1. #1
    sunnyblue88
    sunnyblue88's Avatar Become A Pro!
    Join Date: 08-11-06
    Posts: 116
    Betpoints: 2648

    live betting fair value odds

    Say the line for one MLB game closed @ 0 9(+101) under 9(-111) in pinnacle. If No score after one inning, what's the fair value total odds should be ? How about for 1 run scored, two run... etc. Is there a place that I can figure out this? If pinnacle offers live betting, I will be able to tell by looking at their odds, but some games, pinnacle does not offer live betting. Thanks.

  2. #2
    mathdotcom
    mathdotcom's Avatar Become A Pro!
    Join Date: 03-24-08
    Posts: 11,689
    Betpoints: 1943

    To get started you could go to covers.com, look at all previous games that had a total of 9, and see how many went over after 0 runs scored in the inning, 1 run scored in the inning, etc.

    If 55.5% of those games go under when there is no run scored in the first, then fair odds should be about -125 on the Under.

  3. #3
    buby74
    buby74's Avatar Become A Pro!
    Join Date: 06-08-10
    Posts: 92
    Betpoints: 21207

    I have amodle which takes the moneyline and total lines and predicts the freall possible scores

  4. #4
    buby74
    buby74's Avatar Become A Pro!
    Join Date: 06-08-10
    Posts: 92
    Betpoints: 21207

    I'll try that again accidentally posted my response.

    Anyway my model predicts that if the first inning is scoreless in a game with an initial total line of 9 over +101 under -111 the fair line for the over under should drop to 7.5 over-108 under +108 or 8 over +108 under -108.

    If runs were scored in the first just add them to the new total. The relative strengths of the two teams has very little effect e.g. if the money line predicts a close game or a heavy favourite it doesnt make much if a dfference with heavy favourite games being slightly higher scoring.

    I would be very interested to see how you database search compares with this result.

    Caveats
    -I havent validated my model on a large dataset as it is a top down model using sabermetrics to work out the distribution of runs scored in an inning.
    - my model assumes that all innings are equally likely to score which is ok if you only consider the whole game but actually the 1st inning is the highest scoring in the mlb particularly in the NL with 15% more runs scored in the first inning than the average inning and 15% less scored in the 2nd due to the terrible batting performance of pitchers. The AL sees a similar but weaker effect.

    Hope this helps

  5. #5
    Arilou
    Arilou's Avatar Become A Pro!
    Join Date: 07-16-06
    Posts: 475
    Betpoints: 949

    A Monty Carlo simulation of some kind is going to be your best bet.

    buby74:

    1. If the score is tied the runs can be added to the total, but if you don't take into account the impact those runs have on when the game is going to end then you are going to get yourself into serious trouble.

    2. As you noted innings are not created equal, and to do so will also get you into trouble especially later on in the game.

  6. #6
    buby74
    buby74's Avatar Become A Pro!
    Join Date: 06-08-10
    Posts: 92
    Betpoints: 21207

    Excellent points. My
    Model is a simple Markova so mcmc is not needed I will run it with either the home or road team scoring 2 in the first and see what happens

Top