Bill Connelly takes a look at the early-season S&P+ outliers and whether his eyeballs agree or disagree with the stats. Spoiler: Alabama is underrated. Alabama! Underrated!
13 Jan 2006
by Aaron Schatz, with additional analysis by Michael David Smith
Last year's AFC playoffs were so much fun, we've invited all the kids back for an encore performance. The Jets couldn't make it, but we really think you'll like these guys we got to replace them. They're called the "Denver Broncos."
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.
WEI DVOA is WEIGHTED 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). This is the same formula used in this week's FOXSports.com power rankings, which means that it includes last week's playoff games but does not include portions of meaningless Week 17 games for New England, Denver, and Indianapolis.
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. 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. While all the games are on the charts, the trendlines do not include Week 17 for New England or Denver, and do not include Week 16 or Week 17 for Indianapolis.
In past years, these preview articles have also served as the in-game discussion threads for playoff games. This year, we've created separate discussion threads for Saturday and Sunday on our open discussion threads page.
No long preview here. Picking the Patriots to lose a playoff game at this point is silly, not to mention disrespectful. The numbers say that New England has a 126% chance of winning this game. Frankly, Denver fans might want to consider doing something else Saturday night. Needlepoint, perhaps.
Are they gone? Did we get rid of the people who don't want to read anything negative about the Patriots? OK, good, now we can get around to discussing these two teams like normal adults.
This is where I usually put in the disclaimer about the fact that I'm a Patriots fan and at Football Outsiders we want to put our biases out in the open. The last two years, I was accused of being a Patriots homer because our numbers said the Patriots were the best team in football. This year, Patriots fans hate me. I'm here to do objective analysis and not let me personal rooting interests decide what I write, but it puts me in the very strange position where rooting for my team -- and they are still my team, nothing will change that -- also means that I'm rooting for another week of nonstop e-mails from Patriots yahoos, only some of which will get intercepted by the intern.
What the Patriots are attempting to do this year would be historic, and I'm not talking about becoming the first team to ever win three straight Super Bowls, as I detail in an excerpt from an article I wrote for Wednesday's New York Sun.
If the Patriots win this year's Super Bowl, they will not only be the first team to ever win three in a row. They will also be the worst regular-season team to ever win the Super Bowl, period. Only one team has ever won the Super Bowl with the same 10-6 record as this year's Patriots: the 1988 San Francisco 49ers.
Patriots fans might object that New England wasn't really trying to win its final game and should have finished 11-5. Still, only two teams have won the Super Bowl with an 11-5 record, the 1980 Oakland Raiders and the Patriots themselves in 2001.
Most readers know of the Pythaogrean projection, which estimates how many games a team will win based simply on points scored and allowed (explained here). In a 16-game season, only one team has won the Super Bowl despite less than 10 projected wins: the 1980 Oakland Raiders, with 9.5 projected wins. This year's Patriots had just 9.1 projected wins. (The leaders were Indianapolis with 12.7 and Seattle with 12.3.)
To win their third straight Super Bowl, New England would face the most difficult playoff path in NFL history. Last year's Patriots defeated three teams with a combined 40 wins to set a record for the strongest opposition in a three-game Super Bowl run. The Baltimore Ravens, however, faced four teams with a combined 48 wins when they won the Super Bowl as a wild card in 2000.
If the Patriots were to win their third straight Super Bowl, they would have to go through Jacksonville (12-4), Denver (13-3), probably Indianapolis (14-2), and either Seattle (13-3) or another NFC team (11-5). That opposition represents 50 or 52 wins depending on the NFC champion.
New England quarterback Tom Brady is often compared to Joe Montana, and the Patriots will try to follow the path of Montana's 1988 49ers, the only team to win the Super Bowl after a 10-6 season.
Though the 49ers were not defending champions, they had a five-year playoff streak and were coming off a 13-2 season. The 49ers struggled their way to a 6-5 record before reeling off four straight wins to claim yet another NFC West championship. (Like the Patriots, the 1988 49ers lost their final game after clinching a playoff spot.) The 49ers got past Minnesota and then won the NFC Championship by beating the league's top team, Chicago, thus avenging an earlier Monday night loss. A 20-16 comeback win over Cincinnati in the Super Bowl completed the unlikely title run.
But while the 49ers may have accomplished a feat that the Patriots are trying to duplicate, they were not the only team to try. The 1984 Washington Redskins were coming off back-to-back Super Bowl appearances. At midseason, after getting creamed by the archrival New York Giants, they were just 5-4. But they finished with six wins in seven games to enter the playoffs as one of the league's hottest teams -- only to lose their first playoff game to Chicago.
The 1974 Dolphins had won the previous two Super Bowls but started the season just 3-2. They won eight of nine to finish the season, but lost in the playoffs to the team with the most wins that year, Oakland. The 1998 Packers and 1979 Cowboys also followed back-to-back Super Bowls with playoff seasons but found that championship experience could not make up for weaknesses that opened up that season, and each failed in the postseason.
Obviously, not all of these teams are alike, and Patriots fans will object, saying that the Patriots now are not the same as the team that was 4-4 at midseason. And that's where our usual preview begins...
How do we identify a great football team? Do we look at the past few weeks, or the entire season, or should we go back to last season too? Perhaps we should toss out numbers and judge based on playoff experience, home field advantage, or unquantifiable qualities like "swagger." No game this year poses this question like the rematch of Denver and New England.
The Broncos are 13-2 since an opening day upset by Miami. The two losses came on the road by a combined five points. The wins included a 28-20 victory over the defending champions, but the game wasn't really that close: Denver was winning 28-3 in the third quarter before New England came back against a prevent defense.
But the Patriots now are very different from the Patriots then. In that loss, the Patriots played without All-Pro defensive tackle Richard Seymour, inside linebacker Tedy Bruschi, third receiver/nickel back Troy Brown, and their top two running backs, Corey Dillon and Kevin Faulk. Patriots cornerback Duane Starks was repeatedly beaten by Denver wide receivers that day, and rookie safety James Sanders was making his first career start. Both have now been replaced by players who, while not as good as the players from the 2004 Patriots, are at least league average. The Patriots are red hot now, with five wins in their last six games, including last week's 28-3 demolition of 12-4 Jacksonville.
What few people realize is that the Broncos today aren't the Broncos of that game either. Early in the season, Denver was eeking out close wins despite terrible play in some very important situations. Through six weeks, the Denver offense ranked 28th in third-down conversion rate, and the Broncos were one of the league's worst red zone teams on both offense and defense.
Since playing New England, the Denver offense is seventh in third-down conversion rate. Red zone DVOA on offense has gone from 23rd in the NFL to ninth, and red zone DVOA on defense has gone from 28th to sixth.
The NFL says that Denver has the fourth-worst pass defense in the league. Football Outsiders says that Denver has the fourth-best pass defense in the league. What's going on here?
The Broncos' defense is a prime example of why the official NFL team rankings by total yardage are ridiculous. The NFL says the Broncos are 29th in pass defense, allowing 228 yards per game. Of course, the Broncos also face 38.3 passes per game, two more than any other team and six more than the NFL average. All three of their division rivals rank in the top 12 in passing yards. Only four defenses had more interceptions and only two caused more fumbles on passing plays.
This wacky NFL ranking is one reason why people don't realize just how good Denver was this season. In reality, Denver's pass defense provides a good match for the Patriots, who have the best air attack outside of Indianapolis. Denver's run defense is mediocre, and so is New England's running game. Actually, the Denver defense is much like the Jacksonville defense that the Patriots faced last week, except healthy. The one big difference is that Denver's pass rush never gets to the quarterback, and the Broncos were last in the league in adjusted sack rate, which measures sacks per pass play adjusted for situation and opponent.
The Broncos are the best defense against tight ends and second-best against slot receivers, but middle of the road against starting wideouts. So expect more of Deion Branch and David Givens, and less of Ben Watson and Troy Brown. And while Denver's defense has improved in the red zone, New England has one of the best red zone offenses in football.
Denver built that victory over New England on big plays: a 72-yard pass to Rod Smith, a 55-yard pass to Ashley Lelie, and a 68-yard run by Tatum Bell. But both pass plays came against cornerback Duane Starks, who was burned so often early in the season that the Patriots finally had to stick him on injured reserve, unable to pay the fire insurance. Starks is recovering, but if he comes back next year, he'll have to wear a mask like Dr. Doom.
Meanwhile, as I've noted numerous times, the Patriots defense has led the league for three straight years when it comes to allowing the fewest rushing yards on long runs more than 10 yards from the line of scrimmage. That run by Bell happens to be the only carry over 20 yards allowed by New England all year. Bell also carried the ball on the second-highest run against New England, the only one of 20 yards. The Patriots played LaDainian Tomlinson, Larry Johnson, Warrick Dunn, Cadillac Williams -- and none of these guys could break one for 20 yards.
Of course, after those two runs, Bell showed why he may never be an every-down back: He tires out too quickly. After gaining 109 yards on his first seven carries, he gained only five yards on his final six carries (and had a fumble). The Patriots should be more worried about the consistent pounding of Mike Anderson than they should be about the potential for Bell to make a big play. But they've allowed just 3.2 yards per carry to running backs since Seymour returned in Week 10, so they match up well with Denver's league-best running game.
More likely, Denver will gain yardage with a succession of medium-length passes. New England's secondary has improved without Starks, but it is still the team weakness. In particular, the Patriots keep getting burned by second and third receivers, while keeping number one receivers in check. They gave up big games to guys like Samie Parker and Donte Stallworth. Last week, Matt Jones, Ernest Wilford, and even Reggie "The Canty Dance" Williams gained more yardage than Jimmy Smith. For Denver, that means Ashley Lelie, who has had a strong second half. Third receiver Charlie Adams can be found way, way down here on our secondary table of wide receivers near guys like Tyrone Calico, Taylor Jacobs, and Bobby Wade.
Patriots fans expect Denver quarterback Jake Plummer to make a mistake like the Plummer of the past, but Plummer has only thrown four interceptions in his last 14 games, while the Patriots were near the bottom of the league with just 10 picks -- half of which came from quarterbacks named "Losman" or "Bollinger." And as good as the New England offense is in the red zone, the New England defense is just as bad.
With longer kickoffs at home, punter/kickoff specialist Todd Sauerbrun doesn't need to worry so much about New England's strong kick returns (ranked eighth in the league). Broncos field-goal kicker Jason Elam actually had a subpar year, once numbers are adjusted for the altitude in Denver. The thin air is also supposed to allow longer field goals, but Elam missed his three longest kicks at home, a 52-yarder and two 53-yarders. Adam Vinatieri also missed a 53-yarder here in the first matchup, but he's still the greatest clutch kicker in NFL history.
And now, I would like to share a secret with the class. As most of you know, the formula used for weighted DVOA is based on an attempt to be as accurate as possible for every single NFL team over the last few years. Obviously, some teams can't be painted with the same broad brush, and this is what New England fans (with the exception of the one living at my house) claim about the current Patriots.
DVOA Since Week 12
So, out of curiosity, I decided to figure out the DVOA rating for New England and Denver, but only since the Patriots got all their players back. That would go back to Week 12, the loss to Kansas City. Patriots fans may object that the current streak of strong play began the next week, but that's not fair to Denver, which lost to the Chiefs on the road in Week 13 just like New England did in Week 12. So we go Week 12-16, plus the first half of Week 17 for Denver, and the wild card game for New England.
Surprise! Even if we include both Kansas City losses -- and Denver's loss was closer than New England's loss -- New England actually comes out ahead since Week 12.
So the Patriots have a much better chance to win this game than the full-season numbers might indicate, but the Broncos are a much better team than people are giving them credit for. On a neutral field, with equal rest, the Patriots would probably be a slight favorite. But they play the regular season for a reason, and the home field and extra week of preparation give the Broncos a slight advantage in a game that could easily go the other way.
As Indianapolis went undefeated week after week, two teams were considered the most likely candidates to hand them a loss: Pittsburgh and San Diego. The Colts whipped the Steelers 26-7, but the Chargers got the upset. For the Steelers to copy that result, they must learn from San Diego's strategy. That's why this preview is different -- we all know how good the Colts are, so I'm going to concentrate on the blueprint for beating them.
In the first meeting of these teams, seven of Pittsburgh's first nine plays went for either no gain or a loss, resulting in three straight three-and-outs. The Steelers' linemen are powerful, but they're sometimes a bit slow to get out of their stances, and there's no team in the league that can exploit that better than the Colts with their lightning-quick front four.
The Colts have always had trouble stopping the run, and Pittsburgh's offense is the most run-oriented in the NFL. Knowing Pittsburgh would run, the Colts brought safety Bob Sanders up to the line on every play, and running backs Willie Parker and Jerome Bettis kept running into a wall of Colts defenders while the lonely Pittsburgh receivers ran around in single coverage.
Quarterback Ben Roethlisberger overthrew Hines Ward on a 15-yard pattern on Pittsburgh's second play -- and the Steelers did not throw another pass longer than 12 yards until the final minute of the first half.
Contrast this to the Chargers' strategy when they upset the Colts. In the first quarter alone, San Diego threw four passes of 15 yards or more, three of which were caught.* Only two of their first 13 plays were runs. When Sanders played up, they threw long, and when he moved back into coverage later in the game, they ran the ball.
To win this game, the Steelers need to reverse their usual run-first mindset. We know they can do it: Roethlisberger averaged 10.4 net yards per pass in the first quarter this season, two yards higher than any other quarterback.
*(A landmark moment: This is officially the first item in a game preview to come from the Football Outsiders game charting project. Huzzah! More next week, I hope.)
Peyton Manning's brain is the engine that runs the Indianapolis offensive machine. He excels at recognizing blitzes and adjusting the play at the line to take advantage of defensive weaknesses. But he seems to have more trouble with 3-4 defenses, including the Chargers and the pre-2005 Patriots.
Constant pressure from the outside, provided by rookie linebacker Shawne Merriman, was a major part of San Diego's upset. The Chargers sacked Manning four times, and more importantly prevented him from setting his feet and forced him to step up in a pocket that had collapsed. This was an impressive achievement, because the Colts rank number one in both our offensive line metrics, adjusted line yards and adjusted sack rate.
You might think the Steelers fit the same mold as the Chargers. Back in that first meeting, for example, DeShea Townsend came on a corner blitz and sacked Manning, and the reason Manning wasn't able to do what he so often does -- throw directly to the spot that the blitzing player had vacated -- was that the Steelers do a great job of covering for the spot the blitzing player vacates. But that play was an aberration; for most of the day, Pittsburgh's pressure didn't affect Manning at all.
The danger for Indianapolis is that Pittsburgh can pressure Manning more and worry about running back Edgerrin James less. For the second straight year, James slowed down in December.
Last season, James averaged 4.8 yards per carry through 11 games, but 3.8 yards in December and January. This season, James averaged 4.5 yards per carry through 11 games, but 3.2 yards in December. The Colts hope that with two weeks of rest, James will be the same consistent back who was successful on a league-leading 62 percent of his runs. And if James is back to full strength, he can have a lot of success in the passing game. In Pittsburgh's 3-4 defense, one linebacker usually drops back into deep coverage. That exposes the one weakness in the Steelers' defense, their inability to stop passes to running backs (ranked 30th in DVOA).
Indianapolis sucks at everything having to do with special teams except kicking field goals. Pittsburgh is pretty much average in every area of special teams, but when you look at Pittsburgh's kickoff returns ranked "-0.4 (16)," note that Quincy Morgan's kickoff returns were worth an estimated 4.1 points of field position, Ricardo Colclough's kickoff returns were worth an estimated -3.7 points of field position, and other players were worth -0.8 points. Morgan will be missed.
A blueprint to beat the Colts does not necessarily mean a win over the Colts. The Colts have dominated the league all year, they're rested, and they have the crowd behind them. The Steelers will stay closer than they did in November, but they need to play mistake-free football to keep the Colts from moving on.
Oh, and one more thing:
|DVOA Since Week 12
(Meaningless Games Excluded)
127 comments, Last at 21 Jan 2006, 1:58am by thad