.. But all tracks are not created equal.. According to your stats, post 1 is "dead" OK But there are tracks where post one is golden , post 2 is terrible, outside posts are better ( Tampa) and so on.. But not doing R & D wont make you aware of track tendencies. And when one of your "universal" UDM's pop one up, how would you bet it pertaining to the track bias's? Here's March.. Post 1 on dirt
Here's Post 2 on dirt
Here's post 3
So betting post 2 plays would have cost you in March.. This game is so difficult..:( What I was considering to do was to do some sort of R & D for whats happening now as compared to the big picture. I'm lost mike ~Edited by: Caveat on: 4/2/2013 at: 7:36:42 AM~ ~Edited by: Caveat on: 4/2/2013 at: 7:40:50 AM~
| |||||||
| jeff 4/2/2013 1:13:13 PM | I'm guessing that the rail position stats in post #1 at the very top of this thread are track specific. Going with that assumption... Purposely setting background color for the inner 3 posts in red to make them stand out - here's a cut and paste of those rail position stats again:
Purposely setting background color for the inner 3 posts in blue to make them stand out - here's a re-post of the dirt surface rail position stats from the big picture sample that I posted above:
Now, compare the track specific (or small picture) sample for posts 1-3 (in red) to the the big picture sample for posts 1-3 in blue. How does the small picture sample compare to the big picture sample? The small picture sample for posts 1-3 shows markedly lower win rate, impact value, and roi than the big picture sample. If the small picture sample is track specific, the first question that needs to be answered is: Why? Why is there such a disparity between the two samples? Are the numbers in the small picture sample simply the result of small sample "noise?" Or is there some physical cause behind the difference? Often times, I can get a sense by watching replays... especially the head on. Is there a dead rail at the track where the small picture sample was taken? Going forward with the idea that I do in fact think the track where the small picture sample was taken does in fact have a dead rail... You asked: "And when one of your "universal" UDM's pop one up, how would you bet it pertaining to the track bias's?" Great question. Suppose I have a speed-pace-form-cxn-value business UDM (universal in nature/not track specific) that produces good results overall. In addition, I also have several negative layering UDMs that are track specific - and some of these negative layering UDMs are designed to do nothing more than point out horses unlucky enough to have drawn an inner post at a track-surf-dist configuration where the small picture stats for rail position are similar to the numbers posted in red above. Let's further suppose a look at my html report reveals a horse flagged by both types of UDMs. The business UDMs tell me the horse has good speed-pace-form-cxn and potentially value... But the negative layering UDMs tell me the value part might be questionable because the horse will likely be compromised by the course layout... • How best to approach this? Every single one of us has probably encountered this situation - or a situation nearly identical to it - many times in the past. What do my wager history records indicate? FYI, this is the approach that I've adopted over the past 15 months. I can easily run the intersection of Business and Layering UDMs through JCapper's WagerHistory Module and break the data out by nearly every data point available in the Data Window. From this I gain the ability to (hopefully) make an informed decision ahead of time how I will act when I am confronted by this (or similar) situations on race day. Based on the performance of my own UDMs - at the tracks that I bet - the numbers in my WagerHistory table suggest that my best course of action in this situation is to: 1. Require higher odds than I normally would if the horse "fit" the track profile. 2. Scale bet size down a little (in accordance with those higher odds.) -jp . ~Edited by: jeff on: 4/2/2013 at: 1:13:13 PM~
|
| Reply |
|
Copyright © 2018 JCapper Software back to the JCapper Message Board www.JCapper.com |