Sigma significance?
December 7th, 2010 at 7:19:53 PM
permalink
Wizard or anyone who can help,
A while back I asked the Wizard how one could judge his NFL betting acumen. It was settled by deciding that ROI should be used since spreads and money lines associated with these spreads would all be taken into account.
Recently, I have received a challenge that I cannot get over 70% winners. The rules are as follows:
1) The lines have to be ascertainable, offered at minimum 3 casinos per vegasinsider
2) They are all spread bets, with 2 team teasers allowed up to 7 points (-130 bet)
3) There is no time limit. If it takes years to get these "special 70% plays" that is quite alright.
Note that the spread bets also could be up to -130 given the line changes.
Given that the challenge has to do with % winners, what statistical method should be employed in order to gain a confidence level that 70% winners are achieved, not due to randomness or chance?
One person that I have asked has said that 3 sigma confidence levels indicate that a 37-19 spread record would do it.
Is this correct? What other methods might one use?
Thanks,
KidTwist
A while back I asked the Wizard how one could judge his NFL betting acumen. It was settled by deciding that ROI should be used since spreads and money lines associated with these spreads would all be taken into account.
Recently, I have received a challenge that I cannot get over 70% winners. The rules are as follows:
1) The lines have to be ascertainable, offered at minimum 3 casinos per vegasinsider
2) They are all spread bets, with 2 team teasers allowed up to 7 points (-130 bet)
3) There is no time limit. If it takes years to get these "special 70% plays" that is quite alright.
Note that the spread bets also could be up to -130 given the line changes.
Given that the challenge has to do with % winners, what statistical method should be employed in order to gain a confidence level that 70% winners are achieved, not due to randomness or chance?
One person that I have asked has said that 3 sigma confidence levels indicate that a 37-19 spread record would do it.
Is this correct? What other methods might one use?
Thanks,
KidTwist
