By |
trainer - rider win percentage |
STUBALL 2/9/2014 7:32:14 AM | I am adding trainer win percentage and rider win percentage to my factors (f34 and f35)and when I rebuild the database I get no rank or gap for either 1 of these fields....(What did I do wrong?) could it be in the settings?
Stuball
|
Caveat 2/9/2014 9:31:29 AM | its very tricky ( for me) to make changes doing this
But maybe this will help
http://www.jcapper.com/Article_AddJRating.html
|
STUBALL 2/9/2014 10:45:23 AM | Thanks for the help but I know how to do it...my problem is that the results have no rank or gap...just the win percentage shows in the val col. can't figure out why..
Stuball
|
Caveat 2/9/2014 11:02:13 AM | Stu..
go to SCREEn #4 in the set-up factor tool...
choose your slot #
and there's a check box for "RANK" & "VALUE"
i dont know if you can get gap..
~Edited by: Caveat on: 2/9/2014 at: 11:01:20 AM~
~Edited by: Caveat on: 2/9/2014 at: 11:02:13 AM~
|
STUBALL 2/9/2014 7:03:45 PM | That is for the html report......I am talking about what is in the fields in the starterhistory file....I guess I did not make myself clear.....it won't help to display what is not there..it has to be in the fields for trainer win percent and rider win percent.... I appreciate your help but this is not the answer I am looking for...Thanks
Stuball
|
jeff 2/9/2014 9:05:09 PM | Here is what you need to know...
The part of the program that deals with Rider and Trainer Win Pct is "older" than other parts of JCapper. (It was written in the late 1990's... ancient compared to the algorithms I would write later.)
Back then I hadn't yet developed the ability to (programmatically) rank and gap a field of horses. (I hadn't even come up with the concept of stored UDMs back then.) The only thing I wanted at the time was the ability to break the data out by Rider and Trainer Win Pct numeric value.
Years later, long after that early home grown program had evolved/become JCapper... and long after I added the ability for the user to create and store UDMs - and with that the ability to rank and gap fields of horses for 100's of different factors...
I revisited the idea of adding rank and gap for Rider and Trainer Win Pct and making it "grab-able" in UDMs.
Unfortunately, the way that I built the original part of the program (back in 1997-1998) that deals with rider and trainer win pct was done in such a way that modifying it - and modifying it in such a way that rank and gap for rider and trainer win pct could be added to UDMs like all of the other factors that came later meant I was going to have to burn a SIGNIFICANT number of programming hrs in order to get that done.
My notes show that I made a command decision at the time.
Instead of burning a week or two of programming hrs, I decided to burn 4 or 5 hrs instead... and I added Rider and Trainer Win Pct to the list of factors enabled in conjunction with the 3 Translate Behavior in the area of the program known as UPR Tools.
It's a quirk in the way the program evolved...
But to the best of my knowledge, the ONLY way to reach out and "grab" rank and gap for Rider and Trainer Win Pct so that you can use either in UDMs is through the 3 Translate behavior in a UserFactor.
As in create a new UserFactor where:
• Behavior is 3 Translate
• RIDERWINPCT (or alternately TRAINERWINPCT) is selected from the Factors Drop Down (or alternately) keyed in manually as your factor.
• One of the following is keyed in for surface: *, D*, D, d, T*, T, or t.
-jp
.
~Edited by: jeff on: 2/9/2014 at: 9:05:09 PM~
|
georgekos 2/24/2014 1:46:33 PM | Another matter on Win% and Trainer%....
Says Data is for "Current Meet".. so on Day1 of Meet is everybody 0%? or is there a way to tell how many Races are being used in %Calcs?
Thank you georgekos
|
jeff 2/26/2014 8:56:50 PM | Didn't see your post George until just now... I don't know the exact answer off the top of my head. (I recall there being a unique way in the Equibase data that a "meet" was defined.) I'll dig through my notes but will probably need to contact HDW first before posting the answer...
-jp
.
|
jeff 3/1/2014 9:50:54 AM | Link to field mapping for .JCP Files: http://www.jcapper.com/MessageBoard/Reports/JCPFieldMap_08052010.txt
The following is a cut and paste from the above linked to field mapping doc:
--quote:" *** Today's Horse/Trainer/Jockey/Owner ***
28 Today's Trainer CHARACTER 30 29 Trainer Sts Current Meet NUMERIC 9999 4 30 Trainer Wins Current Meet NUMERIC 999 3 31 Trainer Places Current Meet NUMERIC 999 3 32 Trainer Shows Cureent Meet NUMERIC 999 3
33 Today's Jockey CHARACTER 25 34 Apprentice wgt allow.(if any) NUMERIC 99 2 35 Jockey Sts Current Meet NUMERIC 9999 4 36 Jockey Wins Current Meet NUMERIC 999 3 37 Jockey Places Current Meet NUMERIC 999 3 38 Jockey Shows Current Meet NUMERIC 999 3
" --end quote
Opening day for Saratoga's 2013 meet was July 19, 2013. I loaded the data file for that race card into the program and generated past performances.
Here's a link to a screenshot showing the rider and trainer stat tables for the #3 horse PREDICATE in R3 run at 5.5 furlongs on the dirt: Screenshot - click here-
Next I checked the data fields internally within the .JCP file by loading the SAR0719.JCP file itself into Excel 2010, scrolled down to the row for PREDICATE in R3, and then scrolled left to right so as to view field numbers 28-32 for trainer and field numbers 33-38 for rider. Here's a link to a screenshot of that: Screenshot - click here-
Clearly, the "meet" stats for rider and trainer on opening day are not 0's.
Of course that raises the question: What are they?
A review of my notes shows I was given a choice (from among different info models) how the "meet" stats inserted into JCapper data files would be compiled.
1. Start each new meet over with 0's and build from there.
2. Equibase or HDW to compile a rolling 365 days of rider and trainer stats for each track code.
There were other options - and a reasonable case can be made for each of them...
However, I chose option #2 above.
If you compare the two screenshots above you should be able to confirm that the stats from the .JCP data file displayed in the Excel spreadsheet synch up with the track code stats in the bottom row of the stat boxes for rider and trainer.
For the final verification step I ran the following sql expression through the Data Window with the data broken out by both rider and trainer:
SELECT * FROM STARTERHISTORY WHERE TRACK='SAR' AND [DATE] >= #07-20-2012# AND [DATE] <= #09-03-2012#
Note that July 20 2012 was opening day for the SAR 2012 meet and that Sept 03 2012 was closing day for the same meet.
Here's a link to a screenshot of the Data Window where I've scrolled down through the indivdual trainer names until Mark Hennig appears as the name at the very top of the Data Window: Screenshot - click here -
Here's a link to a screenshot of the Data Window where I've scrolled down through the indivdual rider names until Rajiv Maragh appears as the name at the very top of the Data Window: Screenshot - click here -
Note that everything almost synchs up. (The Data Window BY TRAINER screenshot stats synch up exactly. But the BY RIDER screenshot shows 1 less win and 4 more starts for Rajiv than the screenshot based on the Equibase stats. Not sure what's going on there but it's close. I'm going to ask and will post an answer if and when I get one.)
Now you know what you are looking at. A "meet" in this case is defined as rolling 365 days at today's track code.
-jp
.
~Edited by: jeff on: 2/28/2014 at: 9:06:49 PM~
~Edited by: jeff on: 3/1/2014 at: 9:50:54 AM~
|
georgekos 3/1/2014 10:04:39 AM | a full rolling Year is ABSOLUTELY the PERFECT thing I want.........****** !!!!!! Nice work
Im finding a few Trax where the Public is very Jockey/Trainer susceptible and OVERBETS them....
Thak you from Poolside in Florida
georgekos
|