1. #1
    outdrawed
    outdrawed's Avatar Become A Pro!
    Join Date: 11-21-07
    Posts: 388

    Teaser Question: Math Class

    I'm very new to sports betting so I am not skilled in the ways of calculating this stuff. I'm crossposting this from another forum I post at. Any help is appreciated.

    I'm looking at a 6 pt 2-team teaser right now. One of the teams, however would be the Eagles +3 (-130).

    Now, obviously I know that teasing across the zero is bad bad bad. My question is, since -3.0 (-130) is effectively -3.5 (-104), I'm getting close to 6.5 points for this teaser as I get the same 10/11 odds regardless of the vig of the games I'm teasing.

    How can I go about calculating whether or not this is worth it?

    Actually, I have another question.

    A 6.5 pt two-team teaser pays out 5/6. That's worth 10 cents more than a 6 pt teaser. If I do the 6.5 pt teaser, I get paid a little less, but I cross the 3. It doesn't help me nearly as much in my other game. (Bears +2/+8.5), but I'm wondering if it'd be better to do a 6.5 pt teaser instead.

  2. #2
    Ganchrow
    Nolite te bastardes carborundorum.
    Ganchrow's Avatar Become A Pro!
    Join Date: 08-28-05
    Posts: 5,011
    Betpoints: 1088

    Quote Originally Posted by outdrawed View Post
    I'm very new to sports betting so I am not skilled in the ways of calculating this stuff. I'm crossposting this from another forum I post at. Any help is appreciated.

    I'm looking at a 6 pt 2-team teaser right now. One of the teams, however would be the Eagles +3 (-130).

    Now, obviously I know that teasing across the zero is bad bad bad. My question is, since -3.0 (-130) is effectively -3.5 (-104), I'm getting close to 6.5 points for this teaser as I get the same 10/11 odds regardless of the vig of the games I'm teasing.

    How can I go about calculating whether or not this is worth it?
    There's of course no way to calculate this from first principles, rather you'd need to look at the historical record and/or an NFL score distribution model.

    Looking solely at the historical record (1985-2006, covers.com data) we see that out of 432 trials the 3.5 point favorites covered a 6.5-point teased spread 272 times and did not cover 130 times.

    So conditioned on not pushing, the win frequency of 3.5-point favorites on a 6.5-point teased spread is 67.66% with a standard error of 2.34%.

    Given that breakeven at -110 is 72.37%, we're a slightly more than 2σ's away from a profitable investment.

    Quote Originally Posted by outdrawed View Post
    A 6.5 pt two-team teaser pays out 5/6. That's worth 10 cents more than a 6 pt teaser. If I do the 6.5 pt teaser, I get paid a little less, but I cross the 3. It doesn't help me nearly as much in my other game. (Bears +2/+8.5), but I'm wondering if it'd be better to do a 6.5 pt teaser instead.
    Using the same data set the win rate of a 7-point teaser off the -3.5 is 302/402 ≈ 69.91% with a std. error of 2.21%. Breakeven at -120 is 73.85% meaning we're about 1.78 std. errs. from profitability.

  3. #3
    outdrawed
    outdrawed's Avatar Become A Pro!
    Join Date: 11-21-07
    Posts: 388

    I was really hoping you'd reply to my thread, Ganchrow. Thanks for the info!

  4. #4
    outdrawed
    outdrawed's Avatar Become A Pro!
    Join Date: 11-21-07
    Posts: 388

    I went to covers.com but couldn't find a link to that data you're talking about. Could you tell me how to find that? Do I need to register first?

  5. #5
    Ganchrow
    Nolite te bastardes carborundorum.
    Ganchrow's Avatar Become A Pro!
    Join Date: 08-28-05
    Posts: 5,011
    Betpoints: 1088

    Quote Originally Posted by outdrawed View Post
    I went to covers.com but couldn't find a link to that data you're talking about. Could you tell me how to find that?
    From the covers.com covers.com/pageLoader/pageLoader.aspx?page=/data/nfl/teams/teams.html, select a team, then click the "Past Results" tab. You can then select the season for which you want data from the dropdown at the top left of the page.

    If you want to pull all the data your best bet is to use an HTML scraper.

Top