AFC Semifinals Preview

by Aaron Schatz, with additional analysis by Michael David Smith


We appreciate those of you who link to our previews at team message boards and help us spread word about Football Outsiders. However, I've noticed some message boards where people have cut and pasted entire articles, sometimes game previews and sometimes Mike's Every Play Counts articles. Please only copy a few quotes and then link to our articles instead. Remember, your web clicks help feed my family. Meanwhile, if you are here for the first time to read this preview, or perhaps discovered our site recently through another of our articles, please do not hesitate to email me at Contact Us with questions about our statistical methods. I'll be running another reader mailbag during the two weeks between the conference championships and the Super Bowl, and questions asked in the discussion thread may get lost in the shuffle.


It is often said that home field advantage is key in the playoffs, but that's not quite accurate. The real advantage comes from a first-round bye. In the wild card round and the championship games, home teams win at the same rate as the regular season. But in the second week of the playoffs, home teams are 45-11 since the postseason expanded to 12 teams in 1990. Just like in the wild card round, each playoff game this week is a rematch of a regular season game. All four of those games were won by the teams hosting this week's contests.

Our ratings say there has been a huge gap between the top three AFC teams and the rest of the NFL this season. But only two of those AFC teams can make it to next week's AFC championship -- and if the Jets can put together an upset, only one of the three teams will actually make it. Both AFC second round games are previewed below using a combination of our innovative Football Outsiders statistics and closer tape analysis by Michael David Smith.

For those who may be visiting this site for the first time to read this preview, some explanations for our statistics. 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 VOA indicates a better defense and worse offense, and a positive DVOA indicates a better offense and worse defense.

Each team is listed with DVOA for offense and defense, split into rush and pass, along with rank among the 32 teams in parentheses. (If the DVOA values are difficult to understand, it is easy to just look at the ranks.) I've also listed each team's red zone performance on offense and defense.

TREND is the WEIGHTED DVOA trend, 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). Weighted DVOA now includes playoff performance, and the weights for all teams were moved an additional week ahead whether they played last week or not. Except for TREND numbers, all DVOA statistics on this page are regular-season only unless otherwise noted.

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.

Each team also gets a chart showing their performance this year, game-by-game, according to total DVOA. Since every game this weekend is a rematch, the weeks where each team played the other one this season appears in the other team's color. These charts include the wild card game for the Jets and Colts. In addition to a line showing each game, another line shows the team's trend for the season, using a third-power polynomial trendline. That's fancy talk for "the curve shifts direction once or twice." Note that even though the chart appears in the section for when each team has the ball, it represents total performance, not just offense.

Rather than a separate open game discussion thread like we do during the season, this thread will also be the place to discuss the games as they happen on Saturday and Sunday. Please limit discussion here to aspects of the Patriots-Colts matchup that ARE NOT direct comparisons of Peyton Manning and Tom Brady. Those comparisons must all go in the OFFICIAL THREAD FOR IRRATIONAL BRADY-MANNING ARGUMENTS so that the rest of us can keep our sanity.

NFC previews will be posted later today. Last week's AFC preview and game discussion thread is here.

New York Jets at Pittsburgh

Jets on Offense
DVOA 23.4% (5) -15.0% (4)
TREND* 18.4% (6) -16.8% (3)
PASS 30.9% (6) -19.6% (3)
RUSH 17.3% (2) -9.2% (6)
RED ZONE 32.9% (2) 2.9% (20)

Steelers on Offense
DVOA 16.6% (7) 2.8% (19)
TREND* 20.9% (5) -0.2% (18)
PASS 28.6% (7) 15.1% (24)
RUSH 9.4% (7) -10.4% (5)
RED ZONE 18.9% (6) 2.9% (21)

Special Teams
DVOA 3.1% (11) 3.8% (7)
NYJ kickoff 1.8 (11) 2.3 (13)
PIT kickoff 1.8 (14) -1.2 (14)
NYJ punts 16.9 (5) -5.8 (18)
PIT punts -6.8 (21) 19.5 (1)
FG/XP 3.7 (13) 6.4 (8)

The Jets are 10-6, the Steelers 15-1, but a closer look at each team's performance in context shows two teams that are virtually identical. Both teams prefer the ground game, ranking first and second in percentage of plays where a running back carried the ball. Both have success passing when they need to, with quarterbacks that throw few interceptions. Both score consistently by limiting red zone turnovers. Both excel at stopping opposing runners. Even their special teams are similar.

The one glaring difference is that Pittsburgh is far better when it comes to stopping the pass, and that -- along with the bye week advantage -- makes them a heavy favorite.


Chad Pennington's struggle since returning from a shoulder injury was the biggest reason why most observers -- including this one -- felt the Jets would likely lose in San Diego. But in beating the Chargers, Pennington threw with more authority than any time since injuring his rotator cuff two months ago. A 47-yard touchdown pass to Santana Moss showed that Pennington's shoulder can handle a long throw better than anyone thought; although it doesn't show up in the official records, a 47-yard incompletion showed that Pennington isn't limited to one such throw per game.

In the first game between these teams, won by Pittsburgh 17-6 in Week 14, Pennington completed only 17 of 31 pass attempts and threw three interceptions -- one-third of his total for the entire season. Now Pennington faces the question: Did his season-worst outing come because he could not solve the Pittsburgh defense, or because it was only his second game back after the injury?

Pennington was certainly not the only quarterback to have a bad day against the Steelers defense, which allowed the fewest points in the league and ranked fourth in yards allowed per pass. Three Steelers made the Pro Bowl, linebackers James Farrior and Joey Porter and safety Troy Polamalu. Left end Aaron Smith should have made it as well; he had eight sacks, unheard of for a lineman in a 3-4 defense, and poses a tough test for Brandon Moore and Kareem McKenzie on the right side of the Jets line.

While Pennington will need a good game for the Jets to win, the ground attack is still New York's bread and butter. But Pittsburgh led the league in fewest yards allowed by opposing running backs this year, 1074, so stopping the run is their bread and butter. In the first game, the Steelers won this Butter Battle, limiting Curtis Martin to 72 yards on 24 carries.

However, the Steelers weren't able to stop the Jets where they are strongest: converting on third-and-short. The Jets were the best team in the NFL this season in what we call "power situations," converting 87% of runs on third or fourth down with one or two yards to go. In the first game they ran four times on third-and-one and were successful on all four.


The Steelers have built their offense around the run to an even greater extent than the Jets. 59% of Pittsburgh's offensive plays were runs, the highest percentage in the league by far. Pittsburgh's preference for the run was greatest on first down; the Steelers ran on two of every three first down plays.

The Steelers like to run up the middle behind center Jeff Hartings and guard Alan Faneca, with fullback Dan Kreider often leading the way. That will make this a big test for Jets middle linebacker Jonathan Vilma, Defensive Rookie of the Year, who had another impressive game against the Chargers. Vilma has shown that he has speed and the ability to inflict big hits, but against the Steelers he has to shed blockers when opposing offenses run right at him. He didn't do so well in Week 14, as the Steelers ran up the middle 16 times for 90 yards.

The Jets actually did an outstanding job of shutting down Vilma's counterpart, Offensive Rookie of the Year Ben Roethlisberger. The Steelers quarterback completed only 9 of 19 pass attempts and threw two interceptions against the Jets (one was a Hail Mary attempt at the end of the first half). But despite Gang Green's strong defensive reputation, this kind of dominance against an opposing quarterback was extremely abnormal.

Net Yards Allowed per Pass
  NYJ Defense PIT Defense
1st Down 7.88 6.49
2nd Down 5.55 5.16
3rd Down 4.77 4.32

The weakness of New York's pass defense was masked this season by a number of factors. (This was discussed in more detail in last week's preview.) The Jets play a slower-paced game than other teams, leading to fewer plays per game and thus fewer yards on both offense and defense. They played a schedule filled with poor passing teams like Miami, San Francisco, and Arizona. And the strength of their run defense lifted their placement in the official NFL defensive rankings enough that nobody noticed that the Jets allowed 6.2 net yards per pass attempt, 18th in the league. The table to the left shows how, even based on conventional statistics rather then our complicated DVOA metric, the Jets were significantly inferior to the Steelers when it came to preventing the pass.

So despite Roethlisberger's struggles last time these two teams faced each other, and as much as they prefer to run, the Steelers would be smart to take to the air. This is especially true on first down, when the Jets will be expecting run and susceptible to a good play fake. New York was above average against passes on second and third down, but one of the league's five worst defenses against passes on first down. And while the Steelers threw fewer first down passes than any other team in the league, they were actually one of the league's three most successful offenses when they did throw on first down.

After Eric Barton's brush with goatness, it is also worth noting that penalties made a big difference in the regular season meeting of these teams. The Jets were called for 12 penalties for 84 yards, including three false starts and two delays of game, while the Steelers were called for only two penalties for 15 yards.


Both the Jets and Steelers are average on field goals, kickoffs, and kick returns. Both were below average on punt returns but strong on punts, although for different reasons: Chris Gardocki of the Steelers punts for a better distance, while the Jets have better punt coverage.


The Jets win over San Diego was an upset because the Jets seemed to have lost momentum at the end of the season, but a close game was no surprise because the two teams had played at such a similar level for the season as a whole. That's not the case with the Steelers. Their focus on the run makes them similar to the Jets, but their dramatically better pass defense makes them superior. Last week's win was a small upset for the Jets; a win this week would be a much larger one.

Indianapolis at New England

Colts on Offense
DVOA 38.9% (1) -9.1% (6)
TREND* 44.7% (1) -7.2% (10)
PASS 63.1% (1) -6.5% (11)
RUSH 9.4% (8) -12.3% (4)
RED ZONE 13.5% (10) -39.7% (2)

Patriots on Offense
DVOA 26.3% (4) 2.3% (18)
TREND* 23.5% (4) -2.8% (15)
PASS 42.1% (2) 1.4% (18)
RUSH 12.2% (5) 3.5% (21)
RED ZONE 16.7% (7) -20.1% (8)

Special Teams
DVOA -1.8% (22) 0.2% (16)
IND kickoff -13.3 (29) 7.2 (8)
NE kickoff 5.8 (11) -4.8 (24)
IND punts 5.6 (15) -16.2 (29)
NE punts -4.5 (15) -0.5 (23)
FG/XP -3.6 (26) 15.2 (1)

This is the most anticipated game of the weekend, matching a record-setting quarterback who calls his own plays against his arch-nemesis, a cerebral coach whose voter registration probably reads "Genius William Belichick." But do not ignore the Patriots offense and Colts defense, two underrated units that also get a say in deciding which team emerges triumphant.

(When looking at the game-by-game trend charts, remember that the Colts and Patriots only played their starters for part of their meaningless Week 17 games. Also, if you are not a regular reader of our website, I must in the spirit of full disclosure make clear before this preview that I am a Patriots fan.) 


Conventional wisdom states that injuries in New England's secondary give Peyton Manning the chance to finally overcome his dismal history against New England. But Manning's history against the Patriots is not quite so dismal. While Manning was horrid in last year's AFC Championship Game, he's thrown for 250 yards in each of the last two regular season meetings just to have last-minute drives ruined by a stuffed Edgerrin James plunge on fourth-and-goal and a miss by kicker Mike Vanderjagt.

Manning's success comes in part from his ability to read defenses and adjust the offense at the line of scrimmage. With so many weapons, and his renowned ability to sell the play fake, Manning can take advantage of any hole the defense gives him.

But the Patriots reduce Manning's effectiveness from otherworldly to merely above average by confusing him with shifting defensive schemes, often getting pressure with four or even three pass rushers. (Note that Manning was sacked nine times in six games against 3-4 defenses this year, but only four times in nine games against 4-3 defenses.) That's why the most important injured Patriot is not cornerback Ty Law but Pro Bowl defensive lineman Richard Seymour, listed as questionable.

Right tackle Ryan Diem, whose mistake allowed Denver's Reggie Hayward to sack Manning on Sunday, is the weakness in the Indianapolis pass protection. He'll have to face linebacker Willie McGinest, who has a history of huge plays against the Colts: last year he kept James from the game-winning score in the final seconds and this year his sack of Manning led to Vanderjagt's long-range miss.

What about those injuries in the secondary? Against the Broncos, Manning showed his ability to find the weak link in a defense and exploit it over and over. America had not seen anyone get picked on the way Manning picked on Roc Alexander since Freaks and Geeks got cancelled. Will Manning choose a Patriots second-stringer to be the recipient of a similar barrage of aerial swirlies?

New England Defense
DVOA vs. Receivers by Type
  Weeks 1-7 Rank Weeks 8-17 Rank
#1 WR -54.6% 1 -46.1% 1
#2 WR 40.2% 28 42.8% 30
Other WR -21.2% 13 28.1% 24
TE -20.9% 4 -38.9% 6
RB -25.8% 12 32.3% 32

The fascinating thing about New England's secondary is that, according to DVOA, the Patriots actually had the best defense in the league against number one receivers both before and after Ty Law's injury. They intercepted a league-high 11 passes intended for opposing number one receivers. But they had one of the worst defenses in the league against number two receivers, allowing big games to guys like Antonio Bryant, Johnnie Morton, and Marty Booker. And when injuries turned nickel backs into starters, the Patriots began to give up huge games to third receivers as well. Brandon Stokley could have a game akin to the one Ricky Proehl had against the Patriots in last year's Super Bowl. Stokley actually gets thrown more passes on third down than either Harrison or Wayne, and he caught a higher percentage of third down passes as well.

The numbers also say that the Patriots do a great job against opposing tight ends -- although Dallas Clark seems to be split wide more often than a typical tight end -- and that since Law's injury they've been the worst team in the league preventing success on passes to running backs. I have no idea where the hole in the defense is for that one; part of the reason for the bad DVOA number is that when opposing quarterbacks pass to their running backs, the passes are hardly ever incomplete. It's not as big a deal as you might think, because the problem seems to go away on third down when it would really hurt the Patriots by creating easy first down conversions.

Speaking of the running backs, the Colts do have one and he was pretty good this year. Edgerrin James has become the NFL's most consistent runner. He had the highest Running Back Success Rate of any back with at least 120 carries this season and he rarely gets stuffed. Of course, he also rarely breaks a long run, and New England was the best team in the league at preventing carries over 10 yards for the second straight season. Because they didn't give up big plays on the ground, and they were very good preventing conversions on third-and-short, the Patriots rate as a top five rush defense even though only Minnesota was worse when it came to stuffing opposing running backs at the line of scrimmage.


Indianapolis has by far the best offense in the NFL, but while everybody was busy paying attention to Peyton Manning's touchdown record, the Patriots developed into a top five offense as well. Since the Patriots first won the Super Bowl in 2001, Tom Brady has gradually evolved from "the quarterback who wins without great numbers" to "the quarterback who wins with great numbers." Our unadjusted ratings favor the Chiefs and Vikings, but when you consider strength of schedule -- remember, the AFC East played the AFC North, and both divisions are filled with strong defenses -- New England was the second-most efficient passing offense of 2004, right behind the Colts.

The Patriots aren't just a better offense now compared with their Super Bowl seasons, they are a more balanced offense with the addition of Corey Dillon. According to our numbers, only Curtis Martin was a more valuable running back than Dillon, and Dillon's hidden talent at picking up the blitz will be an important part of neutralizing the strength of the Colts defense: the pass rush.

The Colts were tied for second in the league in sacks, but the Patriots were fifth in fewest sacks allowed. An injury to Colts pass rush specialist Robert Mathis means the Patriots can pretty much control the pressure by doubling NFL sack leader Dwight Freeney with Dillon or tight end Daniel Graham. The downside of this strategy is that the weakness of the Colts pass defense comes against running backs and tight ends, and you can't catch a pass when you're busy blocking. (Also, the Colts were using a lot of stunts to try to get inside pressure against Denver, and their sacks of Jake Plummer came from inside rushers. But I'm guessing Charlie Weis has probably seen this on the game film and prepared his offensive line for the possibility.)

While Brady, like Manning, likes to spread the ball around with multiple receivers, the cold weather and need to control the clock will mean a lot of carries for Dillon. And although the Colts pass defense improved over the course of the season, the Colts rush defense declined.

But the Colts run defense does have one strength this year, and it directly counters a favored Patriots strategy. The Patriots go for it on fourth down more often than most other teams, but the Colts were fourth in the league in preventing conversions in power situations: runs on third or fourth down with two or fewer yards to go. The league average is 67% success, the Patriots had 76% success, but Colts opponents had only 54% success.

This is a big reason why DVOA says the Colts had the league's best overall defense on third downs. Of course, since they give up plenty of yards on first and second downs, their overall rating comes out as average.


In a close game, this is a big advantage for New England. Adam Vinatieri is considered the NFL's pre-eminent cold weather kicker. He also led the NFL this year in field goal percentage and didn't miss a single kick under 47 yards. The talkative Mike Vanderjagt had an off-year, missing five field goals and even an extra point. Indianapolis does have an advantage when they have to punt, as Hunter Smith was second in the NFL in gross punting average and the Patriots are poor on punt returns. Strange statistic: Only five punts against Indianapolis were downed by the punting team, fewest in the league. New England had the second-fewest, six.

Watch kickoff returns, as both teams are strong on returns and weak on kickoffs. Bethel Johnson, in particular, has given the Colts fits; he had two kick returns over 30 yards in the season opener and returned a kickoff for a touchdown in last year's regular season meeting.


Not since Charlton Heston has anyone had this many monkeys riding his back. Manning has to prove that he can beat the Patriots, that he can win in the cold, and that he can win big games in the playoffs. The injuries to the New England defense make this the best opportunity he may ever get to prove all three, but an improved Patriots offense will score plenty of points as well. This game will resemble the last two regular season games between these teams more than last year's AFC Championship: not a rout, but a close game fought until the final seconds, possibly decided by the random bounce of a fumble or a slip on the cold, muddy turf.

Nine of the past 15 Super Bowl titles have been won by the team with the highest number of projected wins based on points scored and allowed during the regular season. 13 of the past 15 Super Bowl titles have been won by one of the top two teams in projected wins based on points scored and allowed during the regular season. New England had the highest number of projected wins this season. Indianapolis, by a very tiny margin over Philadelphia and Pittsburgh, was second.

*TREND numbers are updated through wild card weekend, all other DVOA statistics are regular-season only unless otherwise noted.


There are no comments yet.