The NFL season unofficially kicks off for bettors when season win totals are released. For many professional bettors, even a league as hyper-popular and overanalyzed as the NFL can provide edges.
Finding an edge, though, requires an awareness of key statistical indicators and an ability to disregard — or at the very least, minimize — biases and overgeneralizations developed from the past season and over the course of the offseason. With a 16-game NFL regular season representing such a small sample size, a few bounces of the ball or missed calls can be the difference between a 10-6 record and 8-8. The key, therefore, is to develop an understanding of what a team’s true performance was in a given season — and the factors that may have affected it — and then to use that as the basis for assessing potential improvement or decline.
There are a multitude of factors that can provide a clear picture of where a team truly stacked up the year prior. Among them:
Record in close games
Performance in close games is, for nearly every team not named the New England Patriots, generally luck. Teams that overperform in close games tend to regress toward the mean the following season in those games, and that usually leads to a drop in wins. Teams that underperform in close games tend to regress (positively) toward the mean the following season, which typically leads to an increase in wins.
Turnovers are another statistical representation of luck and, similar to close games, the same principles apply. Teams that have a poor turnover ratio tend to see that ratio improve the following year and, as a result, tend to win more games. The reverse is true for teams that benefited from a surplus of takeaways the year prior.
Wins and losses are not a true reflection of a team’s performance. The most genuine reflection is scoring — both points scored on offense and points allowed on defense. Bill James first introduced the method for quantifying a team’s true win rate based on points scored and allowed, an analytical formula called “Pythagorean expectation.” When a team’s actual wins supersede its Pythagorean expectation, that team represents a prime candidate for underperformance the subsequent year. Conversely, teams that have fallen below their Pythagorean expectation the prior season are traditionally undervalued by the market.
The Plexiglass Principle
Teams that experience precipitous rises or falls in their win total from one year to the next tend to regress toward the mean in the third year. The Dallas Cowboys have been a good recent example. They went 13-3 in 2017, an increase of nine wins from their 4-12 record in 2016. They then dropped four wins to 9-7 in 2018. Bettors who can put in the time to research and identify similar opportunities will be able to isolate value more often than not.
Injuries naturally impact a team’s performance. A high overall injury rate or cluster injuries at one position can be crippling. Look at how bad a team’s injury rate was and compare it to performance. If a team suffered multiple key losses on offense but still posted solid scoring-efficiency metrics, they have the scope to improve with a healthy season. If a team had good health on offense but still struggled, their offensive issues may be more systemic.
A team’s schedule is critical to analyzing its risk profile. Teams with soft schedules can pile up wins even when their form dips. Those facing a difficult schedule have significantly greater downside. Understanding a team’s schedule is essential to properly handicapping season win totals.