|
JCapper Message Board General Discussion
|
| Home |
Register
Log In |
| By | Predicting today's variant | |
| Windoor 11/14/2012 2:48:53 PM | I have been playing with this idea for some time now. Can it be done? In software? I remember years ago when I used to go to the track with my dad, the old timers who came in late, would always ask, "how is the track playing today"? Some would simply not play the first few races, just so they could see how races unfolded. These guys put great value on what type of horse was winning the early races. Was it early or late speed getting to the wire first. With good reason. Once you know how the track is playing, you can adjust the value of the factors you are using to rank your contenders. It's not as simple as playing the best of early or late, (unless the gap is large enough) but to use the numbers to separate real contenders. Making changes to a factors value (depending on the "type of race") is something I have been doing for the last four years with great success. "The numbers have hinges". I try not to use too many performance factors for this reason. Most (All?) are very dependent on the track surface. Yes, you can see dirt, turf and synthetic, but what is the variant going to be? Is there a bias? Just how much energy needs to be expended to get around the track in a winning effort today? Is it raw speed or stamina that will have the advantage today? I know pace can play a key role here, but no one can really predict how a race will unfold. It looks good when they run like you think they will, but how many times are you left scratching your head when the fraction times are no were near what you thought they would be, and that long shot runs wire to wire? Or the front runners burn themselves out and the plodder pass's all in the stretch? In my never ending search for trying to improve the bottom line profits, I found some performance factors within J/Capper the really seems to help. At least for a while. Then it goes south and the next month looks much worse. Sometimes turning a winning month (without it) into a losing one, (with it) at some tracks. Random noise? Maybe, but I have enough data now (nearly 200K races)to make me think it is not. I am beginning to think the racing surface has indeed changed enough to take away any advantage this factor gave me in the first place. Pulling some PP's up for some races I wagered on and lost, I can see the variant number is much different than the previous months average for this track, or indeed were this horse has run well. Not all of them, but enough times to make think something is there worth investigating. Something that might help the bottom line considerably. And that is what we all should be looking for. I'm thinking of building a look-up table of tracks to get a modified DRF variant number (much like I do for class ratings) base on past history for, time of year, listed track condition, (fast, sloppy, muddy, etc) average class level for the day and a weather report for the area. As in temps, humidity, and air density. Maybe just a air density number would be enough. Assign a modified number for each, then use a look-up table to get a number to use for today. It still would not have any unusual maintenance practices, but much better than nothing I think. When that is in place, write some UDM's based on these numbers, (maybe a range of variant numbers) Plug in today's track, date, average class number, and expected temps, humidity and air density. It will be a lot of work to get it up and running, but once its done it done. Any thoughts? Regards, Windoor
| |
| Windoor 11/14/2012 4:15:50 PM | --quote: "I know pace can play a key role here, but no one can really predict how a race will unfold. It looks good when they run like you think they will, but how many times are you left scratching your head when the fraction times are no were near what you thought they would be, and that long shot runs wire to wire?"--end quote Red board alert. 11/14/2012 PMX Race 3 Horse 6. Pays 38.20. Sorry. Couldn't resist:) Windoor
|
| Reply |
|
Copyright © 2018 JCapper Software back to the JCapper Message Board www.JCapper.com |