In 2005, the average NFL team score was 20.8 points per game, implying a typical game total of 41.6. By last year, that figure crept up to 45.6 points per game, on average.

In response to the more-or-less consistent climb in NFL scoring, we would expect bookmakers to increase the average betting totals. One measure of oddsmakers' reactiveness is the change, over time, in number of games with closing totals north of 50. The following chart shows the uptick in such occurrences over the years, which is, in fact, highly correlated with rising scores over time.

Despite a greater propensity for bookies to jack totals up these days, so far this season the O/U is a staggering 30-19. The Action Network cites increased pace and efficiency, favorable officiating and the greater number of fourth down attempts as justification for the ridiculous cover rate for the OVER in 2020, noting that scoring is up more than five points versus last year. There have already been seven games this season with a closing total of at least 50.

We contend that bettors need not universally fear the OVER in high total situations--generally speaking, and especially this year. It is true, on the one hand, that the familiarity bred by more frequent meetings between teams in the same division seems to depress scoring versus expectations in games with high totals. Since 2003 the O/U is 59-80-3 (44%) in regular season divisional matchups with totals of 50 or better. However, in non-divisional play, where the teams are less aware of each others capacities, the O/U is a more neutral 138-132-3 (51%).

Thus, a high number, per se, has not historically indicated over-optimism in inter-divisional action.

The O/U in high total games this year is 4-3. When this observation is considered in tandem with the broader advantage for the OVER this season, there seems to be a case to be made that bookies may not have fully adjusted to the increased productivity on display this year.

We present below a simple system designed to identify high-probability OVER opportunities.

Backing the OVER in non-division meetings between two teams each allowing at least 27 points per game, on average, has been good for a 19% return on investment since 2003.

This system is 2-1 so far this year, before the four games that are triggered for week 4.

At present, the totals for the featured contests range between 53.5 and 56.5--a zone that might have been regarded as the nosebleed section in years past. However, there is a case to be made for biting the bullet and pulling the trigger on the OVER in each of these instances. For one, in non-division action, betting high-total OVERs is not gambling sacrilege, as is taking the OVER in divisional play. Further, the OVER is on a terrific roll this season(!), aided by environmental factors that have yet to be fully baked into lines. And finally, when bad defenses collide in non-divisional matchups, the OVER has been the best bet.

Happy betting!!