• Tom
    89
    I want to make an odds line based on lower-is-better numbers,

    3
    7
    11
    15
    18
    Any suggestions how to turn these into probabilities?
  • Dave Schwartz
    303
    Can you offer us a scale for the range of numbers?
    From what to what?
  • RanchWest
    503
    Yes, I would invert to the range.
  • RanchWest
    503
    If you do not want it weighted, also convert to percentages.
  • Dave Schwartz
    303
    If you do not want it weighted, also convert to percentages.RanchWest

    Yes, like 100 - score.

    Or 20, or whatever.

    That's why I asked for the scale.
  • Steven
    113
    You could also rank the values in ascending order and then

    IF RANK = 1 then OddsValue = 100*
    ELSEIF RANK = 2 then OddsValue = 90
    ELSEIF RANK = 3 then OddsValue = 80
    ELSEIF RANK = 4 then OddsValue = 70
    .
    .
    ELSEIF RANK >= 12 then OddsValue = 1
    END IF

    * These values can be any value you want. A Fibonacci series would probably work better.
  • Tom
    89
    Say lowest would be 7 and highest would be 35.

    I think you guys may have already given me the answer/
  • Dave Schwartz
    303
    Say lowest would be 7 and highest would be 35.

    I think you guys may have already given me the answer/
    Tom

    What would that answer look like?
  • Tom
    89
    Rank 1 = 5
    Rank 2 = 4
    etc. to
    Rank 5 =1
  • RanchWest
    503
    Another approach is to assign Fibonacci numbers to the rankings.
  • Dave Schwartz
    303
    Another approach is to assign Fibonacci numbers to the rankings.RanchWest
    My personal favorite.
    ___________________
    I want to make an odds line based on lower-is-better numbers,
    3
    7
    11
    15
    18
    Any suggestions how to turn these into probabilities?
    Tom

    Are these similar to Sheet Numbers, by chance?
    If so, I have a schema for that.
    ___________________
    Rank 1 = 5
    Rank 2 = 4
    etc. to
    Rank 5 =1
    Tom

    I'd call this one, FUZZY handicapping.

    I'd weight it as follows:
    5 = 100 (Excellent)
    4 = 62 (Good)
    3 = 38 (Average)
    2 = 24 (Fair)
    1 = 14 (Poor)

    I'd also add a 7 & 9.)
    7 = 162 (Think Off-the-scale horse.)
    9 = 262 (Secretariat entered in a $20k claimer.)

    Then I'd normalize the whole thing to 100%.
  • RanchWest
    503
    I take my totals and raise them to a power of 1.2 to 1.6 because the low odds horses will usually be overbet and this takes that into account.

    But, in the end, I tend to generally not put a lot of credence in the specific odds. I just want to know which horses should be most likely to make my contenders list. Not all low odds horses are contenders, though.
  • Tom
    89
    The numbers ranks, ie like Sartin outlined in the Yellow Manual, To simplify:
    EP SP AP
    1 3 2 = 6
    2 4 3 = 9
    3 1 4 = 8
    4 5 5 = 14
    I'm playing around with Randy Giles' Optimum Pace Model and want to incorporate either class or Bris PP into a final number.
  • Dave Schwartz
    303
    EP SP AP
    1 3 2 = 6
    2 4 3 = 9
    3 1 4 = 8
    4 5 5 = 14
    Tom

    Maybe add Reynolds numbers?

    Best 2 multiplied together
    Best 3 multiplied together

    Thus
    1 3 2 = 2 6
    2 4 3 = 6 24
    3 1 4 = 3 12
    4 5 5 = 20 100
  • RanchWest
    503

    I use the ranks of the 7 primary Sartin factors.
  • Dave Schwartz
    303
    I use the ranks of the 7 primary Sartin factors.RanchWest

    What are those these days, please?
  • RanchWest
    503


    I am not an expert on Sartin, but here's what I use (by rank of the pace line):

    Early Pace
    Late Pace
    Total Energy
    Turn Time
    Hidden fraction
    FW
    FX

    To correct myself, though, I use this as a significant number, but I don't think I include this in my odds line, though I certainly couldn't disagree with anyone who does.
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.