In the book, Win That Pool!, I describe a way to estimate win probabilities using point spreads and by using the standard deviation of the against-the-spread margin (ATSM).
For visitors to this site who have not read the book the ATSM tells you how the eventual point margin compared to the point spread. In other words, how good of a predictor was the spread?
For example, in the WildCard weekend for the 2018 season, the Bears were favored against the Eagles by 6 points (-6), but lost by 1 point. So the ATSM for the favorite on this game was -6 minus 1, or -7.
If you calculate this measure for all the games for several seasons, you will find that the average is close to zero and the standard deviation is somewhere between 13 and 14, but closer to 13.
This result is important, because it means the spread is an exellent predictor of the eventual point margin. That means you can use Excel or some other calculator to translate a point spread into a Win Probability. I tend to assume the standard deviation of ATSM is 13.3.
For example, in today’s conference championship games, as I write both home teams are favored by -3. That implies, with a normal distribution of (mean=0, sd=13.3), the win probability for both favorites is approximately 59%.
For the time being, since I published the book so recently, you can just assume the ATSM standard deviation is 13.3. I wrote that you can calculate your own estimate of the standard deviation of ATSM because it’s possible that number will change in the future, especially if the NFL rules around safety change.
The question is, where can readers obtain this information?
As I write in the book, I rely on a site called sportsdatabase.com. If you are new to that site, you may have to spend some time learning how to query their database, but it is pretty powerful.