Sep 7, 2017
Sep 7, 2017

NFL regular season betting: What should you consider?

What do winning margins tell us?

Using a Pythagorean expectation in NFL betting

How important is turnover differential?

NFL regular season betting: What should you consider?

The NFL season hasn’t even started yet and already all the talk is about who will win the Super Bowl in 2018. There’s plenty of games to be decided before then and there’s already data available that can be used to inform your regular season betting. Read on for some expert insight ahead of the new NFL season.

The 2017 NFL regular season begins on September 7 with the defending Super Bowl champions, the New England Patriots hosting the Kansas City Chiefs and culminating with Super Bowl LII on February 4 in Minneapolis at the U.S. Bank Stadium. This article analyses several factors that bettors should consider when looking to bet on the NFL. 

How does the NFL work?

The season consists of 16 regular season games spread over 17 weeks to allow for a bye week, with the 32 teams split equally between the NFC and AFC conferences. Each conference is further divided into four, four team regional divisions; the north, south, east and west.

Year after year, sides with win/loss records at the extremes regress towards 0.5 win percentages in the following season.

Six teams from each conference compete in a seeded knockout postseason format, comprising the winners of each division and the two teams with the best regular season record who didn’t top their division. 

The winning team from each conference post season then contests the Super Bowl in February. It is worth noting that one new team name appears in 2017, as the Chargers relocate from San Diego to Los Angeles.

The competition provides an attractive betting proposition, where fancied teams are excused an occasional regular season slip up, but that is not a luxury that carries over into the post season. Of course, you need to know the basics of NFL betting before you can find value in the odds.

Matches are relatively high scoring, with double figure scoring events being common place, but other significant, game changing events, such as fumbles and other turnovers, are often heavily influenced by randomness.

The New England Patriots apart, the structure of the game and the draft, tries to ensure that few sides can remain at the top for long, while those poorly performing sides can benefit and rebuild by the allocation of higher draft selections.

Although there is an element of skill and defensive scheming involved in a turnover of play there is also a degree of randomness that should be taken into consideration.

This mix of waxing and waning teams, combined with randomness in a side’s regular season win/loss record tends to draw the record of very good or very bad teams from one season towards a league average 0.5 win % of an 8-8 record.

Year after year, sides with win/loss records at the extremes regress towards 0.5 win percentages in the following season, as randomness that partly influences the previous season returns to more usual levels.

Fortunately, such factors that temporarily inflate or depress the records are readily identifiable and we can at least understand why less is expected from last season’s dominant teams.

What do winning margins tell us?

A string of narrow wins is often seen as a positive, sustainable team trait, but in reality it is simply the inevitable consequence in a sport where many games are decided by seven points or fewer and some teams will naturally and often randomly accumulate more narrow wins than losses in a short 16 game season.

Firstly, let’s see how this metric performed when predicting likely performance during last year’s regular season.

In 2015, Denver, Arizona, Carolina and Indianapolis had the four largest differentials between narrow wins and narrow defeats. Their combined regular season win percentage was 0.75 and none of the four teams had a losing season.

In 2016, their combined regular season win percentage fell to 0.48 and two of the successful four from the previous season posted losing seasons. 

At the opposite end of the scale, Cleveland, Dallas, Tennessee, New York Giants and San Diego (now in LA) all suffered more narrow defeats than narrow victories when winning just 25% of their regular season games in 2015.

Pythagorean expectation uses the ability to score and concede points as a measure of team talent, rather than how those points are shared around within a limited number of regular season games.

Although Cleveland dropped two more wins in 2016 compared to the 2015 season, all the others improved their record and had a combined win% of 0.48 and three of the five teams had winning seasons.

Looking ahead to the new season, there were a host of teams who lost significantly more narrow games than they won during 2016 and might be expected to improve their records with a more usual split in such games in the upcoming season.

The newly named LA Chargers had a -7 game differential in closely contested games, although they also have a change of venue to contend with as they attempt to improve on their 5 win 2016 campaign.

The list of likely improvers then includes Jacksonville, Chicago and Philadelphia, with Carolina, Cincinnati, Buffalo, Cleveland and San Francisco also recording a relatively large, negative differential.

The 2016 sides whose record was bolstered by an imbalance of narrow wins that they may not be able to repeat in 2017 include Oakland, Miami, Dallas, Houston and New York Giants.

Using a Pythagorean expectation in NFL betting

A related metric that might herald to movement towards the middle ground is a Pythagorean expectation that over or under performs a side’s actual win loss percentage.

Pythagorean expectation uses the ability to score and concede points as a measure of team talent, rather than how those points are shared around within a limited number of regular season games to produce wins or losses.

The structure of the NFL tries to ensure that few sides can remain at the top for long, while those poorly performing sides can benefit and rebuild by the allocation of higher draft selections.

A team whose actual win record exceeds their Pythagorean expectation in one season tends to perform more in line with their Pythagorean in the upcoming season - this can then be used as part of an NFL spread betting strategy.

Some sides were duplicated in this list and the previous one, but the four sides who most greatly over performed their Pythagorean in 2015 averaged a win% of 0.72, falling to 0.48 in 2016 and a similar bounce towards 0.5 was seen for the five sides whose win loss record most underperformed compared a points based approach.

This season LA Chargers, Cleveland, Jacksonville and San Francisco are the familiar names where improvement might be expected and Oakland, Houston and the Giants are joined by Tampa Bay and New England as teams who may struggle to emulate their win totals from last season based on their Pythagorean from 2016.

How important is turnover differential?

The final category that greatly influences win/loss records, but rarely in a sustainable way is turnover differential or how often a side takes away possession from their opponent, either by interceptions or fumble recovery compared to how frequently they cede possession themselves.

Oakland, Houston and the Giants are joined by Tampa Bay and New England as teams who may struggle to emulate their win totals from last season.

Although there is an element of skill and defensive scheming involved in a turnover of play, particularly when taking the ball away by interceptions, there is also a degree of randomness that should be taken into consideration.

Dallas, Baltimore, Jacksonville and Tennessee had the worst turnover differential in 2015, but in 2016 they turned a combined win percentage of 0.27 into one of 0.51 and of the four previously losing sides, Baltimore returned an 8-8 season, Tennessee moved to 9-7 and Dallas 13-3.

In contrast, the four teams with the best turnover differential fell from a combined 0.8 to 0.5.

Oakland, Kansas, Minnesota, Atlanta and New England had the best turnover differentials in 2016. Teams with the worst figures, who may see natural regression this season and an improved return, included Chicago, Jacksonville, New York Jets, Cleveland and LA’s other franchise, the Rams.

Want to learn more that will inform your NFL regular season betting? Find out how to use yards per play rankings to your advantage.

Betting Resources - Empowering your betting

Pinnacle’s Betting Resources is one of the most comprehensive collections of expert betting advice anywhere online. Catering to all experience levels our aim is simply to empower bettors to become more knowledgeable.