I am searching for the approximate value std dev for the NBA totals. Preferably, not the data for all the games, but instead for a subset of say closing line = 200-205 or any similar subset.
Hmm, as no one cared to answer, I had to dust off my stats memories from school.
From SBR tools I got that 1.5 point in totals (200-209.5) in average transform a +100 bet to a -114.7, or 53.42%. (I used 1.5 points as it is the maximum data in the screen; it is slightly more precise than 0.5 or 1 point). In the standard normal table I get that it is equivalent to 0.086 std devs. So 1 point is 0.086/1.5 = 0.0573 std devs. Thusly, the std dev is 1/0.0573 = 17.4
My last stats class was over 20 years ago, I hope I did not make any error. Could anyone check it?
Average total for each team and the standard deviation of the total is given in the second and third columns. At the bottom, the overall is shown as 18.3 for all teams so your 17.4 wasn't far off. The last four columns show average PF, PF stdev, average PA, and PA stdev for the NBA this season
Thanks very much. My math was right, I feel very intelligent (OK, it was an intro to stats problem easy for college freshmen).
I assumed that the big scoring teams would have a larger std dev, but Phoenix has one of the lowest. Oklahoma and Indiana are in the 20+, only below Golden State. Strange.