AFC Championship Preview 2021
This year's AFC Championship Game, like the NFC Championship Game, is a rematch of a Week 6 game. Buffalo's first contest with Kansas City was much closer than the first meeting of Green Bay and Tampa Bay, and ended with Kansas City taking a 26-17 victory. This week's game may be similar, or it may be completely different. We may be able to learn something about each team's strategy by looking back at the Week 6 meeting. But what we definitely know is that these are not quite the same teams that met on that rainy October night in Buffalo.
In particular, the Buffalo defense has been much better in recent weeks while the Kansas City defense has declined. Buffalo ranked 30th in defensive DVOA for Weeks 1-6 at 13.2%. Since Week 7, and including the playoffs, Buffalo ranks sixth in defensive DVOA at -12.5%. Kansas City's defensive decline is clearer if we look just at pass defense. For Weeks 1-6, through that first meeting with Buffalo, Kansas City ranked fourth against the pass at -8.5% DVOA. Since Week 7, Kansas City ranks 19th at 9.8% DVOA against the pass. That includes last week's game against Cleveland, which was their best defensive performance of the last couple months.
There has been a lot of discussion about whether the Bills can slow down the Kansas City offense. That's very hard to do: Kansas City's only offensive game below average this year was the week Patrick Mahomes threw three interceptions against Miami. But a bigger question might be whether the Chiefs can slow down the Buffalo offense. Given how well the Bills are playing and the decline of Kansas City's defense until last week, that's going to be pretty difficult as well.
For those who may be unfamiliar with the Football Outsiders stats, they are explained at the bottom of the page. Scroll down or click this link. Game charting data appears courtesy Sports Info Solutions, unless noted. All stats represent regular season only, except for weighted DVOA and anything else specifically noted.
Buffalo at Kansas City
|DVOA||23.8% (4)||23.0% (5)|
|WEI DVOA||38.0% (1)||25.5% (5)|
|Bills on Offense|
|BUF OFF||KC DEF*|
|DVOA||15.6% (5)||2.4% (18)|
|WEI DVOA||19.9% (4)||1.8% (18)|
|PASS||43.3% (3)||3.3% (13)|
|RUSH||-15.0% (22)||1.2% (29)|
|Chiefs on Offense|
|BUF DEF||KC OFF*|
|DVOA||-2.2% (12)||24.9% (2)|
|WEI DVOA||-10.7% (8)||24.3% (3)|
|PASS||2.2% (12)||50.0% (2)|
|RUSH||-8.2% (17)||-3.9% (12)|
|ST DVOA||5.9% (4)||0.5% (17)|
|*Week 17 not included.|
All readers can click here for the open in-game discussion thread. If you have FO+, you can click here to see all the matchup of DVOA splits for this game.
Note that we've removed all Week 17 stats for Kansas City from the table above, since the Chiefs sat so many starters on both sides of the ball that week.
WHEN THE BILLS HAVE THE BALL
Buffalo is a pass-first offense, although they don't necessarily have to be against Kansas City because the Chiefs are weak against the run. For the year, including playoffs, the Bills ran on only 30% of plays in the first half of games. That was the third-lowest rate in the league. (This includes scrambles as pass plays.) However, in six games against the 10 worst run defenses in the league, Buffalo's run/pass ratio increased to 38%. Even after adjusting for opponent, Buffalo was a little better against these defenses: -7.8% DVOA and 4.4 yards per carry against bottom-10 run defenses, compared to -21.5% DVOA and 3.8 yards per carry in the other 10 games. The one bad run defense that Buffalo really struggled against was ... Kansas City in Week 6. The Bills had only -30.6% DVOA and 3.4 yards per carry in that game. The Bills rushed on nearly half of their plays in the first half of that game. In the second half, after falling behind, they called only two runs.
The Bills did not struggle in the first game when Allen took the ball to run himself. His four designed runs included conversions on second-and-10 and third-and-2. And speaking of places where the quarterback often runs the ball himself, one place where it's good for the Bills to run is in short-yardage situations. Both of these coaches are aggressive on fourth downs. The Bills were only average running the ball in short yardage, converting 66% of the time, but the Chiefs defense ranked dead last, allowing 78% conversions on short-yardage runs.
So the Bills will mix it up a little more, but even with the Chiefs' weakness against the run it is more efficient for Buffalo to pass the ball. Their run offense only ranked 22nd in the NFL; their pass offense ranked third, trailing only Green Bay and Kansas City during the regular season. It's particularly good for the Bills to air it out right away on first down. They ranked first in pass offense DVOA on first downs and just 29th in run offense DVOA.
Numbers suggest that the Bills want to spread it out and spend a lot of time in 10 personnel with four receivers. They want to take advantage of their depth at the position, with rookie Gabriel Davis having an excellent season (10.7% receiving DVOA) as the fourth receiver. (Davis is listed with an ankle injury and has missed practice this week; if he can't play, the fourth receiver will be Isaiah McKenzie, a speed jitterbug who had two touchdowns in Week 17.) During the regular season, only Arizona used 10 personnel more often than Buffalo, and the Bills had 39.5% DVOA and 7.0 yards per play with this personnel group on the field. The Bills also excelled in 10 personnel in the first game against Kansas City. In 33 plays with 11 personnel, the Bills averaged 2.8 yards per play with -34.1% DVOA. But in 18 plays with 10 personnel, the Bills averaged 7.9 yards per play with 70.0% DVOA. (There's no sense in quoting season-long numbers about the Chiefs against 10 personnel, since Buffalo was their only opponent to use this grouping heavily.)
The Bills also had the league's biggest gap between DVOA in shotgun (third) and DVOA with Allen under center (18th). Both the passing game and the running game were better with Allen in shotgun.
The Chiefs brought a lot of pressure against Josh Allen in the first matchup between these teams. SIS recorded pressure on nearly half of Bills pass plays (48.5%). For the season, Allen was one of the league's best quarterbacks under pressure, but that still means there was a big gap between his performance under pressure (-11.3% DVOA, 4.9 yards per play) and his performance without pressure (65.9% DVOA, 8.7 yards per play). As an aside, while Allen was the league's No. 3 passer with pressure by DVOA, Patrick Mahomes was the league's No. 2 passer with pressure. (No. 1 was Ryan Fitzpatrick, who was unsustainably almost as good with pressure as he was without pressure.)
The Chiefs are probably going to bring a lot of pressure this week as well. For the season, SIS has the Chiefs ranking seventh with a 29% pressure rate on defense, and the Bills were just 27th with a 30% pressure rate allowed on offense. A lot of Kansas City's pressure comes up the middle, as defensive tackle Chris Jones tied for third in the league with 39 hurries. He's not the only danger on the inside; rookie Tershawn Wharton had 15 hurries this season. It's good for Buffalo that they have Jon Feliciano back in the lineup at right guard; he missed the first meeting of these teams.
Kansas City also brings pressure by bringing extra defenders. In the Week 6 matchup, the Chiefs blitzed on a little more than half of all pass plays. Five-man blitzes worked well, holding Allen to just 2.1 yards per play on nine plays. Bigger blitzes did not do well, as Allen had 8.6 yards per play on eight plays. During the entire regular season, Allen was very good against blitzes, with a slightly higher DVOA and roughly the same yards per play as when he faced four pass-rushers. However, Allen faced a ton of defensive back blitzes, more than any other quarterback in the NFL. Opponents sent at least one defensive back at Allen over 20% of the time according to ESPN Stats & Info. Buffalo's pass DVOA did drop a little to 7.8% on these plays. And it just so happens that the Chiefs send a lot of defensive back blitzes, more than any defense other than Seattle and Buffalo, and the Chiefs do well with these plays, with their defensive DVOA improving slightly to -0.7%.
One way to get away from the pass rush is to dump the ball off to your running back, assuming that he's not staying in to pass protect. Kansas City ranked 31st in DVOA against running backs in the passing game. However, in the first matchup, Devin Singletary had just one catch for 13 yards.
Derrik Klassen wrote in this week's Film Room about the Chiefs' frequent use of man coverage. They played 40% man coverage this year, one of the higher rates in the league. The played man and zone roughly 50-50 in the Week 6 game. Man coverage does allow Josh Allen to scramble, which he did four times in the Week 6 game with three conversions for a new set of downs.
For the season, the Chiefs were better in man than zone. They allowed 6.8 yards per pass and 47% success rate in man, compared to 7.6 yards per pass and 54% success rate in zone. And Josh Allen on the season had a lower success rate against man coverage (52% compared to 56% against zone) although he had most of his interceptions against zone (eight of 10). In Week 6, however, Allen was better against the man coverage. He averaged 5.6 yards per pass with a 50% success rate against man, including two DPIs but not including his four scrambles. He averaged 5.0 yards per pass with a 40% success rate against zone, plus the game-ending interception.
The Chiefs did not have L'Jarius Sneed in the Week 6 matchup, and the rookie fourth-round cornerback has had a very good year. He doesn't have enough targets to be ranked among cornerbacks, but he allowed just 5.1 yards per pass with a 56% success rate. Sneed plays mostly in the slot, but not entirely. Charvarius Ward was the weaker of Kansas City's other two cornerbacks. Ward allowed 7.4 yards per pass with a 50% success rate, both ranking in the 30s. Bashaud Breeland allowed 6.8 yards per pass with a 54% success rate, both ranking in the 20s. Breeland is in the concussion protocol but is probably going to play on Sunday night.
One last note: Kansas City had the worst red zone defense in the league by DVOA. However, there's some question as to how much that really matters, as the difference between defense overall and defense in the red zone doesn't seem to show any predictive value from one season to the next or from the first half of the season to the second half of the season. The Bills offense was as good in the red zone as it was overall.
WHEN THE CHIEFS HAVE THE BALL
Let's start with the elephant in the room. Yes, Patrick Mahomes is in the concussion protocol. There's a small chance that he will be unable to play on Sunday, but he has been practicing this week and all indication is that he will pass the necessary tests and start on Sunday night.
On the off chance that Chad Henne has to start the AFC Championship Game, obviously the Bills become substantial favorites. But this does not mean the Chiefs are toast! Andy Reid has a very good history with his backup quarterbacks. Last year, the Chiefs had 18.4% offensive DVOA in three games with Matt Moore (the game where Moore replaced an injured Mahomes, followed by two starts). This year, Henne managed the Chiefs to 5.9% offensive DVOA in Week 17 even though he was playing without his best offensive teammates. The Chiefs can still make it a game with their backup quarterback. They probably won't have to, and we'll write the rest of this preview assuming that Mahomes is the guy.
UPDATE: Mahomes announced on Friday afternoon that he's out of the concussion protocol and will start Sunday.
The biggest question about this side of the matchup is whether the Bills are going to play defense the same way they did back in Week 6, or whether they will mix things up. On the season, for example, the Bills blitzed 31% of the time, higher than the NFL average. In the Week 6 game, they essentially never blitzed at all; there is one play where more than four pass-rushers came but it was play-action and the defenders may have just been following the fake. Also on the season, the Bills played man coverage on 27% of passes, zone coverage on 61%, and some kind of mixture on 12%. In Week 6, they played nothing but zones. They played the entire game in nickel, with the safeties back, essentially inviting the Chiefs to run on them.
And the Chiefs did just that. Even in the first half of the game, before pulling out to a lead, the Chiefs had close to a 50-50 split between passes and runs. Those runs in the first half of the game gained 8.0 yards per carry. The Chiefs may want to be a pass-first team but they'll take 8.0 yards per carry every time. The Chiefs took advantage of big holes and broken tackles. The good news for Buffalo is that improved run defense has been a big part of their improvement on defense in the second half of this season. Buffalo allowed 16.3% run DVOA against Kansas City, close to their worst run defense game of the year. But since their bye week, including the playoffs, Buffalo has -24.2% run defense DVOA, seventh in the NFL. The return to health of linebacker Matt Milano has played a big role in that improved run defense.
The one place the Chiefs may not want to run is when we usually encourage teams to run: short-yardage situations. The Bills ranked seventh against short-yardage runs this year, allowing a 61% conversion rate. The Chiefs ranked dead last, converting these runs only 51% of the time. But their aggressiveness on fourth down is fine, because they were actually more likely to convert via the pass: including last week's game, the Chiefs converted 70% of passes in short-yardage situations (1 or 2 yards to go, goal line or third/fourth down).
Returning to the pass, safety Jordan Poyer made statements this week suggesting that the Bills may take a different defensive approach to the Chiefs this Sunday. That's probably a bad idea. If it means man coverage, well, Patrick Mahomes gained 8.4 yards per play against man coverage compared to 7.7 yards per play against zones, with the same success rate. Don't forget that he can scramble for good yardage on these plays, just like Josh Allen. (As for the foot injury Mahomes suffered last week, Andy Reid told reporters this week "I think he'll be OK with the toe part.") And if a different approach means blitzing Mahomes, well, it's going to get bad. Yes, the Bills liked to blitz this year, and they had slightly better defense with a blitz (-6.0%) than with four pass-rushers (7.6%). But Mahomes absolutely destroys the blitz. This year, Mahomes has a league-leading 78.2% DVOA when opponents blitzed him. And when opponents sent defensive back blitzes, something the Bills like to do a lot? Mahomes had 52.6% DVOA, still higher than his performance against the normal four pass-rushers.
The Bills' conservative defense from Week 6 actually did a pretty good job of slowing down the Chiefs passing game. Their biggest gain, 37 yards to Byron Pringle, came on a scramble drill. Tyreek Hill only caught three passes for 20 yards. It's Travis Kelce who might be the biggest problem for Buffalo. He had two touchdowns in the Week 6 game. Although the Bills had an average DVOA against tight ends, they did allow more passes and more yards per game than the league average. Kelce excels at finding holes in zones and also made Tremaine Edmunds look bad a couple times when the structure of the zone calls led to specific player-on-player matchups.
The other problem for the Bills is that if they're going to play the conservative defense, they need to accept that they aren't going to get a lot of pressure on Mahomes. As noted above when I mentioned how Josh Allen did against pressure, Mahomes was the second-best quarterback in the league when pressured this season -- but it's still better to pressure him than not. The Bills had a really hard time getting to him with just four pass-rushers despite some weaknesses on the Chiefs offensive line. SIS only recorded pressure on seven pass plays in the Week 6 game, with one sack. The same player never got marked down with specific pressure twice.
For the season, the Bills finished second in the league in ESPN's pass rush win rate. They were only 15th in SIS pressure rate, however, and they were below average at getting pressure rate with only four pass-rushers. (They got pressure on 23% of these plays, compared to a league average of 26%.) The Bills are going to have to bring pressure with their front four in order to disrupt the Chiefs offense a little bit. Even if they can limit the Chiefs to just a couple fewer first downs than they had in the Week 6 game, it could make a big difference. Buffalo's top defenders in hurries were Jerry Hughes (30), Ed Oliver (19), and Mario Addison (18).
Of course, the problem with bringing the pressure on Mahomes on early downs is that he'll still roast you on later downs. The Bills struggled with the pass on first down, ranking just 26th in DVOA, but the Chiefs "relatively" struggled on first downs, ranking only 12th in passing. However, Mahomes has a history of being unstoppable on third-and-long. This year, the Chiefs ranked first on third-and-medium (3 to 6 yards to go) and second on third-and-long, trailing only -- surprise -- the Los Angeles Chargers.
There's a surprising gap between these teams, as the Bills had one of the league's best units and the Chiefs were around league average. In truth, the difference between the teams probably isn't that big. The Chiefs' rating for this year is dragged down by one horrible special teams game in Week 3 against Baltimore. They were second in 2019 and 11th in weighted DVOA for 2020. The "true quality" of their special teams is probably closer to those rankings than to this year's No. 17 finish.
Interestingly, although punter Corey Bojorquez led the league in gross punting value, our metrics rate the Bills getting more value from kickoffs, punt returns, and kick returns. (Short version: Bojorquez had such a high gross average in part because he only kicked three times all year from the opposing side of the 50.) Andre Roberts has been one of the top return men in the NFL for years. Rookie Tyler Bass was average on field goals during the regular season, though he missed two kicks in the Buffalo winds last Saturday night.
For Kansas City, Mecole Hardman was average returning kickoffs and punts and had a 67-yard punt return touchdown against Miami. Punt returns were the biggest negative for the Kansas City special teams and that was fluky, as most of the negative value came from the occasional return attempt by Tyreek Hill (he muffed his one try) or Demarcus Robinson (remember the safety against New Orleans)? The other big negative was kickoff coverage, which was subpar (and allowed a touchdown in the Baltimore debacle mentioned above). Kicker Harrison Butker only missed two field goals during the regular season but missed six extra points (one blocked).
When two teams play each other in the regular season and again in the playoffs, our predictions for the second game should be guided much more by how well those teams played all season than by one specific game where they faced each other early in the year. Nonetheless, the Week 6 game between Kansas City and Buffalo can tell us a lot about the strategies these teams might follow on Sunday as they try to get past one another and make it to Super Bowl LV.
A lot of Buffalo's strategy here depends on whether they see the Week 6 game as a win for their defense. I think they should see it that way. Look at the Buffalo weekly defensive chart above. The yellow dot is the Kansas City game. It certainly doesn't stick out among Buffalo's first-half defensive performances. And the Bills defense is better now. There's plenty of reason to believe that the Bills defense can play a similar strategy against the Chiefs on Sunday night and perform better than it did in Week 6, especially when it comes to running plays. Holes should be a little bit smaller and tackles should be a little bit better. All the Bills need to do is limit the Chiefs to maybe one or two fewer series, stop them before one more score. Just holding the Chiefs to 21 or 24 points instead of 26 points will help.
The bigger issue in the first game was the Bills offense. Buffalo had one of its worst offensive games of the year against Kansas City in Week 6. They had another negative offensive game last week against Baltimore. The Bills absolutely cannot have a bad offensive game if they want to beat the Kansas City Chiefs. We need to see the Buffalo offense that we saw for the second half of the regular season. The Bills need to hold the Chiefs offense the way they did in the Week 6 matchup, and then they need to outscore them.
I think they can do it. Kansas City is the better of these two teams on offense, and they have the better quarterback (assuming Mahomes plays), but Buffalo is the better all-around team, especially over the last two months. And the Buffalo offense hasn't been too shabby. Conventional wisdom is giving the Bills a better chance to win this game than I would have expected -- I thought the Vegas line for this game would have Kansas City favored by more than just three points. But I also think the Buffalo Bills have a better chance to win than that. This is a very close 50-50 matchup between two very good teams, but I still think how the Chiefs have played over the last two months matters. I think the Bills offense will show up, and I think their defense can be just a little bit better than it was in Week 6. I favor the Buffalo Bills ever so slightly to take out the defending champions and move on to Super Bowl LV.
DVOA (Defense-adjusted Value Over Average) breaks down each play of the season and compares it to the NFL average based on situation and opponent. You'll find it explained further here. Since DVOA measures ability to score, a negative DVOA indicates a better defense and worse offense, and a positive DVOA indicates a better offense and worse defense.
Team DVOA numbers incorporate all plays; since passing is generally more efficient than rushing, the average for passing is actually above 0% while the average for rushing is below 0%.
SPECIAL TEAMS numbers are different; they represent value in points of extra field position gained compared to NFL average. Field goal rating represents points scored compared to average kicker at same distances. All special teams numbers are adjusted by weather and altitude; the total is then translated into DVOA so it can be compared to offense and defense. Those numbers are explained here.
Each team is listed with DVOA for offense and defense, total along with rush and pass, and rank among the 32 teams in parentheses. (If the DVOA values are difficult to understand, it is easy to just look at the ranks.) We also list WEIGHTED DVOA (WEI DVOA), which is based on a formula which drops the value of games early in the season to get a better idea of how teams are playing now (explained here).
Each team also gets two charts showing their performance this year, game-by-game, according to offensive and defensive DVOA. In addition to a line showing each game, another line shows the team's trend for the season, using a rolling average of the last five games. Note that the defensive chart is reversed so upwards is a more negative defensive DVOA (which is better).