Originally posted on 10/26/2020:

Quote Originally Posted by KVB View Post
It is not likely wiser to arbitrarily take what you think are key numbers from a a different sport to apply to the NFL but you could always back test.

Another suggestion would be in fact to make a more meaningful adjustment to the key number and use a recency effect with a larger sample.

Going with the "classic" numbers may throw you off a little, but perhaps not much.

Try looking at the last 20 years or so, then see what you consider the top 5 in order. It will be different that what you have. You can weight the recent higher scoring as well but be careful, the market because the market may not be what you think it is.

Why adjust at all right now?

The lines are within reason of the distribution of NFL scores and the Total count based on the closing line is basically a 50% split with 51 games going UNDER and 51 games going over with a couple of pushes.

The public favors OVERs and favorites and from your post it sounds like you do the same. Scores might be higher on average, but what about medians? Keep an eye on the variance and standard deviations of the lines and results for the league as a whole and the teams specifically. It will reveal quite a bit. As with every year, the market has overestimated scores for a slew of teams.

Taking numbers from a large sample of NCAAF games and laying them over the NFL might be a decent hypothesis to start with, but I think there are better ways figure NFL key numbers and to use them effectively.

Keep in mind theres been a few major changes in NFL land, one people miss extra points because its kicked from further out. Second thing, it may be that coaches are now going for 2 more often, and the starting drive goes from 25 instead of 20 for touch backs.