|
JCapper Message Board JCapper 101
|
| Home |
Register
Log In |
| By | Is there a way to pull specific claiming tags? | |||
| Windoor 8/26/2011 4:24:14 PM | I finally have an Excel program that will let me change the information in one of the fields in the JCP file to identify my spot plays. Thanks to Harry (HCap) I can pull my spot plays and name them SpotA , B, C etc., as the sires name and use J/Cap to filter for them. Pretty neat, but a bit cumbersome as presently I can only change the Sires name. I am hoping I can get Owner, Siressire, Dam, Damssire to make the program more useful to me. I would think any field is doable, but I may have pushed Harry to his limits with my constant requests. I'll let him rest for a while:) I can use this for what I am asking but would have to build a database for every class/claiming level. A tedious task and means duplicating many and more records. I see a 5K claimer much differently than a 25K claimer and treat them accordingly. Is this capability there and I just don't see it? I do see a purse value, but that will not work for what I want. Or do I need to buy a 2 terabyte drive and build databases for the next few weeks? Regards, Windoor.
| |||
| jeff 8/27/2011 2:08:32 PM |
Just curious, why the need to change sire (or dam, damssire, siressire, or owner) names in the data file? (as all are currenlty available for use in JCapper UDMs.) Switching gears, numeric claiming price isn't (currently) a grab-able factor for UDMs in JCapper. Why not? In my R&D, I've always used large sample hit rate and flat win bet roi to evaluate how horses within a given race that have previously raced under a variety of classes/conditions stack up against each other class-wise. That R&D has always indicated other class based factors such as RaceStrength, ClassRating (based on pars), ClassShift (based on purse value), CMI, and more recently ClassConsensus (which are all grab-able for use in UDMs) actually do a better job in capturing "the essence of class" (if there is such a thing) than numeric claiming price. In short, that is why claiming price numeric value has never been made grab-able for use in UDMs. What I have not done in my R&D is evaluate the strategic effectiveness of using numeric claiming price as a way to segregate races by race type... although I have used Track Code, Distance, intSurface, Purse, ClassDescriptor, FieldSize, RV, PaceIndex, PacePressure, and ESPRaceShape, etc. to do that with a reasonable degree of success. That said, numeric claiming price DOES represent the intuitive as a way to latch onto class. Also, within the past year several people have asked me to make it grab-able. Windoor, tell you what - I'll take a look at the inner workings of the Profile Marker and take a run at making numeric claiming price grab-able for use in UDMs. If you will let me know whether you are operating in playlist file mode or sql mode - I'll make every effort to enable it in that mode first. -jp . ~Edited by: jeff on: 8/27/2011 at: 2:08:32 PM~
| |||
| Windoor 8/27/2011 6:53:01 PM | I have an Excel program than can filter for my spot plays and give the name and race ID of my contenders on the result sheet. In all, I have an Importer, a Filter, a XRD file changer and a JCP file changer. I take this information and have the JCP file changer look for these horses in the JCP file. When it finds a match, I have it use a field that I don’t need (such as sires name) and have it change to something I can filter on. Such as “SPOTE”. Then I can pull all of my contenders by having J/Capper look for SPOTE as a requirement for the UDM while also having use of all of the other filters in J/Cap. It has proven to be quite useful if a bit convoluted. Especially for daily play. While claiming price may be an indicator for class, I find it has more value to me as a trainer/owner’s intent for today’s race. There are higher purses for a lower claiming price at times and the decision to not risk “this horse” to a claim, gives some indication as to when the horse is ready for a good effort. I have a form cycle test that bears this out. I am still in playlist mode at the moment. I really want to learn the SQL as it appears to be a superior method and I use many “IF, AND, OR” statements in my own rules for spot plays. I just can not find the time to make the switch, and I am getting good results with the playlist mode so far. Any help that can save me some time and hard drive space is greatly appreciated. My original idea was to have you do a custom add on within J/Capper to allow me to filter on the factors I require. (See earlier posts on the subject) Then I would not need any of the Excel programs except for record keeping, but I think it might be beyond the scope of the software. If you think you might be interested in such a project, I could send you a copy of the Excel importer and filter I am using. The filter includes the basic requirements for one of my spot plays with the capability of breaking down the four “OR” statements in the rule. Listed as A, B, C , D. Along with a bypass switch to test for other plays. If not, that's ok, I know your a busy guy. What I have will work, it just not very efficient when it comes to time and hard drive space, and I still need to fine a way to change the target field in the .JCP file. Regards, Windoor.
| |||
| Windoor 9/6/2011 7:41:26 PM | As I find myself building databases for six months of data, my machine takes about 3.5 hrs to do six months of all tracks. Is this normal? My machine is fairly new with Win7 Pro 64 bit with X86 support. Eight core Intel 2.8 MHz processor and 6 gigs of ram. Hard drives are one 300 Gig 10K Raptor and 1 Terabyte at 7500 rpm’s. It takes my Excel importer program about an hour to do six weeks of races at a time. I can’t get a full quarter (3 months) into Excel due to the 64K line limit. Six weeks is about 46K lines At this rate it will take about 168 hours of number crunching to get all of databases built for the research I want to do for every six months of racing. I really want at least two years of consecutive data at the minimum. Run the PC for about one month at 24/7? I think I will build another PC just for importing files and database building. Regards, Windoor.
| |||
| Charlie James 9/6/2011 10:13:42 PM | "Is this normal?" Umm... no. I'm running in sql mode -- running db builds in mode 5 -- where the db builder first scans the target folder for race card files not yet in the database for my current folder. After identifying new data and results files to be added the db builder then appends new race card files to the database. 8 days out of 10 I do this as part of my daily routine. It takes maybe 5 minutes each day.
| |||
| ryesteve 9/7/2011 10:40:44 AM | I think you two are comparing a full rebuild with an append. Windoor, you're not doing a rebuild from scratch every time, are you?
| |||
| Windoor 9/7/2011 5:43:49 PM | Yes, from scratch every time. I have to do this in order for me to use J/Cap filters on my spot/angle plays. J/Capper by itself does not have the capability to filter for these horses at the present time. I use an Excel program on the JCP files to find the horses that qualify as a spot/angle plays. I then use another Excel program to change the name of the sire (to spot A, spot B etc.) in the JCP file for every horse that qualifies as a spot/angle play. This way I can use Sires name as a filter within J/Capper to flag these horses. Hopefully Jeff will find a way to keep from having to do this for claiming tags too,(see above) or I'm going to need a much bigger hard drive. Unfortunately I have to do this for every basic play and build a separate database for each one. Very time consuming and many, many gigabytes, but until I can find an easier way to filter for these horses this has got to do. The good new is, that it works, and has been very successful in creating profitable UDM's. Will they hold up over time??? That remains to be seen. Regards, Windoor.
| |||
| ryesteve 9/9/2011 10:22:11 AM | I understand that you're doing some extra things with the data, but I still don't see why you need to recreate from scratch each time. Once you've built the data to fit your needs, why couldn't you just process the incremental data each time?
| |||
| Windoor 9/11/2011 12:02:33 AM | Yes, you are correct. After I build a database for “SpotA” horses, I can do the incremental thing every month to add to it. What I meant to say, was that I have to build a database for each spot play, and for each variation of same if I use a filter outside of J/Capper. It looks like 3.5 hrs for every six months worth of data. Does that seem right to you? For each one of my Key Factors (I have about a dozen) that Excel filters for, I have to build a new database “from scratch” within J/Capper. And if “SpotA” is tweaked a little for a different class or distance (Pick one of the seven) with filters outside of J/Capper it now becomes SpotA2, SpotA3, etc. Again a new database for each one is needed. I am now up to SpotG (naturally I call it G_Spot) with no end in sight. Regards, Windoor.
|
| Reply |
|
Copyright © 2018 JCapper Software back to the JCapper Message Board www.JCapper.com |