[pokerjoe]

Let's say we have a league and a sport with an average non-OT game total of 50.

Two teams meet, each with an average expected score of 40.

You have no additional information.

What do you make the total?

I've always thought it was the the team's individual score expectancy multiplied, then divided by the league average score. So, in this case, 40*40 = 1600, divided by the league average of 50, = 32.

OT impact would be added by the likelihood of OT occurring (less likely the higher the spread) added on as a point value, so if the chance of OT is 2%, and 2% was equal to 1/2 point at that total point, then 32.5

(I didn't name the sport to avoid being sidetracked into actual values and chances)

But, I haven't actually done any work on this (in all these years, hard to imagine).

So, any one whose put more thought into this, want to weigh in? Because in some early database fooling around, I don't find this formula particularly accurate. A little better than just adding the individual team scoring and dividing by 2, but not that much more accurate.

Also, in lieu of dividing the product of the team's scoring by the league average, maybe dividing it by the average opponent's score would be more accurate, especially if you removed the team's own impact on that average opponent's score.