JCapper Message Board

JCapper 101
-- Tossout preset filter

Home	Register Log In

Tossout preset filter

pabennett
1/22/2009
10:36:45 PM

Jeff,
My database has almost 2 years of data from most, if not all, tracks. When I run the xToss negative expectation UDM, I get all zero's when I query this UDM against my database. If I run the xPacefailure UDM through my database, I get proper data. Is there something I'm missing regarding the Tossin / Tossout filter?

Thanks,
Patrick

jeff
1/23/2009
11:26:54 AM

Without seeing what's stored as the UDM Definition in both Profile Tables it's hard to say.

Try this... Use the UDM Wizard to create your own Tossout UDM:

1. In the Main Module or Data Window click UDM Wizard.

2. In the UDM Wizard, give the UDM a unique name... something like TossThese and set the mark characters to ^^ neg.

3. Set surface and distance to * and * so that the UDM flags horses at all surfaces and distances. Then click NEXT. At this point the base template for the UDM is now created.

4. From the factors drop down select Filters and bring up the Preset Filter Tool.

5. In the Preset Filter Tool, select TOSSOUT require TOSSOUT Profile and click ADD FILTER followed by SAVE and CLOSE.

At this point you have a negative expectation UDM that finds horses fitting Jack B's TOSSOUT profile.

Run the new UDM through the Data Window and you should see results like these:


     Data Window Settings:
     999 Divisor  Odds Cap: None
     Filters Applied: TOSSOUT
     `
     Surface: (ALL*)  Distance: (All*) 
     From Index File: C:\2008\Q4_2008\pl_PTC.txt

     `
     Data Summary         Win     Place      Show
     Mutuel Totals    1426.00   1400.80   1321.60
     Bet             -2358.00  -2358.00  -2358.00
     Gain             -932.00   -957.20  -1036.40
     `
     Wins                  23        63       122
     Plays               1179      1179      1179
     PCT                .0195     .0534     .1035
     `
     ROI               0.6047    0.5941    0.5605
     Avg Mut            62.00     22.23     10.83

-jp

.

pabennett
1/23/2009
1:13:42 PM

Thanks for the reply Jeff. I followed your instructions and created my own Tossout UDM but I still get all zero's when I query this against any of the databases that I have created. The only difference I am seeing between your output and mine (besides the zero's) is that where it says "Filters Applied:", yours reads TOSSOUT and mine reads TOSSOUT-. I don't know if this could be the issue but everything else is identical. Thanks again for your help.

Patrick

~Edited by: pabennett on: 1/23/2009 at: 1:13:42 PM~

jeff
1/23/2009
5:22:26 PM

Patrick,

Email me a copy of your c:\2004\JCapper.mdb file and I'll have a look.

jeff @ jcapper . com

(remove the blank spaces first)

-jp

.

pabennett
1/23/2009
8:16:32 PM

Jeff,

You have mail.

Thanks for taking a look.

Patrick

jeff
1/24/2009
10:58:25 AM

Ever watch a Simpsons episode and crack up when Homer Simpson does something really stupid and then says "DOH!"

One of the factor constraints in the TOSSOUT definition is that rank for JRating must be >= 5

If you aren't JRating enabled then your copy of the program isn't able to make the JRating rank evaluation for Jack B's TOSSOUT - and the result is zero plays found.

DOH!

I (almost) can't believe nobody complained about this before - or if they did - that it didn't register with me.

Looks like I'll need to create a TOSSOUT2 filter where rank for JRating >= 5 is replaced with a similar cutoff using JPR.

Until that happens, you could create a pretty strong negative expectation UDM - something along the lines of:

JPR min rank = 5
Late Pace (best 3) min rank = 4
Prime Power min rank = 6
Bris Speed Figure (best 10) min rank = 5
Lifetime Starts min val = 1
Preset Filter: FORX avoid Foreign Horse

And I have a sneaking suspicion that'd be really close to the the TOSSOUT2 filter I have in mind. < G >

-jp

.

jeff
3/24/2009
1:17:03 PM

Was playing around this morning with negative factor combinations. Came up with something fairly simple and straightforward. It uses factors from JCapper 2007 and later and finds a lot more weak horses than the original Tossout preset filter.


     UDM Definition:   MiToss
     Divisor:          # UDM Def  Divisor: 999
     Surface Req:      *ANY Surface*
     Distance Req:     *ANY Distance*
     ~
     Bris Fig (best10) MinRank= 5  MaxRank= 999   
                       MinVal= -999  MaxVal= 999   
                       MinGap= -999  MaxGap= 999
     JPR:              MinRank= 5  MaxRank= 999   
                       MinVal= -999  MaxVal= 54.99   
                       MinGap= -999  MaxGap= 999
     Late Pace (last3) MinRank= 4  MaxRank= 999   
                       MinVal= -999  MaxVal= 999   
                       MinGap= -999  MaxGap= 999
     Lifetime Starts:  MinVal= 1  MaxVal= 999
     PRating:          MinRank= 4  MaxRank= 999   
                       MinVal= -999  MaxVal= 999   
                       MinGap= -999  MaxGap= 999
     Running Style:    ALL
     ~
     ~
     query start:         3/24/2009 10:00:14 AM
     query end:           3/24/2009 10:01:34 AM
     elapsed time:        80 seconds
     ~
     Data Window Settings:
     Divisor = 999      Odds Cap: None
     Filters Applied: FORX-
     ~
     Surface: (ALL*)  Distance: (All*) 
     From Index File: C:\2009\Q1_2009\pL_profile.txt

     ~
     Data Summary         Win     Place      Show
     Mutuel Totals   10342.50  11508.40  11605.20
     Bet            -21078.00 -21078.00 -21078.00
     Gain           -10735.50  -9569.60  -9472.80
     ~
     Wins                 329       853      1533
     Plays              10539     10539     10539
     PCT                .0312     .0809     .1455
     ~
     ROI               0.4907    0.5460    0.5506
     Avg Mut            31.44     13.49      7.57

-jp

.

JCapper Message Board

JCapper 101 -- Tossout preset filter

JCapper 101
-- Tossout preset filter