• Steven
    113
    Working with Dave and Tom I was able to put Tom's Ultimate Software concept into a little program for the HorseStreet users. It is patterened after his PDF as closely as I could get it.

    And Dave was kind enough to provide a link to the package > the program and some short PDF directions.

    We can use this thread for comments, questions, criticisms and other stuff.

    Tom's Ultimate Odds Line Software and Directions
  • Dave Schwartz
    262
    I actually did nothing but sweep the file for viruses and host it on our site.

    Steven did all the work.
  • Jim Pommier
    49
    Very good work. A few comments. 1). There are no ratings in the ACL3 column. 2). There appears to be a calculation issue. SAX R4, #10 Et Toi has the highest Prime of 142 or 10 points. 10 points would be 10-1. The entry also appears to be missing 1 point in SRLR and EPLR, and 2 points in LPLR. Total 14 or 7-1. Thanks, Jim.
  • Steven
    113
    And that's why we have this thread.

    The ALC3 column uses the last three class ratings. I'm not sure all the BRIS files contain that information. If your data file does not contain these ratings, then it will not compute the ALC3.

    Let me look at the other items.
  • Steven
    113
    Jim, SAX1215?
  • Steven
    113
    And which data file are you using? .drf, .eg, .mcp, .pcs, .jcp????
  • Jim Pommier
    49
    SAX0124.DRF R4, the race example in the instructions.
  • Steven
    113
    You would think that I would have caught that?!

    That DRF file does not contain class ratings.
  • Jim Pommier
    49
    I'll default to Tom and the rest of the group, but maybe add a different rating? Something related to class, or a high impact value rating that's in the DRF file. Also, easy to code.
  • Steven
    113
    Well, it's Tom's concept.
  • Tony Kofalt
    393
    Thank you Tom and Steven!! great work and thank you for sharing.
  • Steven
    113
    It appears that the $1 DRF files do not include the BRIS CLASS rating per race but the $5 multicaps files do. $5 Yikes.

    I'm open to changing the class requirements to something included in the $1 file if someone can come up with a good idea.

    And I did get that PRIME POWER bug fixed. Fix is in V1.7 zipped up below.
    Attachment
    UOL1.7 (137K)
  • Steven
    113
    And Ultimate is spelled Ultimate, not ulitmate. Spelllng was never my strung soot.
  • Tony Kofalt
    393
    Thanks Steven for your great work. Since this program is about 'odds lines' I'd suggest some sort of a trainer stat in place of the blank class category. Trainers do impact the odds line and who wins the race. I'll wait for some feedback from Tom, Ranch and William as well. They always present good ideas
  • RanchWest
    503
    I think it would be good to model for the track being handicapped. Or, a simple approach would be to use the provided morning line. Tony has a good point about trainers impacting odds. Jockeys impact odds a lot, too. One big factor in jockeys winning is that their agent gets them good mounts. But jockeys tend to run hot and cold in the short term.

    Just depends on whether you're trying to evaluate a realistic probability of the horse winning or you're trying to forecast the public wagering. I prefer the former.
  • Dave Schwartz
    262
    FWIW, in our software we use these factors to predict the odds.

    100  Morning Line
     62  PSR (like BRIS Prime Power)
     38  Cramer Power  (like BRIS Prime Power)
     24  RTG (like BRIS Prime Power)
     15  Trainer wins at the track last 365 days
      9  Composite Final Time (hi-level speed rating)
    
  • Steven
    113


    I'm pretty sure I've seen most of those in the "Practical Handicapping" workshop. When Dave says "High level", he means "complicated formula, but not outside the scope of a spreadsheet handicapper".
  • Steven
    113


    Trainer win percentage would be an easy substitution. We just need to maybe get a consensus.
  • Steven
    113


    Modeling is far outside the scope of what I intended this project to be so nix this idea.

    And, like trainer win percentage, morning line could also be substituted.
  • Jim Pommier
    49
    I'm okay with adding the trainer's overall win % with the 10-6-4-2-1 versus any high-level coding. Also, would like the ML column next to the UOL (no coding, just for reference and comparison). There may be others like me that are not able to bet "live", and make their bets early. Thanks, Jim.
  • Steven
    113
    Jim, that's a good idea. It may be a day or two before I get back into this one but it looks like:

    Adding the ML column
    Using trainer win percentage instead of ACL3.
  • Tom
    89
    Steven, great job!
    I'm all in on anything you want to substitute. The real name of the idea is the Ultimate Lazy Guy's Oddsline. LOL.
    Thank you for the great work you put in.
  • Tony Kofalt
    393

    Tom- thank you for your work as well. After all it's named Tom's Ultimate Odds Line
  • Steven
    113
    Let's try version 1.9

    Added ML and used trainer win percentage instead of average class last 3 races.

    As always, let me know if and when you find a bug.
    Attachment
    UOL1.9 (137K)
  • Jim Pommier
    49
    1). The LPLR column now reads UOL when you open up a race. Should go back to LPLR. Click on a race and it changes. 2). Also, I'm not coming up with the 10-6-4-2-1 total for each column that equals the UOL. Seems to be off a few points. If you can pick a track and date for today or tomorrow, we can go over a few examples. I'm going to be out most of today so may not get back to you right away. Thanks, Jim.
  • Steven
    113
    That’s why I said “when you find a bug”. Let me look at what’s going on. Thanks.
  • Tony Kofalt
    393
    Lol- too funny Steven. Thank you for your work here!!! I seem to find a bug or two in my selections every day.
  • Steven
    113


    Just to clear a thing or two/three...

    For example:

    If the speed ratings go

    85
    79
    82
    85
    60
    44
    39
    32
    The rankings are line this

    85 - 1
    79 - 4
    82 - 3
    85 - 1
    60 - 5
    44 - 6
    39 - 7
    32 - 8

    There is not "2" since #1 is tied. Numerical values would then be

    85 -> 10
    79 -> 2
    82 -> 4
    85 -> 10
    60 -> 1
    44 -> 0
    39 -> 0
    32 -> 0

    No "6" rating due to there being no "2" ranking. I hope that's the issue.

    And I fixed the LPLR and UOL thing. V2.0
    Attachment
    UOL2.0 (137K)
  • Jim Pommier
    49
    This is helpful. I'll recheck some examples tomorrow when I get home. I was taking the ties (85 & 85)-- (10 + 6 / 2 = 8 each) versus equal 10 each.
  • Jim Pommier
    49
    I need to check a few things with the coding. 1). If the Score falls in between two Odds, which Odds are you defaulting to? For example, Scores 12, 14, 16, 18, 20, 22, 24, 25, 27, 28 etc. For example, if the Score is 16, is the Odds 6-1 or 7-1? Score 28, 3-1 or 7-2? Seems in Tom's examples he went both ways. I would recommend the lower, 16 = 6-1, and 28 = 7-2.
    2). How are you Scoring multiple ties? For example, 102, 87, 81, 81, 81. [10 - 6 - ? - ? - ?]. For example, 20, 19, 16, 16, 15. [10 - 6 - ? - ? - ?]. Thanks, Jim.
  • Jim Pommier
    49
    Coorection: I would recommend the lower, 16 = 6-1, and 28 = "3-1" (I had 7-2).
bold
italic
underline
strike
code
quote
ulist
image
url
mention
reveal
youtube
tweet
Add a Comment

Please register to see more

Forum Members always see the latest updates and news first. Sign up today.