Predicting NFL Season Win Totals: AFC North


*This is part seven of an eight-part series looking at each division*

Previous: NFC East l NFC South l NFC North l NFC West

AFC East l AFC South l AFC West

The AFC North is the third most challenging division. The Baltimore Ravens are the favorites from a win total perspective. The model is not as confident in the Ravens this year as the bettors seem to be. The model sees this division shaking out in an uncommon manner.

Based on the Vegas over/under win totals, the following graph ranks the relative strength of each division in the NFL. The NFC South has a projected win percentage of 53.68%.

The AFC North plays the AFC West and the NFC North this season. Those divisions project to be the second and sixth most difficult divisions, respectively. The AFC West enjoys a projected win percentage of 55.15%. The NFC North, brought down mainly by the Lions, sits at 45.59%.  

The NFL schedule has expanded to seventeen games this season. The two conferences will rotate yearly between hosting the ninth game and going on the road for a ninth game. This year, the AFC gets the benefit of the added home game. Each team in the division will face three different opponents that correlate to their position in the 2020 final divisional standings. This change in scheduling is a statistically significant datapoint in assessing win totals and division winners. 

In years past, teams within a division shared a commonality of opponents in 14 of their 16 games, 87.5%. This year, teams within a division will share a commonality of opponents in 14 of their 17 games, 82.3%. This additional 5% disparity will exert an outsized role in determining win totals, division titles, and playoff births. These three different opponents will be referred to throughout the article as non-standard opponents.

The AFC North’s non-standard opponents come from the AFC South and East and the NFC West. The particular opponents for each team are listed in the table below:

TeamOpponent OneOpponent TwoOpponent Three
Cleveland BrownsHoustonArizona@ New England
Baltimore RavensIndianapolis@ MiamiLA Rams
Pittsburgh Steelers@ BuffaloSeattleTennessee
Cincinnati BengalsJacksonville@ Jets49ers

The first thing you should notice from the table is that it is imbalanced. All AFC teams will get two home games from their pool of non-standard games bringing their total number of home games on the season to nine. There is also a massive disparity in the strength of each team’s non-standard schedule. While this was also true in the past, the additional game this season amplifies that imbalance. These games now comprise 17.6% of the schedule.   

To illustrate the disparity in the strength of schedule of non-standard opponents, the graph below indicates the combined projected win percentage of each team’s non-standard opponents. 

The information from the graph pops off the page. Rational decision-makers would much prefer Cleveland’s schedule to that of Baltimore or Pittsburgh. Remember that Baltimore and Cleveland are separated by only a half a game in their over/under win totals. 

Looking objectively at the chart, we can deduce that Cleveland’s opponents expect to win 1.26 times in the three games. Conversely, Baltimore’s opponents expect to win 1.55 times over three games. That is a .29% edge to Cleveland. (Pittsburgh’s opponents expect to win 1.76 times in three games) Since results are binary, we expect this to result in a net +1 win for the Browns – which covers the disparity in win totals between the two teams. 

Vegas bakes the schedule into the win total, and the NFL has had two games determined by this method for years. That is accurate. The model finds that the books are underestimating the significance of a third such game.  

There are three things needed to win in the NFL; talent, schedule, and scheme. The following end-of-season division projections derive from an objective mathematical model that has simulated the 2021 NFL season tens of thousands of times.

Cleveland Browns

Projected Finish: 1st, AFC North

Projected Range: 11 – 13 Wins

Las Vegas Win Total: 10.5

Model Win Projection: 11.54

This division will be determined from week thirteen through fifteen when the Browns and Ravens play twice. Week thirteen is in the DMV. The Ravens will travel to Pittsburgh in week fourteen before traveling to Cleveland for the week fifteen game. Cleveland, well, they have a bye week in between. The model believes if Cleveland can split the games, they are highly likely to win the division. It believes Baltimore will likely need a sweep to win the division.

The Browns drafted JOK primarily to spy on Jackson. JOK is a hybrid linebacker/safety with wheels and instincts. It is a lot to ask of a rookie, to be sure, but he will have time to ramp up to the week thirteen game. He does not need to stop Jackson cold. The Browns need him to impede Jackson’s effectiveness by a few percentage points.  

Cleveland has been a “disappointing-darling” of sharp bettors and syndicates since 2016. Between 2016 and 2019, the Browns went 23-40-1 against the spread, 36.5%. They finally delivered a successful season in 2020 but disappointed their backers with an 8-10 record against the spread, 44.4%. The math has liked the Browns for some time but has little to show for it. This indicates that the books were ahead of the curve on the Browns.

This year is the first year I see a backlash against the Browns. Perhaps it is bettors’ fatigue. It could also be the misleading way in which they lost to the Chiefs. The narrative that emerged is that they should have won the game. I agree. However, not because they are better, but because Mahomes got hurt.

The model trusts the Browns because the coaching, talent, and scheme fit perfectly. Their non-standard opponent win percentage is only 42%. They get a massive advantage with the placement of their bye-week. And the model places them securely in the top seven of its power rankings.   

Baltimore Ravens

Projected Finish: 2nd, AFC North

Projected Range: 10 -12 Wins

Las Vegas Win Total: 11

Model Win Projection: 11.23

As the season approaches, the Ravens are losing backers by the score. The model is not as convinced. Some are pointing to JK Dobbins’ injury. Dobbins is a fine player, but there is no reason his injury should change your mind about the Ravens. Running backs are not that important. Baltimore has made do with a different leading rusher for five-plus straight years. It has not impacted them, and it will not this season. 

Did that just happen?

There was a quirk that occurred after Dobbins’ injury. Several books, including two off-shore books, still had season-long props on Dobbins. As you might imagine, a large number of people tagged the unders on these numbers for the better part of 12 hours in some instances. 
This is an example of straight arbitrage. Some have cried foul, but I do not see it. The books make it clear that we are responsible for the accuracy of our tickets. They are responsible for the viability of the lines they deal.

This is not the same situation where a typo led to an inaccurate line being posted. In this scenario, the books left a valid bet up beyond its viability. It is more similar to the books simply being beaten by bettors moving more quickly on the information.

There are causes for concern with the Ravens. Particularly with the schedule starting in week ten. The Ravens will play four of their next five games on the road. In weeks ten and eleven, Baltimore will travel to Miami and Chicago. They return home week twelve for a critical game against Cleveland. Then hit the road, albeit a short trip, to Pittsburgh for week thirteen and Cleveland in week fourteen. Cleveland will have their bye week between the two matchups. Baltimore then returns home to face Green Bay. That is a brutal stretch and includes the two divisional games against their most significant threat to the title.   

There has been a widespread, recent cooling on the Ravens in the analytics community. There are regression indicators that are significant in Baltimore. They were the recipient of uncommon field goal luck in 2020. They recovered fumbles at an unsustainable rate. Most importantly, they faced one of the easiest schedules in the league last season, and this season they face one of the most challenging schedules. 

They could only score 20 on a shaky Tennessee defense in the first round of the playoffs. Then only three points on Buffalo. Their style of play is radically different from any other offense in the NFL. The theory is starting to percolate that maybe Baltimore is an excellent regular-season offense. While that is common in the NBA, where the teams play a series in the playoffs, it is not common in the NFL. 

Conversely, Baltimore went 11-5 last season. They outscored their opponents by a mind-blowing 10.4 points per game. The organization relies heavily on analytics. They play a unique style which always creates an edge. Under Harbaugh, the team has won at least ten games in the previous 13 seasons, 76.9% of the time.

The model may end up looking foolish, it hasn’t so much cooled on Baltimore as it has warmed to the Browns. The margin it sees between Baltimore and Cleveland is only .31 wins. It is entirely determined by schedule, especially weeks 10 – 16. And if you believe that Baltimore will finish second in the AFC North, there is value to be found. 

Pittsburgh Steelers

Projected Finish: 3rd, AFC North

Projected Range: 6 – 8 Wins

Las Vegas Win Total: 8.5

Model Win Projection: 8.3

The Steelers have had an incredible run under Tomlin. Though he can make some head-scratching decisions, his teams are successful by any metric. This year is going to be tough in Pittsburgh. Talent-wise they are the third-best team in the division. They have the most challenging schedule in the division, and it is not particularly close. To be successful, not only will they have to be better than the Browns and the Ravens, and they will have to do it against better competition consistently. Mathematically, that is unlikely. 

Pittsburgh had played one of the three easiest schedules last season, including the NFC East and the AFC South. They did not implode late in the season. They played above their ability early against easier competition. The regression was foreseeable. This year they will face one of the three most demanding schedules in the NFL. The schedule includes the NFC North and the AFC West, with non-standard games against the Bills, Titans, and Seahawks.

The Steelers fired OC Fichtner and his paint-by-numbers offensive playbook after last season. Quarterbacks coach Matt Canada replaced him. This constitutes a massive upgrade. Fichtner’s offense was one of the most predictable offenses in the NFL. Unlike Fichtner, Canada favors using different formations, motions, and deception to give his offense an edge. The change should help lower the predictability of their offense and Ben’s air yards specifically. 

The disturbing news is that from ownership down through the coordinators, Pittsburgh has been vocal about wanting to get back to running the ball this season – they shouldn’t be. The Steelers are not just vocal about it. They acted on it by drafting Harris in the first round – they shouldn’t have. Harris is a fine player, but that was a mistake. The problem with that approach is that running does not help you win. Running backs need an offensive line to be productive, and Pittsburgh’s line has massive question marks.  

Back to Running the Ball?

Sharp Analytics has pointed out that the Steelers best season running the football in the last decade came in 2015. The year Bell was lost to injury, and a 32-year-old Deangelo Williams, a nine-year vet that Carolina had cut during the preseason, was the lead back. That season, the team ended with a 0.0 EPA on running plays. So running did not help them score, but at least it did not hurt them. That was their best season in the last decade, and ownership wants to run the ball more?

The reason for the improvement in 2015 was neither Williams nor that Bell was injured. It was an elite offensive line and Munchin’s coaching. If you have that, you can find somebody to run the ball without spending a first-round pick. If you do not have that, a first-round pick will not help.  

The defense remains an outstanding, top-ten unit that out-performed the league-median substantially last season. Still, there will be regression-based purely on the quality of the 2021 opponents’ offensive prowess. The defense is capable of keeping them in games. To win them, the Steelers need to recover fumbles, benefit from field goal luck, and prevail in one-score games at a rate well outside of the norm.

This is likely Ben’s last season in Pittsburgh. If the model is correct, the Steelers are in purgatory already. They will be the only team in their division without their quarterback of the future. (I am not aware of anyone that believes that player is on the roster) Plus, they will be a mediocre team that is likely to get shut out of the top prospects in the draft.        

Cincinnati Bengals

Projected Finish: 4th, AFC North

Projected Range: 5-7 Wins

Las Vegas Win Total: 6.5

Model Win Projection: 6

The Bengals have a disconnect between Zac Taylor’s offense and Joe Burrow’s strengths. Before we get to that issue, perhaps we can agree that he needs protection. The Bengals remade their offensive line in the offseason, and there is some cause for optimism with Reif and Williams starting. Many, myself included, thought they should have taken Sewell with the fifth overall pick. Instead, they reunited Burrow with his college teammate wide receiver, Ja’Marr Chase.

Taylor is either on the receiving end of an unequaled run of bad luck, or the outlying data points to something that needs to change in his coaching. Taylor has been the head coach of the Bengals for 32 regular-season games. Cincinnati has led in 24 of those games, 66%. The team has only prevailed in six of the games in which they had a lead, 25%. The Bengals are only winning 10% of their one-score games. It is necessary to wonder if Taylor is blindly stubborn to a philosophy that does not fit his team’s personnel. 

The biggest issue is that Taylor has Burrow doing what he does worst, more than 29 other teams in the league. Burrow is an accurate passer with touch. He does not have a cannon for an arm. Yet, Taylor is dialing up outside the numbers, routes as though he had a young Elway at quarterback. The results have not been good. The other issues are moot until Taylor figures out play calls that protect Burrow from pressure and play to his strengths. If he does not figure it out this season, it will be someone else’s problem next season. 

The Bengals schedule is tough. They have to play Cleveland, Baltimore, and Pittsburgh twice each. They also have to play the AFC West and NFC North. They get slight relief in their non-standard games with Jacksonville and the Jets, but they also draw the 49ers. This is a team that has gone under on its win totals for five straight years. The model says it is about to be six. 

The path to the over this season is straightforward. The Bengals’ defense performed precisely on the median last season, which is good enough. If the offense can suddenly figure out their approach and pick up a positive regression in a few key areas, then the over is in play.  

The Model Sees Value in the Following Sides:

  1. Cleveland to Win the Division, +155
  2. Pittsburgh 6-8 Wins, +160
  3. Baltimore to Finish Second, +180
  4. Pittsburgh to Finish 3rd, +120
  5. Beckham Over 6 Touchdowns, +100 
+ posts