Last week I posted my NFC win predictions using cumulative team value (AV), which has a strong correlation to win percentage. Now here is the AFC.
For background of the model, please see this post: https://phillycovercorner.com/2021/08/projecting-win-totals-using-team-approximate-value-av-metrics/
AFC East
| Team | Exp Team Value | Exp 2021 Wins | vs. Over/Under |
| Buffalo | 253 | 13.3 | Over (11.0) |
| New England | 214 | 9.2 | Push (9.5) |
| Miami | 183 | 8.6 | Under (9.5) |
| New York Jets | 175 | 5.1 | Under (6.0) |
Buffalo
Josh Allen had a great 2020, finishing 3rd in composite EPA+CPOE but is a puzzle. Accuracy (as measured by CPOE) is a pretty stable metric from college to pro and QBs usually don’t improve much after around 200 career pass attempts (some data here from @ConorMcQ5). But Allen took his a mid-50% completion percentage and a 0.04 CPOE in college and performed the same his first two years in the league, finishing 41st of 43 eligible QBs in CPOE in 2019-2020. Then in 2021, Allen became an elite QB. Could 2021 be an outlier? Maybe, but there are many reasons to believe that Allen is the QB we saw in 2021. The Bills forecast out improving on offense from 5th in the league to the top offense in 2021 as they return the majority of their offense and have the 4th rated OL for pass block win rate. On the defensive side, the Bills should again be near a top 10 unit as they also return the majority of their starters and have added depth this year. Buffalo also goes from an average schedule in 2020 (16th) to the 3rd easiest schedule in 2021. While they did outperform their expected win total last year by 12% (2 wins), they are another team when you look at their 2020, the close games they won they should have won. There is risk of regression or “bad luck” in 2021 here, but when you look at their schedule it’s hard to find more than 2-3 games that could/should be losses.
New England
The Patriots will be interesting – they have a good OL, ranking 6th in run block win rate and 13th in pass block win rate, will seemingly be a run-heavy team and appear to be a run-heavy offense, but maybe not. Like all teams with a rookie QB, the season is going to be highly dependent on how Mac Jones does. The Patriots are all in on Jones, cutting Cam which makes sense – Jones had elite CPOE in college and in a limited preseason sample size, showed the same. He was the top-rated rookie QB with a 92.2, 1st in adjusted completion percentage, 3rd in CPOE, and 2nd in EPA, 1st in under pressure rating, and 2nd in passing grades in play-action with a 93.7 rating and 114 passer rating. Cam was 27th in the league in composite EPA+CPOE and near the bottom of the league in explosive passes. Damien Harris is good – he is 85th percentile in explosive runs and one of the few RBs with a positive EPA. Their OL is good. And they have hopefully improved their offensive weapons with the addition of maddeningly inconsistent Nelson Agholor, but who was 90th percentile in explosive receptions last year and 79th percentile EPA. There is a distinct offensive EPA advantage when play-action is effectively used and the Patriots are set up to have a good, balanced offense this year that pairs with a dramatically improved defense with the return of Dont’a Hightower and additions in free agency and the draft. Their schedule is more difficult than it appears with 5 games you pencil in as “should be easy wins” (Jets twice, Texans, Saints, and Jacksonville), but if Jones is good, they should improve on their 7 wins last year.
Miami
A lot of people love the Dolphins as a team to surprise with their strong defense and hope that the offense improves, but there are areas of caution. First, no team had better “luck” in 2020, both outperforming their expected win total by almost a full win and benefiting from turnovers. I mentioned elsewhere that I don’t think turnovers are as random or “lucky” as most others believe and there is a bit more year-over-year stability of turnover data than most realize, but the source of turnovers matter. A QB that throws interceptions will be more stable year to year while recovering fumbles or interceptions are less so. The Dolphins are on the wrong side of both types – while Tua had the 12th best interception rate in the league last year, when you adjust for dropped interceptions he goes to 31st in the league. And Xavien Howard had an amazing year with 10 interceptions which helped Miami generate the 2nd most turnovers in the league, but this is very likely to regress (as a comparison, in 2019, Miami was 24th in the league in turnovers generated). Similar to Hurts, I expect Tua to improve in his second year. In 2020, he was 30th in the league in composite EPA+CPOE but like Hurts, Tua had a high CPOE in college which translates to the NFL. The addition of Will Fuller will help as he is 95th percentile in explosive receptions, an area of need for the Dolphins as Tua was the 3rd worst QB in the league (ahead of Goff and Darnold) at explosive plays generated. The Dolphins have a decent floor on wins though because of their schedule – they play the Jets twice, the Texans, Giants, Raiders, Saints, and Jacksonville. Like Philly, it all comes down to their OL improving, Tua’s advancement, and how their new weapons help the offense.
New York Jets
Continuing the theme of the AFC East, the Jets season will largely be determined by a new QB. They have two new starters (rookie Alijah Vera-Tucker and Morgan Moses) on the 30th ranked OL which should be better. Like Mac Jones, Zach Wilson graded out well (second behind Jones) in a very limited preseason with only 20 dropbacks. But since Darnold was 36th in the leage last year in composite EPA+CPOE (only ahead of Wentz and Dwayne Haskins), it is right to think Wilson will be an upgrade. As for weapons, they have effectively swapped Breshad Perriman for Corey Davis, an 8 AV improvement which equates to 4.5% of the total team value. The defense forecasts to still be bad, forecast to be 30th in the league. The Jets get a slightly easier schedule in 2021, going from the 3rd hardest to the 9th hardest. While there are few teams on their schedule that look like definite wins, the team value should return to near their 2019 level where they were still a bad team but snuck some unexpected wins.
AFC North
| Team | Exp Team Value | Exp 2021 Wins | vs. Over/Under |
| Cleveland | 243 | 12.3 | Over (10.5) |
| Baltimore | 228 | 10.7 | Push (11.0) |
| Pittsburgh | 189 | 6.6 | Under (8.5) |
| Cincinnati | 179 | 5.4 | Under (6.5) |
Cleveland
The Browns significantly outperformed their expected win total last year by 19%, equal to 3 wins. And while they have the 3rd easiest schedule in 2021, they also had the 3rd easiest last year. Despite that, they still model out to improve on last year with 12.6 wins as they added more than they lost in free agency. Their weakness last year was defense and it should be be improved, going from 25th to 18th in the league. The worst part of the defense was the secondary and they have added Troy Hill (71.9 coverage grade) and John Johnson (86.6 grade) who are big upgrades over departed Terrance Marshall and Karl Joseph. And while it is difficult to rely on rookies, they drafted Greg Newsome and JOK (a LB that had the second best slot coverage grade in the NCAA last year) that will see good playing time. The Browns DL is still not great, 14th in pass rush win rate and only 30th in run stop win rate last year and neither Clowney nor Malik Jackson should be expected to improve that much. On the offensive side, the Browns should have one of the top offenses in the league – their OL returns and ranked 2nd in pass block win rate and 13th in run block win rate last year. Baker ranked 11th in the league in composite EPA+CPOE and is 77th percentile in explosive passes, despite the Browns being a heavier run team.
Baltimore
The Ravens underperformed their expected win total last year, going 1-4 in one score games. Their defense should be ranked near top 10 with a good secondary and their offense will be better, predicted to be 6th in the league. Lamar Jackson was 16th in the league in composite EPA+CPOE in 2020 but what he can do running is what matters, being one of the highest value QBs in the league when you add rushing. The Ravens had a top 10 OL in 2020 and are changing several players on the line (Villanueva and Ronnie Stanley back from injury for Brown and Fluker at the tackles, Bozeman for Skura at center), but net should be equal to last year. This is one of the forecasts that is my least favorite – it’s hard to find a lot of losses on their schedule and I think 11 wins could be low.
Pittsburgh
Along with the Cardinals, the Steelers is probably the team I am most different from consensus on. And it’s a prediction I’ve looked at over and over because it is saying the Tomlin will do something he has never done which is win fewer than 8 games. But I just don’t see a good 2021 season for them. First, the Steelers are likely to regress in almost every area: in 2020, they benefited from good luck and outperformed their expected wins by 9% (1.5 games), going 7-2 in one score games. They were the 5th healthiest team in 2020, which is likely to regress. And they go from one of the easiest schedules in 2020 to the league’s most difficult schedule in 2021. Because of their cap situation, they lost much more in free agency than they gained, losing 3 of 5 starting OL, Steven Nelson and Mike Hilton in the secondary, and LB Bud Dupree while only adding OL Trai Turner as an expected starter. Their OL was 24th in run block win rate and 28th in pass block win rate last year and will be worse in 2021, predicted to be the second worst OL in the league. Nobody should question Kevin Colbert as the Steelers have been amazing at personnel, but it reeks of “let’s try one more time with Ben”, ignoring analytics and going RB and TE in the first two rounds when impact OL prospects like Cosmi, Radunz, and Creed Humphrey all were there to get in the 2nd at their pick or with a small trade-up. Now, the Steelers swap Villanueva (74.8 blocking grade) for Okorafor (57.5) at LT and bring in Zach Banner at RT who may not be recovered from his ACL injury. Rookie 4th rounder RT Dan Moore is likely to play to start the season. He played well in the preseason but it is a lot to ask a rookie tackle to be impactful. They had the top-rated defense in 2020 and it is still expected to be very good, but with a worse secondary (even after the late trade for Ahkello Witherspoon), their defense is forecasted drop some to 9th in the league. And Ben has clearly been regressing, finishing 24th in the league in composite EPA+CPOE last year, leading the Steelers to 30th in the league in explosive play rate and a plodding offense with one of the lowest yards per completion in the league and trying to reinvigorate a running game while ignoring the OL as the common issue. If somehow the OL is better and the secondary holds up, I could be horribly wrong here but this is one of the team predictions I feel most comfortable with.
Cincinnati
5.5 wins is one that still feels high to me as the Bengals have the 10th hardest schedule in 2021, Burrow will only be 9 months post his injury and while cleared to play, I cannot imagine he will be near 100%, and the Bengals look to have bottom of the league offensive and defensive units. Much has been written on their draft and I am one of many that thought a tackle was way more valuable than a WR (even if Chase was lighting it up). The Bengals were 29th in pass block win rate and 27th in run block win rate and the only change was Riley Reiff (71.4 grade), an upgrade over Bobby Hart (66.3). But that isn’t enough. It’s hard to find more than 4 wins on their schedule, but I will leave the model where it is.
AFC South
| Team | Exp Team Value | Exp 2021 Wins | vs. Over/Under |
| Tennessee | 222 | 10.1 | Over (9.0) |
| Indianapolis | 199 | 7.6 | Under (9.0) |
| Jacksonville | 175 | 5.0 | Under (6.5) |
| Houston | 156 | 3.0 | Under (4.0) |
Tennessee
The Titans had the second easiest schedule in 2020 but now have the league’s easiest schedule in 2021. They did have a good amount of “luck”, outperforming their expected wins by 12% or 2 wins. While there is a lot of debate analytically about Tannehill, he has been at the top of the league in composite EPA+CPOE, finishing 2nd behind Aaron Rodgers last year. He is clearly helped by Derrick Henry and benefits from getting to pass against a stacked box more than most QBs, but his results are what they are. Their offense will be top 10 in the league and they replace Corey Davis with Julio, and there is some hope their OL can be a bit better than last year where they finished 20th in run block win rate and 24th in pass block win rate. On the defensive side, they should improve from one of the worst defenses in the league to 24th – they added in free agency with Bud Dupree and Janoris Jenkins and should expect a better secondary with advancement of their youth (Fulton, Farley, Molden). I think this will be a worse division than expected and despite some negative regression, Tennessee should still be the top team here and exceed their win expectations.
Indianapolis
The Colts are a difficult team to model. On paper, they still have a good team with top-half of the league offensive (12th) and defensive units (13th). But Wentz is a question mark and is clearly a downgrade from what Indy had last year with Philip Rivers – Rivers finished 12th in the league in composite EPA+CPOE (right with Dak Prescott, Baker Mayfield, and Russell Wilson) and Wentz has not had a season like that since his 2017 season. And, while it is difficult to model injuries as the Colts were about league average last year with adjusted games lost due to injury, it sure seems like 2021 is headed towards a high impact injury year. Wentz had foot surgery and is missing time due to Covid exposure, three offensive linemen are out with Quentin Nelson (foot surgery), Eric Fisher (achilles), and Ryan Kelly (less serious elbow injury and Covid), and TY Hilton with neck surgery. They keep a relatively easy schedule with the 6th easiest schedule, but had the league’s easiest schedule in 2020. I don’t love this forecast as so much comes down to health and how Wentz plays. But even with their injuries, there are 7-8 games on their schedule that should be wins as they get Houston and Jacksonville twice as well as Arizona, the Jets, Raiders, and Dolphins. But I am confident they are a worse team than last year and not likely to be in the playoff hunt.
Jacksonville
Jacksonville’s three QBs last year finished 25th (Minshew), 43rd (Glennon), and 45th (Luton) in composite EPA+CPOE – Trevor Lawrence doesn’t have to have an elite rookie season to expect a sizable upgrade at QB. Of rookie QBs the past two seasons, five (Tua, Burrow, Herbert, Lock, Murray) had rookie seasons better than the Jaguars QBs last year and only three (Daniel Jones, Haskins, and Minshew) were worse. Lawrence will be a definite upgrade. That said, they are still expected to be near the bottom of the league on offense (27th) and defense (26th). the upside is that Jacksonville underperformed their expected win total last year by 11% (2 games) and have the 6th easiest schedule after having a middle of the league SoS last year. This is a forecast I don’t love and think Lawrence gives them a chance to outperform their 5.1 expected wins just because of the teams they play.
Houston
Where to start… Houston will obviously be bad as it is not expected Watson will play. They will get worse on defense, forecasted to be 31st in the league, and will obviously be worse on offense, dropping from middle of the league to 23rd. They were big net losers in free agency, losing one quarter of their expected team value led by Nick Martin, JJ Watt, Will Fuller, and Zach Fulton and replacing their holes with lesser players (WR Anthony Miller, DL Maliek Collins) or question marks (C Justin Britt who hasn’t played in two years). The Texans underperformed their expected win total last year by a good amount (11%) and went 2-5 in one score games, but they just don’t have much talent and sometimes losing close games isn’t just luck.
AFC West
| Team | Exp Team Value | Exp 2021 Wins | vs. Over/Under |
| Kansas City | 247 | 12.7 | Push (12.5) |
| LA Chargers | 227 | 10.6 | Over (9.5) |
| Denver | 225 | 10.4 | Over (8.5) |
| Las Vegas | 172 | 4.7 | Under (7.0) |
Kansas City
The Chiefs actually outperformed their expected wins last year by a league-leading 19% (or 3 wins). Under or over performing expected wins, which are based on actual point differentials, is generally an area of regression but sometimes they aren’t luck, especially when you have Mahomes – the best QB in the league will win you close games. The Chief go from the 8th easiest schedule to the 6th hardest, but again, when you are one of the top teams in the league it is hard to find many losses on their schedule. On the positive side, their clear weakness last year was the offensive line and it has clearly been upgraded with Orlando Brown at LT and snagging Creed Humphrey (who should have been taken earlier by center-needy teams like the Rams or Steelers) is a big upgrade – he had a 73.7 grade in the preseason and is slated to start under Reid who rarely starts rookies. When your expected win total is 12.5, it is hard to exceed that but I think the Chiefs could do so.
Los Angeles Chargers
The Chargers model is one I had to re-look at as it has a fairly big jump in their defensive ranking for 2021, going from middle of the league to 6th in 2021. But they should be much better – they have 7 players on defense that are young in career where year-over-year improvements are historically seen and modeled in. They also get Derwin James back from missing all of 2020 and Joey Bosa who missed almost half the season – the difference of those two alone account for 7% of the teams total expected value. And they didn’t lose anything in free agency – Casey Hayward (59.5 coverage grade) will be replaced by rookie Asante Samuel, but Samuel doesn’t have to be great to be a plus and Uchenna Nwosu equals Melvin Ingrahams output (both graded out at 72 and have equal AV value). On the offensive side, Herbert was middle of the league statistically in his rookie season, finishing 18th in composite EPA+CPOE and middle of the league in explosive plays and QBs will typically improve into the second year. I am not as bullish on Herbert as many today as his college CPOE (which is one of the most stable metrics from college to pro) was average.
Denver
Denver goes from the 5th hardest schedule in 2020 to the 5th easiest in 2021. Obviously QB is the question area and Denver made a mistake passing on Fields, but Bridgewater is an upgrade over Lock – Bridgewater was 19th in the league in composite EPA+CPOE vs. Lock at 38th. Their defense is forecasted to improve and be the 4th best defense in the league with the return of Von Miller and riches in the secondary. Their offense was one of the worst in the league last year but should improve all the way to 25th in the league… joking aside, Bridgewater should be a good enough upgrade to make use of their talent at WR and TE and improve on their 4-7 record in one-score games.
Las Vegas
There isn’t really anything that you can point to as improving the Raiders over 2020. They lost more in free agency than they gained and are counting on rookie Alex Leatherwood at LT and Nick Martin or Andre James to replace C Rodney Hudson. They outperformed their expected wins last year by almost a full win, meaning regression is likely. They also go from the 6th hardest schedule in 2020 to the 3rd hardest schedule in 2021. Carr is a good QB, finishing last year 9th in the league in composite EPA+CPOE. The upside for the Raiders are a bigger jump by their young players (Ruggs, Bryan Edwards, Josh Jacobs, Amik Roberson, and rookie Trevon Moehrig). This is a forecast that I think could be low because of the young players, but the Raiders are more likely to be worse than their line than better.
AFC Summary
Below summarizes the AFC with each team’s modeled 2021 win projection, the current over/under win total betting line, difference from O/U line, 2020 actual win total, and difference from 2020 wins.
A few call outs:
- The Bills, Browns, Titans, Chargers, and Broncos are the teams modeled to most beat their current over/under win totals (more than a one game positive difference).
- The Bengals, Steelers, Colts, Jaguars, Texans, and Raiders are the teams most expected to underperform their over/under win totals.
- The biggest year-over-year win total improvements are the Broncos (+5.4), Jaguars (+4.0), Chargers (+3.6), Jets (+3.1), and Patriots (+2.2).
- The biggest year-over-year win decreases are the Steelers (-5.4), Colts (-3.4), Raiders (-3.3), Dolphins (-1.4), and Chiefs (-1.3).
As explained in many of the team descriptions, there are some of these I like more than others. In the AFC, I feel most comfortable with the predictions for the Steelers, Broncos, Colts, and Browns.
I included this in the NFC win prediction post, but wanted to include again here – this shows the average year-over-year win total changes in the league. This shows that it isn’t uncommon for large win total swings year-to-year.
