Stat's guys or Excel Guru's

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • ScreaminPain
    SBR High Roller
    • 09-17-08
    • 246

    #1
    Stat's guys or Excel Guru's
    Proposal:
    I'd like to quantify winning percentages into a meaningful stat.

    Let me mention, I'm NOT able to write code and I'm adaquately serviceable with Excel, but somewhat limited when using it's more complex capabilities. For discussion's sake, lets surmise the Angels are playing the Rangers and I want to reach a winning probability using the following stats ONLY!

    Overall Seasonal Records
    Home/Away Record
    Record against >.500 teams, <.500 teams

    Obviously, this will not become a profitable betting solution, but I need to understand the process for other similar endeavors.

    Angels:
    Overall: 55-48 53%
    H/A: 27-25 52%
    >.500 24-23 51%

    Rangers:
    Overall: 59-44 57%
    H/A: 34-19 64%
    >.500 26-24 52%

    My question is this. How do I derive a winning probability of a Texas/Angels matchup using 3 sets of data from each team? What would the formula look like if written into Excel?

    I understand Bill James' work and I currently make use of his Log5 formula. Of course, that's using winning percentages from both teams. I don't know how to assimilate sets of data so that the winning percentages can be converted to correct probability.

    Any help is greatly appreciated...
  • uva3021
    SBR Wise Guy
    • 03-01-07
    • 537

    #2
    The log5 Formula should suffice for overall W% and H/A. The result is a win probability. You can assume the average opponent for each team hovered around 50%, which would make the inclusion of the >.500 stat negligible in this case
    Comment
    • ScreaminPain
      SBR High Roller
      • 09-17-08
      • 246

      #3
      Originally posted by uva3021
      The log5 Formula should suffice for overall W% and H/A. The result is a win probability. You can assume the average opponent for each team hovered around 50%, which would make the inclusion of the >.500 stat negligible in this case
      I can construct a formula using Overall stats or H/A stats, but how would one construct a formula using BOTH sets of data in a single formula?
      Comment
      SBR Contests
      Collapse
      Top-Rated US Sportsbooks
      Collapse
      Working...