Writers of Pro Football Prospectus 2008

Most Recent FO Features

WareDeM05-1.jpg

» Week 7 DVOA Ratings

Denver: great team, or the greatest team? Would you be satisfied with "one of the ten greatest teams?" Plus: hard times in the NFC South, where defense goes to die.

15 Jan 2004

AFC Championship Preview

by Aaron Schatz

(Before we get started here, let's get this out of the way: I am a die-hard Patriots fan so while I've tried to be even handed here, it shows through.  This isn't like FOX News or the New York Times -- I'm letting you Colts fans know up front where my bias lies so you don't think I have some hidden agenda against your team.  On with the article...)

Last year, going into the Super Bowl, the Oakland Raiders offense was clicking on all cylinders.  Rich Gannon had thrown for a combined 569 yards in two easy wins over the Jets and Titans, completing 49 of 71 passes for five touchdowns and just one interception, plus 41 yards and a touchdown rushing.  The battle was on: the league's best defense against the league's best offense, and league's best offense was coming into the game as strong as they had ever been.

At which point they got smoked.  Gannon threw for five interceptions, and most of his 272 yards as well as his two touchdowns came in the second half with Tampa already leading 31-3.

How long is momentum?  Is momentum two games, or three games?  Three weeks ago, the Colts struggled to beat the 5-11 Houston Texans, in a game they had to win to get home field advantage.  What if momentum is four games?  Four weeks ago, the Colts were blown out of their own building by the Denver Broncos.  Clearly they've overcome this game, but they were not exactly clicking on all cylinders in Weeks 16 and 17.

As Tom Curran pointed out in Tuesday's Providence Journal-Bulletin:

In 2001, there was a quarterback for the Colts named Peyton Manning who came in to Foxboro Stadium at the start of the season with a 2-0 Colts team to face the 0-2 Patriots in the debut start of young Tom Brady.

In his first two games of 2001, Manning was 45 of 61 for 652 yards, six touchdowns and two picks and Indy had scored 87 points.

The Colts lost that day, 44-13, and Manning was humanized, going 20 for 34 for 196 yards with a touchdown and three interceptions.

So, with the media story going from "Peyton can't win the big one" to "Peyton is unstoppable" fast enough to give whiplash to even the most neckless offensive lineman, I hereby advise everyone to just calm down.  Yes, the Indianapolis Colts offense is clicking on all cylinders.  Look down at the graph below (under "When Indianapolis is on Offense") and you will see that even after penalizing the Colts for playing one of the league's worst defenses, DVOA ratings say that the Colts have played two of the best offensive games of the season over the past two weeks.  Peyton Manning may be the best player in the game right now.

If the Patriots offense plays well, and the Colts defense plays badly, it means, to quote Captain Cragen of Law and Order, "minus zero negative bupkis."  Why do I say that?  Read on.

Instead of going through an explanation of our stats for the 100th time, I think I'm just going to link to the Wild Card playoff preview, and you can read the stat explanation that starts off that column, or just go to the good ol' methods column that explains most everything.  All the line yards stats are explained here.  Some notes:

  • I've separated those fun DVOA week-by-week graphs into offense and defense, and I've added the playoffs to the graphs (so you can see what a hot streak the Indianapolis offense is on right now).  Remember, since a positive DVOA means more scoring, and defense is better when it is negative, improving defense has a downward slope.
  • Otherwise, the stats here are regular season only unless otherwise mentioned.
  • HFA (home field advantage) represents how much better the Patriots are at home and how much worse the Colts are on the road.
  • When one team has a significant advantage over the other team, that segment will be in bold (example: Pats prevention of 10+ yard runs).

When Indianapolis is on Offense

Colts on Offense

  IND OFF NWE DEF
DVOA 21.0% (2) -20.0% (3)
TREND 20.5% (3) -25.8% (2)
PASS 39.6% (2) -28.7% (4)
RUSH -1.1% (16) -7.7% (13)
HFA -8.5% (16) -3.8% (22)

Peyton Manning is good.  I don't think that's a surprise.  The Pats defense is good.  That's also not a surprise.  According to DVOA, the Colts had only one game where their offense was significantly below average.  The Patriots had only two games where their defense was even slightly below average.

The Pats defense had 39 turnovers, second in the league behind St. Louis.  The Colts turned the ball over only 18 times, second fewest in the league.

While the Pats sack rate per pass attempt was middle of the pack, their total number of sacks (40) was among the league leaders.  The Colts offense, with only 19 sacks allowed, was third in the league (and second in sack rate per pass attempt).

Hmmm, that New England pick in our fantasy playoff draft doesn't look so keen now.  See those two dots way up in the upper right corner of the graph here?  Yes, that's the Colts offense over the first two games of the playoffs.  Pretty sweet.

If you haven't had a chance to read Bob Cook's article on the Indianapolis off-tackle draw, do so.  This is the play that the Colts use for much of their play action, and it is that play action that makes the Colts one of the top offenses in the league.

The DVOA numbers definitely agree with the article's comment that this play works far better when the Colts are in their own part of the field.  For all their offensive dominance, the Colts have a below average offense once they get into the red zone.  The Indianapolis play-action is based on getting the receivers behind the defense.  In the red zone, there's no defense to get behind. On the other side of the field, the Patriots had 2003's second-best defense in the red zone (behind Miami), and in fact got better on defense as opposing offenses matched from one side of the field to the other.  It's that bend-but-don't-break thing that gets talked about a lot:

 


DVOA by location IND 1-20 IND 21-39 Between 40s NWE 21-39  NWE 1-20 (IND in Red Zone)
IND OFF +41% +28% +9% +41% -5%
NWE DEF -2% -9% -22% -31% -55%

I'm not quite sure why the Colts are so much better between the opponent 21-yard line and 39-yard line than they are in the middle of the field, but the decline in the red zone is real and plays into the strength of the Patriots.  One of the reasons why the Colts are so much better outside the red zone is that they complete so many more passes than the NFL outside the red zone.  Inside the red zone, though, Manning actually has a slightly lower completion rate than the NFL average.  Here is the completion rate for the Colts along with the NFL average, splitting the field into fifths (the Pats are basically the same as the NFL average, in case you are wondering):


Completion
Rate
Own 1-20 Own 21-39 Between 40s Opp. 21-39  Opp. 1-20
(Red Zone)
NFL 63% 62% 61% 58% 53%
IND 67% 74% 70% 68% 52%

When the Colts do throw in the red zone, the most common target is not Marvin Harrison but Reggie Wayne.  The red zone was the only part of the field where Manning threw to Wayne more often than he threw to Harrison, and he had better results throwing to Wayne as well.  This trend continued in the playoffs, where Wayne has three red zone receptions, as many as the rest of the Colts combined, resulting in two touchdowns and a first down.

If the Patriots can keep the Colts away from completing long bombs, we're going to see a lot of Mike Vanderjagt trying chip shot field goals on Sunday.  So the question becomes, how much of a problem is it to keep the Colts from completing long bombs.  I ran numbers on this for Carolina over the past two weeks, so I've run the numbers for the Colts and Patriots as well.  The Colts had 7.1% of their completed passes go for 30 yards or more; only Baltimore and Minnesota had more.  However, only 2.7% of completed passes allowed by the Patriots went for 30 yards or more; only Chicago allowed fewer.  So at first, it looks like the Pats should be able to counter the Indianapolis long-distance air show.

Except that, while Ty Law should be able to keep up with Marvin Harrison and Tyrone Poole should be able to keep up with Reggie Wayne, there's this other guy out there.  His name is Brandon Stokley, and as we all know, the one-time Raven has gone completely nuts over the first two weeks of the playoffs.

There are worries about how the Patriots will handle him in three-receiver sets, since rookie Asante Samuel will be the nickel back.  Well, in Week 12 Stokley wasn't healthy, so Troy Walters was the slot receiver, playing the same role Stokley will on Sunday.  And Walters caught five balls out of the eight thrown to him for 56 yards, a touchdown, and a +55% VOA.  Yikes.  That touchdown came on third-and-goal, and was the one that tied the game, and Walters also had three other first downs including one on 3rd-and-9.  Assuming Stokley can play on Sunday -- he may not, due to recent family illness -- he is going to be a major problem for the Patriots.  If he doesn't play, Patriots fans should not get excited, because Troy Walters is going to be just as big a problem. 

There's one other aspect of the play action, which is when Manning isn't actually faking it and he hands the ball off to James.  As Cook noted, the Colts are much better running to the sides than up the middle, because the fakery gets the defensive end so revved up to get the quarterback that he isn't in position to tackle James coming around him.  The Colts should especially have an advantage going left -- that's where the Colts were the second-best team in football this year -- although that's also where Willie McGinest usually plays, and as we know from a few weeks ago he seems to have the Colts' offensive line calls memorized.  Here's a closer look at the line yards stats for when Indianapolis has the ball (you'll find these stats all explained here):


Team Line Yards   Left   Middle   Right   Power   10+ Yd   Stuffed   Sacks Sack Rate  
IND OFF 3.81 9 4.49 2 3.41 22 3.79 12 54% 30 11% 30 23% 12 19 3.4% 2
NWE DEF 3.52 13 3.57 9 3.57 18 3.34 8 62% 8 6% 1 25% 14 40 5.7% 20

You can officially be shocked whenever Indianapolis can pull off a long running play.  The Pats were the best team in the league at preventing runs over 10 yards, while Indianapolis was third-worst.  This loss of explosiveness since his injury is why, even though he's steady as ever, James is no longer an elite back.  (Note: a good day running against the Kansas City Chiefs does not qualify James as "back to being one of the league's best backs.")  The Pats also have an advantage in "power" situations, third-and-short and similar plays, but of course we already know that from a few weeks ago.  That advantage is part of the reason why the Patriots defense gets better on third down, and the Colts offense gets worse -- although it should be noted that the reverse has actually been true in the Colts two playoff games:


DVOA by down 1st Down 2nd Down 3rd Down
NWE DEF -13% -19% -25% 
IND OFF (regular season) +31% +18% +3%
IND OFF (playoffs) +42% +83% +171%  (Egads!)

So we know that Asante Samuel and Eugene Wilson are important in the passing game, to cover Brandon Stokley (or Troy Walters).  In the running game, a lot depends on Ted Washington.  Washington needs to be double-teamed fairly frequently, and the more attention the Colts offensive line has to give him, the more the other New England defenders are free to either guard for James coming around on the handoff or drop into coverage to help prevent a completion on play action.

When New England is on Offense

Patriots on Offense
  NWE OFF IND DEF
DVOA -0.2% (12) 0.2% (17)
TREND 3.0% (13) 3.8% (18)
PASS 10.2% (10) -11.6% (7)
RUSH -11.2% (25) 12.0% (30)
HFA +35.0% (2) +7.3% (16)

John Clayton of ESPN.com writes in his preview of this game: "The AFC championship game comes down to two people. Colts quarterback Peyton Manning looks unstoppable. Patriots coach Bill Belichick usually finds a way to stop any individual."

I disagree.  I think this game will be decided on the other side of the ball.  Will the Patriots offense repeat the first half of their 38-31 win in Week 13, when they looked unstoppable, or the second half of that win, when they looked like a Pop Warner team and were saved by their defense?

Yes, Peyton Manning had a great comeback.  But the reason the Patriots almost lost in Week 13 was the complete turnaround of their offense and the Indianapolis defense.  Here is are the VOA ratings, by quarter, for that Week 13 game.  Without adjusting for opponent, remember, each defense has the same rating as the opposing offense:


Week 13 Q1 Q2 Q3 Q4
IND OFF/NWE DEF -118% 54% -19% 26%
NWE OFF/IND DEF 70% 60% -114% -103%

The Patriots' first drive the third quarter was 11 plays, 61 yards, and a touchdown that made it 31-10.  From then on, here are all their offensive possessions:

  • -3 yard pass, interception
  • incomplete, 3 yard pass, interception
  • 7 yard run, 2 yard run, 0 yard run, punt
  • 15 yard pass, 3 yard pass, incomplete, 13 yard pass for touchdown
  • 4 yard run, -2 yard run and lost fumble
  • 5 yard penalty on Colts, incomplete, aborted snap, incomplete, (shanked) punt

What the heck happened?  There is one good drive there, and it only looks good because Bethel Johnson's kick return started it on the Colt 31-yard line.  That is a terrible job of holding onto the ball, much less running out the clock.  Thanks to New England's offensive meltdown, Peyton Manning got to engineer a great comeback with drives starting at the New England 44 (touchdown), New England 26 (touchdown), New England 11 (field goal), and New England 48 (the final drive that stalled at the goal line).  Gee, give those starting positions to Ryan Leaf and he could tie up the score.

The Colts need to go back and look at whatever they did to take Tom Brady off his game, and they need to do it again.  The Patriots need to go back and look at whatever the Colts did to take Tom Brady off his game, and they need to prevent it.  Whichever of those two goals is reached likely determines the winner of this game.

The strategy against the Colts -- and who am I to disagree -- is to win the time of possession battle and keep Peyton Manning off the field as much as possible.  In the first Denver-Colts matchup, this strategy worked to perfection.  In the second Denver-Colts matchup, offensive line penalties led to first-and-long, which led to punts, which led to destruction.  In the first half of the first Colts-Pats game, the Pats led time of possession 16:15 to 13:45.  In the second half, the Colts led time of possession 18:15 to 11:45.  Yikes.

If I read their preview correctly, the ESPN Scouts Inc. preview (no point in linking, it costs money) is picking the Colts in this game for two main reasons.  The first is the momentum argument listed above.  The second is that the Pats have trouble running against "even the weakest run defenses", and so they cannot take advantage of the Colts poor run defense and will not be able to keep the ball out of Manning's hands.

While the lack of a strong running game is New England's obvious weakness, I think the idea that the Pats have trouble running against "even the weakest run defenses" is ridiculous and easily disproveable. Do you want your 163-yard game on the Titans? How about 240 on Buffalo over the two games, mostly in the second? 126 yards on Houston? 84 yards against the defense that rates atop the defensive line yards stats, Jacksonville?

How about a look at the line yards stats for when New England has the ball (you'll find these stats all explained here):


Team Line Yards   Left   Middle   Right   Power   10+ Yd   Stuffed   Sacks Sack Rate  
NWE OFF 3.49 23 3.37 30 3.62 14 3.29 26 63% 21 11% 29 23% 11 31 5.3% 9
IND DEF 4.03 30 4.38 31 3.84 27 4.16 30 64% 16 19% 17 22% 23 29 6.4% 12

These statistics tell you a lot about how the Patriots running game works, beyond the simple word "bad."  They rarely break a run over 10 yards.  They're better than most teams at not losing yardage on running plays.  They're horrible to the sides, where most of the long runs are, but reasonable when running up the middle.  All these statements point to a running game that, while it will not break the game open with a big gain, should be able to chew clock time if necessary (and if Charlie Weis isn't feeling like he has something to prove).

Two changes in New England personnel will make things interesting.  First, Antowain Smith wasn't even active for the first Pats-Colts game.  He's nobody's idea of a great running back, but his straight ahead style fits the chew the clock strategy as well as the New England blocking strength up the middle.  However, one of those great blockers, guard Damian Woody, is not going to play.  Russ Hochstein, this is your moment to shine!

Scouts Inc. also seems to forget the main idea of the Pats offense: use of the short passing game as a running game.  That's something that has worked well all year; whether it works well this Sunday is another issue.  The Colts were one of the better teams in the league at containing passes to running backs (-11% VOA receiving by RB) but I'm not sure what that says about their ability to stop the 7-10 yard receiver pattern.  It would seem to indicate that the Pats should use the screen pass to Kevin Faulk once or twice rather than over and over.  Not to mention that, as a Pats fan, I would personally like to see an end to the "quick screen pass to a receiver who has a defender standing right in front of him on the line of scrimmage and no blockers to speak of" play that Charlie Weis seemed awful fond of against the Titans.  Blockers, Charlie, screen passes need blockers.  And while we're at it, no wacky lateral-reverse-wideout passes to Dan Koppen garbage, especially in the second half, okay dude?

Saying "penalties are bad" is pretty obvious, but as we learned in the first round of the playoffs, offensive line penalties will kill you when you play the Colts.  Penalties lead to 1st-and-15 or 1st-and-20, 1st-and-20 leads to 3rd-and-12, 3rd-and-12 leads to punt, punt leads to Peyton Manning, Peyton Manning leads to loss.

It comes down to this: For most of the season, the Colts defense has been average.  For one half against the Patriots, they were spectacular.  One set of coaches will seek to duplicate that performance.  The other set of coaches will seek to prevent it.  Which one wins out, well, I don't know, you don't know, the folks at ESPN Scouts Inc. don't know, and the numbers sure as heck don't tell you.  That's why they play the games.

Special Teams/Weather

Special Teams
  IND NWE
DVOA 0.1% (14) 0.8% (8)
IND kickoff +5.5 (9) +12.6 (4)
NWE kickoff -9.5 (26) +10.6 (6)
IND punts -17.8 (32) +1.4 (14)
NWE punts +4.2 (14) -5.9 (24)
FG/XP +19.1 (1) -10.0 (28)

I considered just leaving the space next to "IND punts" blank, but the fact is that when Hunter Smith was forced to punt this year, the Indianapolis punt coverage team was the worst in the league.  Maybe it is because they are so out of practice.  The Colts aren't much better on kickoffs.  The kickoff team was worth more points than average almost entirely because Eddie Berlin of the Tennessee Titans cannot hold onto the ball.  Those fumbles aside, the Colts give up a lot of long returns.  They gave up two to Bethel Johnson in the first meeting of these teams, and don't forget the kickoff return by Dante Hall that kept Kansas City in the game last week.

When the Pats punt, Indy gets the advantage, although Ken Walter is pretty good when punting into a short field.

Mike Vanderjagt has been perfect all year, but you have to wonder -- will he have trouble kicking field goals with a frozen ball and possible light snow, since he plays his home games in a dome, or is he used to such conditions, since he is Canadian.  New England's kicker Adam Vinatieri has been slumping horribly all year but is still known for clutch kicks, particularly clutch kicks in weather, and he hit another one to beat Tennessee.

Remember, here in Boston, the weather is like an extra player on the field.  Heck, the weather gets more playing time than Hunter Smith.

I asked Anthony Brancato, inventor of "THE SYSTEM," for Manning's cold weather statistics.  According to Brancato, "Including Sunday's playoff victory over the Chiefs, Peyton Manning is 9-7 straight up and 8-8 against the spread in his NFL career as a visitor in cold weather (both figures are, of course, well above the average records of all the warm-weather and indoor teams over the years Manning has been in the NFL)."  I'm not sure whether that statistic accurately portrays the difference between cold weather and COLD WEATHER, however.  THE SYSTEM lists Kansas City as a "cold weather" city.  A lot of people are saying "Indy has no problem playing in cold weather" because they won in Kansas City.  Please, mid-50s in KC only counts as cold if you live in Egypt.  It's going to be in the 20s here Sunday.

As for non-weather home field advantage, as I pointed out last week, our statistics disagree with the conventional wisdom about why the Patriots had a decisive home field advantage this year.  The team's defense didn't play much better at home than it did on the road, but on offense the Pats had the second-best offensive home-field advantage in the league behind Chicago.  At home, their offense scored +18.1% DVOA, but on the road they were almost opposite, with -16.9% DVOA.  At home, the Pats' offense plays like Minnesota or Seattle; on the road, more like Buffalo or Detroit.  Even in the six Pats road wins, ignoring the Buffalo debacle and the Washington turnover-fest, the Pats offense totaled a -7.9% DVOA.

(By the way, on both offense and defense, the Colts home field advantage was league average.  They weren't really better on the road, even though they were 7-1 but only 5-3 at home.  They just managed to win the close road games and lose the close home games.)

A Very Hesitant Pick

There was a very good argument to be made that last week's Patriots-Titans game featured the two best teams in football.  There is an equally good argument to be made that this week's Patriots-Colts game features the two best teams in football.  Not to give the Eagles and Panthers any bulletin board material (like they care about our little website) but there is a pretty good chance this game is, like the NBA Western Conference Finals, the real championship.

The teams are evenly matched.  Each offense is slightly better than the defense that opposes it.  The Pats have an advantage on returns, the Colts on field goals.  Each team could win this game and I would not be surprised.  Here's one more statistic I didn't include above, because it applies to whichever team has the ball, down by a touchdown or field goal, in the last minutes of the fourth quarter.  Both the Patriots and the Colts had offenses that were better than usual in the second half of close games.  Both the Patriots and the Colts had defenses that were better than usual in the second half of close games.  If the game was within a touchdown in the second half, the 2003 Colts saw their offense improve from +21% to +28% DVOA, and their defense improve from 0% to -12% DVOA.  In the same situation, the 2003 Patriots saw their offense improve from 0% to +13% DVOA, and their defense improve from -20% to -29% DVOA.  So not only are these two teams great, they are even better when the chips are down.

Colts or Pats fans expecting an easy victory are kidding themselves.  But with everything so close to call, only one team has the home field advantage.  One team is used to playing in the cold, while the other team is used to playing in a dome.  The Colts could become the first dome team in eight tries to win a conference championship on the road.  But the slight edge has to go to New England.

Posted by: Aaron Schatz on 15 Jan 2004

comments