AFC Championship Preview 2017
by Scott Kacsmar
The path to get here was difficult, but this is what we expected the 2016 AFC to come down to all along. In our projections for Football Outsiders Almanac 2016, we had New England and Pittsburgh as the top teams in the AFC, refusing to bat an eye at season-opening suspensions for superstars Tom Brady and Le'Veon Bell. Nor were we concerned that Sebastian Vollmer (surgery) and Martavis Bryant (also suspended) were expected to miss the entire season, or that neither Dion Lewis nor Bud Dupree would make their season debuts until Week 11 due to injuries. We may have gulped a bit when the Patriots traded Jamie Collins on Halloween, and again when the Steelers had to place Cameron Heyward, quite arguably their best defender, on injured reserve two weeks later.
We weren't correct about the New England at Pittsburgh game in Week 7 being the AFC's (Regular Season) Game of the Year, because Landry Jones was at quarterback for an injured Ben Roethlisberger in a 27-16 loss. Roethlisberger's torn meniscus threatened to sink the Steelers' season after a four-game losing streak, but it did not. Neither has another season-ending injury to Rob Gronkowski for the Patriots. In Week 10, the Steelers lost to the Cowboys and the Patriots lost to the Seahawks in two epic games, but neither team has lost since, setting up the long-awaited playoff rematch in this season's AFC Championship Game.
It is hard to believe, but the last time the Steelers met the Patriots in the playoffs, Roethlisberger was just a rookie, and I was a college freshman who had never heard of VLOOKUP. This is the seventh postseason since then where both teams qualified for the playoffs, but a single outcome prevented the rematch from happening five times, including Pittsburgh's blown lead in Denver last year that otherwise would have made this last year's AFC Championship Game. So this has been a long time coming.
While we will make several references to the Week 7 meeting between these teams, Bill Belichick said it best at his Wednesday presser when asked about how relevant that contest was to this week's game.
Belichick said the Week 7 Pats-Steelers game "isn't really relevant" now. "I mean, it is, but it really isn't."
— Zack Cox (@ZackCoxNESN) January 18, 2017
The single biggest difference from Week 7 is obviously the addition of Roethlisberger at quarterback over Jones. The other big change also benefits Pittsburgh with Gronkowski out. Pittsburgh will also have Eli Rogers, who did not play that day, and its full offensive line with Marcus Gilbert back at right tackle. Sean Davis has taken over for Robert Golden at safety and Dupree has become a factor in recent weeks.
The Patriots have also seen some changes, particularly during this winning streak. Rookie wide receiver Malcolm Mitchell has shown some playmaking ability, though his playing status is uncertain, and the team acquired Michael Floyd from Arizona. Lewis could be a big factor in this game, and sixth-round rookie Elandon Roberts has essentially replaced Collins in the linebacker corps.
However, the biggest change may be the venue, where the Steelers will have to deal with some bleak history for road teams in New England. Since 2001, including the playoffs, the Patriots at home are:
- 103-1 (.990) when leading by at least eight points at any time in the game.
- 51-2 (.962) when defending a one-score lead in the fourth quarter.
- 87-3 (.967) when winning the turnover margin.
- 88-4 (.957) when scoring at least 25 points.
- 117-17 (.873) in games started and played into the second half by Tom Brady.
Don't lose the turnover battle. Don't fall behind. Don't get into a shootout. Simple, right? Even for how sloppy the Patriots looked against Houston last week, they still won by 18 points. Meanwhile, Pittsburgh mustered all 18 of its points in Kansas City on field goals. That won't fly this week, and the Patriots did allow the fewest points per drive (1.42) in the NFL this season.
That last stat is the other striking similarity in this matchup. Sure, neither the Steelers nor Patriots have lost since Week 10. They ranked seventh and 11th, respectively, in weighted defensive DVOA at the end of the season. That's up to fourth and seventh if we include the postseason. The big question is: who the heck have these defenses beaten in that time?
We know the Patriots faced the easiest schedule in the NFL this season, but the defense especially had a soft ride to this title game. Four of New England's last seven games have been against the three worst offenses in the NFL by DVOA (Texans, Rams, and the Jets twice). The Broncos, ranked 28th, were also part of that run.
The highest-rated quarterback the Patriots faced this season was Andy Dalton, who finished 13th in DVOA. Out of the 108 teams to reach the Conference Championship Game since 1990, the 2016 Patriots faced the sixth-easiest slate of quarterbacks based on average DVOA, and were the only defense to face quarterbacks with negative total DYAR. The 2016 Patriots and the 2002 Buccaneers are the only teams to not face a single quarterback who finished in the top 12 in DVOA that regular season. We can argue that Russell Wilson, albeit in a down year with some injuries, was the best real-life quarterback the Patriots have faced, and he led Seattle to 31 points in a road win in Week 10. This defense, which also has had the best starting field position thanks to some stellar special teams and the fewest giveaways (11) in the league, has simply not been challenged this season. The Patriots even drew the worst offense to make the playoffs (Houston) last week.
As for Pittsburgh, we can say a lot of the same things. Sure, the Steelers have played three of the top five quarterbacks in DVOA, but those games against Brady and Dallas' Dak Prescott did not go well -- and Pittsburgh was at home for both. Out of the 28 combined wins by Pittsburgh and New England this season, the highest-rated quarterback either team beat was Kirk Cousins (fifth), who lost to Pittsburgh back in Week 1. Also, similar to how the Patriots got Landry Jones in the only game Roethlisberger missed due to injury this year, the Steelers got Scott Tolzien on Thanksgiving in the only game Andrew Luck missed. During this nine-game winning streak, the Steelers have beaten offenses ranked 10th (Bills), 11th (Bengals), 12th (Colts), 13th (Chiefs), and 14th (Dolphins) in DVOA, but without players such as Luck and A.J. Green active, is that really that impressive?
No matter which team wins, it will have to take down an excellent quarterback and offense in the Super Bowl in two weeks. As for Sunday evening, this is the toughest test either defense has faced yet this season.
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. Please remember that all stats represent regular season only, except for weighted DVOA and anything else specifically noted.
This week, we're breaking out separate charts for offensive and defensive DVOA each week. The defensive charts are "reversed" so that better games are still higher on the chart even though they have negative DVOA. The games with white dots on the charts were started by backup quarterbacks.
Pittsburgh at New England
All readers can click here for in-game discussion on our message boards. If you have FO Premium, you can click here to see all the matchup of DVOA splits for this game.
WHEN THE PATRIOTS HAVE THE BALL
I am not sure why Pittsburgh head coach Mike Tomlin was mad that the Patriots had an extra day and a half to prepare for this game. Doesn't New England just dust off the 2002 season opener game plan for every Pittsburgh matchup anyway? Hasn't Tomlin just followed Dick LeBeau's "3-4 for Dummies" script for his defense? I wish this was more of a joke, but too many of the Steelers-Patriots matchups have looked all too familiar over the years. Brady and Belichick are 9-2 against the Steelers, and since Tomlin became Pittsburgh's head coach in 2007, Brady is 5-1 with 19 touchdowns to zero interceptions.
New England's lack of an emotional attachment to establishing the run helps to expose the weaknesses in Pittsburgh's scheme. The pass rush from the outside linebackers is rarely able to get to Brady, and he releases the ball quickly in a spread attack that takes advantage of the cushions provided by Pittsburgh's usually beleaguered secondary. While working on FOA 2016, I spent time in July watching these matchups over, recalling the way the Steelers would also suffer mental lapses in leaving receivers, including Gronkowski, wide open at times, or just falling victim to the same short pass over and over. Pittsburgh is starting two rookies in the secondary in Artie Burns and Sean Davis, and the youngsters cannot get caught peeking in the backfield this week. New England almost always has one well-timed bomb ready for the Steelers, or two if you recall Deion Branch's performance in the aforementioned 2004 AFC Championship Game.
2007: Brady to Welker on 6 consecutive plays (last 5 are all same drive) pic.twitter.com/9SU3KIuzcE
— Scott Kacsmar (@FO_ScottKacsmar) July 8, 2016
I'd like to say this lack of attention to Gronk was due to him being a rookie, but stay tuned. pic.twitter.com/qJKSXvlVce
— Scott Kacsmar (@FO_ScottKacsmar) July 8, 2016
Not sure what you'd call this Gronk strategy. Another NE TD. pic.twitter.com/TxMpGUbOvY
— Scott Kacsmar (@FO_ScottKacsmar) July 8, 2016
That is not to say that Tomlin has failed to learn anything over the years. Pittsburgh went blitz-crazy in New England in 2007, sending at least five pass-rushers after Brady on more than 70 percent of the plays, but Brady shredded them for 399 yards and four touchdowns -- one of his best games in his best season. Pittsburgh has backed off more in the other matchups, and in Week 7 this season, the Steelers blitzed just one time in the whole game. Brady still managed to find Gronkowski for 37 yards, New England's biggest gain in the game.
However, Pittsburgh has brought more pressure on defense this season since the New England loss. There was a time early in the season when Pittsburgh's only sack came after Andy Dalton held the ball for seven seconds. Since Week 8, Pittsburgh has seen its overall pressure rate nearly double, while becoming one of the more blitz-happy teams with the fourth-highest pressure rate with five-man rushes, according to data from SIS charting.
|Pittsburgh Defense: Increased Pressure Since Patriots Loss|
|Split||Pres%||Rk||Blitz%||Rk||Pres%, Rush 5||Rk|
Just as New England will likely not attack Roethlisberger, the Steelers should cool off on the blitzes this week, though pressure is still critical to beat Brady. He was off last week against Houston, throwing two interceptions and completing less than 50 percent of his passes, after the Texans hit him often. While James Harrison has had another awesome postseason at age 38, he has one career sack in seven games against Brady. The Steelers might need to get creative and try to pressure Brady from the interior instead of the edges. Credit to Twitter user @mchorowitz (and Jeff Howe at the Boston Herald) for inquiring about Brady's recent tendency to throw deep when flustered. Since 2015, Brady's average pass has traveled 10.5 yards down the field while pressured, just a tick away from Tyrod Taylor's league-leading average under pressure. In previous seasons, Brady was closer to the league average, while his counterpart Roethlisberger was actually the league's mad bomber while pressured from 2011 to 2014. Pittsburgh's four hits and zero sacks in Week 7 won't be nearly enough on Sunday to disrupt Brady.
Brady typically averages around 8.0 air yards per throw against Pittsburgh, which is about the league average in this era. In 2016, the Steelers have faced an average throw of just under 7.0 yards beyond the line of scrimmage, and Brady's 6.8 air yards per attempt was his shortest game yet against Tomlin's Steelers. The Patriots dropped a couple of passes, and Chris Hogan lost an early fumble that seemed to knock him off his game that day. Hogan can be a solid deep threat, and the Steelers were only 18th against deep passes (thrown further than 15 yards beyond the line of scrimmage) this season.
Pittsburgh cannot afford to give up big pass plays to an offense that failed to make any with players not named Gronkowski in Week 7. The 2011 matchup, won 25-17 by the Steelers in Pittsburgh, was easily Tomlin's best approach for New England. The defense got physical and pressed the receivers more, as Brady's longest completion was 23 yards to none other than Gronkowski. Are you starting to see a pattern here? Meanwhile, Roethlisberger controlled the clock with short passes of his own and the Steelers had more than 39 minutes in time of possession. Brady's QBR in the 2011 game was 64.9, by far his lowest against Tomlin's Steelers. It has been 89.2 or higher in his other five games against Tomlin's defenses, and it was 93.0 in Week 7.
If there is hope here for Tomlin's defense, it is that this is the first time Brady will face him without an all-world talent such as Randy Moss or Gronkowski in his arsenal. Martellus Bennett is a very good tight end, who might be called on to perform like Gronkowski this week, but he also has been banged up and has eight games this season where he failed to crack 25 receiving yards. Julian Edelman could exploit a defense that ranks 32nd against No. 1 wideouts. Edelman has nine straight games with at least 73 receiving yards after hitting that mark once in the season's first eight games. The Steelers bottled him up pretty well in Week 7, holding him to 60 yards on nine grabs.
[ad placeholder 3]
Early downs should be interesting in this matchup. Pittsburgh was No. 1 in DVOA on first down this season, but the New England offense was No. 2. If the Steelers can force Brady into third-and-long situations, then they will have to defend them much better than they have this season. The Patriots rank second on third-and-long, while the Pittsburgh defense is just 26th in DVOA.
New England's offense ranked No. 1 in DVOA (45.7 percent) from the shotgun this season, though only used it 53.3 percent of the time, 27th in 2016. In Week 7, the Patriots were in shotgun on 51.8 percent of plays, including both of Brady's touchdown passes. The Patriots average 2.0 more yards per play in shotgun compared to non-shotgun plays, the largest differential in the league, so one could argue that this is a good week to go shotgun and spread out the Steelers. Of course, the Patriots can be very unpredictable in their approach.
For all the attention on Brady against the pass defense, this could be a big opportunity for LeGarrette Blount, who had a season-high 127 rushing yards and two touchdowns against the Steelers in Week 7. That was the worst DVOA for Pittsburgh's run defense (28.6 percent) all season. At the time, the loss of Cameron Heyward was thought to be the main reason for that poor performance, and the Steelers have gotten better without their stout lineman since. Some believe the lack of Blount touches against Houston was due to an illness. On the other hand, Dion Lewis and James White are not bad alternatives, and could take advantage of the Steelers in the receiving game. White had a receiving touchdown in Week 7, and Lewis had three scores in three ways (run, catch, and kick return) a week ago against Houston.
WHEN THE STEELERS HAVE THE BALL
It is getting hard to find new things to say about the Killer Bs. At least they will all be on the field together against the Patriots for the first time since 2013, which was back when Le'Veon Bell was a rookie rather than the most exciting back in the NFL to watch. The bottom line is that Bell has to be great again, Roethlisberger has to play better, and Antonio Brown can probably get consistently open against any cornerback the Patriots put on him.
So what does New England focus on this week? Belichick is famous for being able to take away what an offense does best. That just might be the run for Pittsburgh right now, but when it comes to throwing the ball, we know Brown carries this receiving corps. Including the playoffs, Brown has 876 more receiving yards than Pittsburgh's next closest receiver (Eli Rogers at 640 yards). Brown will likely match up again with Malcolm Butler, who finished 20th in coverage success rate this season according to SIS charting (subscription required). In Week 7, Brown caught 5-of-9 targets for 94 yards against Butler, who allowed a 51-yard gain but also had an interception in the end zone. Brown also had 133 yards and a touchdown in the 2015 season opener in New England with Roethlisberger at quarterback.
In the regular season, the Patriots ranked 20th in DVOA against No. 1 wide receivers, so this is not likely a game where Brown will be shut down. It is going to be important for Pittsburgh to find contributions from other receivers. Rogers has really earned a lot of Roethlisberger's trust out of the slot, but tight end Ladarius Green would be a nice boost if he is able to return from a concussion. Jesse James did very well in Green's place last week, but the Steelers lack that reliable tight end target this season since Heath Miller retired. The Patriots only rank 20th against receiving backs, but Bell has just seven catches for 18 yards in his last three games. This might be a matchup where the Steelers should use Bell more as a wideout, with DeAngelo Williams in the backfield. Williams was excellent in Bell's place in Week 1 last year in New England with 127 rushing yards. However, Pittsburgh has not shown any interest in this in recent weeks. In Kansas City, Bell played all but three snaps, which was the total number of snaps for Williams.
[ad placeholder 4]
Bell has received at least 20 carries in each game he has played during Pittsburgh's winning streak. It is hard to believe he was held to 3.52 yards per carry and 58 rushing yards per game from Week 5 to Week 10 (including 21 carries for 81 yards against the Patriots), but Pittsburgh's offensive line has really stepped up as the team has transitioned to a run-heavy approach down the stretch. Perhaps this has hurt the effectiveness of the passing game to a degree, but it is hard to argue with the results Bell has provided. Since Week 11, Bell has averaged 5.33 yards per carry and 146.5 rushing yards per game. Only Buffalo, the No. 1 rush offense in DVOA, was able to post such rushing numbers against the Patriots this year, but Pittsburgh should be very committed to doing so.
On paper, this seems like a juicy matchup for Roethlisberger, going against the No. 4 run defense, but No. 23 pass defense. We mentioned the lack of quality quarterbacks the Patriots have faced this year, but Roethlisberger is not coming into this game at his sharpest. This is why Belichick should focus on run-blitzing the Steelers to turn those slow-developing runs by Bell into big stuffs. Surprisingly, Pittsburgh had the fourth-lowest stuffed run rate this year despite all of the times that Bell literally stopped cold in the backfield before picking his spot for another good gain. The Patriots were only 21st at stuffing runs, but did finish first in open-field yards.
Pittsburgh's offense has almost gotten predictable. On the season, no team used play-action passing less often than Pittsburgh (14 percent), which uses it primarily for the big plays. When Roethlisberger used play-action, he averaged 13.5 air yards per attempt (second highest in 2016). With Roethlisberger still barely using play-action this postseason, New England really should consider loading up against the run to deal with Bell.
"Force Roethlisberger to beat you" might sound crazy, but this is what happens when a great offensive line is paving the way for a determined Bell. Also, credit to ESPN Pittsburgh's David Todd for pointing out that Roethlisberger has not thrown for 300 yards in his last eight games, a streak he last matched in 2007. Roethlisberger has still been given very good protection, and he must take advantage of that against a New England pass rush that was 25th in pressure rate this season, according to SIS charting (subscription required). Kansas City ranked 26th, and the Chiefs were really unable to penetrate Pittsburgh's line too. Interestingly enough, the Patriots (25.5 percent) and Chiefs (19.9 percent) led all defenses in percentage of passes where they rushed fewer than four defenders, so they may have voluntarily sacrificed some pressure for more coverage this season. Still, a clean pocket for Roethlisberger is not good for New England, and the Steelers kept Landry Jones very clean in the Week 7 matchup. Pittsburgh just hopes that Roethlisberger provides the needed boost over Jones, who had the Steelers in a competitive game, but faltered down the stretch.
The red zone will get plenty of focus this week after Pittsburgh failed to score a touchdown in Kansas City. The Steelers had three plays inside the 5-yard line, but they were all Roethlisberger passes, including a tipped interception. We may see more of that this week too. While the Pittsburgh offense is No. 2 in red zone rushing, the Patriots are also No. 2 against the run on defense. In fact, New England last allowed a rushing touchdown in Week 8, but that stat is a little misleading. The Patriots have faced just four rushes from inside the 5-yard line in that time, the second fewest in the NFL, and backs like rookie C.J. Prosise and Benny Cunningham aren't exactly Le'Veon Bell. However, the Patriots only rank 24th against red zone passes, so Roethlisberger may need to find his groove again.
In fact, he better, because the recent red zone results have been terrible. Since Pittsburgh's winning streak started in Week 11, Roethlisberger's red zone stats pale in comparison to those of the other three remaining quarterbacks:
|Red Zone Passing Stats Since Week 11|
Yikes. Roethlisberger's numbers really pale in comparison to every NFL quarterback's in that time, but especially to those of these three. Roethlisberger was a liability in his last playoff start against the Patriots because he was a clueless rookie in over his head. This time, he is a 13-year veteran with one of the best supporting casts he has ever had on offense, and one of his best chances yet to get a road win over Brady and Belichick. While Bell is going to get his touches, this game is still largely on Roethlisberger to play well and expose a New England defense that has not played a Hall of Fame passer this season.
In his 20th playoff start, Roethlisberger needs to remind us why he is in fact one of those future Hall of Fame quarterbacks.
New England had a huge special teams advantage over Houston last week, and it mostly showed in a 98-yard kick return touchdown by Dion Lewis, though he did also fumble deep in his own end. Lewis has six kick returns since 2012 to his name, so this is not really something he is adept at doing. Both teams like to use big-name players as returners, but neither has been that dangerous on returns this season. Pittsburgh just did very well against the No. 2 return unit in the league in Kansas City last week. The Steelers bottled up All-Pro returner Tyreek Hill very well, holding him to 72 yards on four kick returns. Pittsburgh only punted once last week, which was important since Jordan Berry has not been strong in that area this season.
Both kickers are trustworthy. Chris Boswell just set a playoff record with six field goals in a game last week, though Pittsburgh better hope he isn't that busy on Sunday again. Stephen Gostkowski had an infamous miss of an extra point in last year's AFC Championship Game loss in Denver. He has missed three more extra points this season, as well as two field goals from 30-to-39 yards away. The weather should be great on Sunday night, with temperatures in the 30s and no precipitation.
The Patriots were 7.5-point favorites in Pittsburgh against Landry Jones earlier this season, and covered 27-16. With all of the changes for Sunday's matchup, the Patriots are still a reasonable 6-point favorite at home. If Brady has his usual Pittsburgh performance, then that should be enough for a win. If Brady has his usual AFC Championship Game performance -- in this round, he has thrown 12 touchdowns to 12 interceptions with just 6.55 yards per pass attempt -- then the Steelers have a great opportunity for an upset. As much as the offensive production and the four quarterbacks are being highlighted this week, this is a game that is likely to be decided by pass protection, run blocking, and ball security rather than in an aerial shootout.
Since 2001, the Patriots are 12-0 in the playoffs against new opponents, but only 11-9 in rematches from the regular season. No matter how much time Tomlin received to prepare for this game, you just get the feeling that he is not ready to outcoach Belichick on this stage. This is the right situation if it was to happen, but between New England's long-running mastery of the weaknesses of this defense and the road sloppiness of Pittsburgh's offense, the Patriots are the safe pick at home to become the first team in NFL history to play in nine Super Bowls.
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 red zone DVOA and 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.