In this week's Varsity Numbers, Bill Connelly takes a page out of baseball's playbook and attempts to isolate power from efficiency.
04 Feb 2010
by Bill Barnwell and Aaron Schatz
The "run the ball and stop the run to win in the playoffs" bandwagon has veered off course yet again. The 2009 NFL season has delivered two pass-happy teams to South Florida, with the Indianapolis Colts and New Orleans Saints expected to produce a shootout. In a season where the mainstream media proclaimed the NFL to officially be a passing league, it seems only logical that we'd end up with two teams built around the pass.
While the Saints lead the league in rushing DVOA, they're a pass-first offense that uses the run to throw teams off the aerial attack's scent. The arrival of defensive coordinator Gregg Williams and the emergence of Jabari Greer as a shutdown corner finally gave them a pass defense to capitalize on the mistakes teams made when they were trying to keep up with Drew Brees and company.
That's the same game the Colts have played with Peyton Manning under center and Dwight Freeney on the defensive line, but that's beginning to change under new head coach Jim Caldwell. While the Colts of the aughts were famed for sacrificing run defense to try and shut down opposing passing games, it's the Saints who were 29th in the league against the run this year. The Colts blitzed more than they ever did under Tony Dungy, and while a 0.0% defensive DVOA against the run isn't anything to write home about, it's impressive considering Bob Sanders made all of three tackles during the 2009 season.
With two evenly matched teams employing similar styles, this game will either come down to one of two things: a huge dose of luck, or the intricacies of matchups and situations. Can the Colts' zones handle the Saints death-by-a-thousand-formations offense? Can the Saints stop the run if the Colts get ahead early? We'll break that and much more down in this, our Super Bowl XLIV preview.
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.
In the charts we've included on this page, games where teams sat their starters are colored differently and are not included in the trendlines. In addition, we've calculated weighted DVOA without those "sit starters" games for those respective teams, and have included those numbers in this article. Note that the Indianapolis charts include a pair of dots for Week 16; the dot that's marked a different color and not included in the trendline is that unit's DVOA after Peyton Manning departed the game in the third quarter.
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.
The Saints' passing offense is built around creating mismatches at the line of scrimmage and exploiting them with quick, high-percentage throws. They have a group of skill-position players that don't fit preconceived notions about fitting player A into position B. While they have your standard deep threat in Devery Henderson and all-around tight end in Jeremy Shockey, they take "tweener" skill sets and make success stories out of them. Wide receiver Marques Colston moves around the field and ends up in the slot or even at tight end, abusing linebackers and safeties with his huge frame. Mostly freed of the constraints of being a traditional halfback, Reggie Bush torments nickel corners and woefully overmatched linebackers in the flats. Tight end David Thomas spends most of his time lining up at fullback. All of them are great after the catch.
The ideal matchup for the Saints is a team that big-blitzes a fair amount and plays lots of man coverage, someone like the Eagles or the Redskins; Brees finds the bad matchup at the snap or stays upright long enough to find the open receiver, they break a few tackles, big plays abound. Although they're unquestionably moving away from the Tampa-2 scheme Tony Dungy favored, the Colts are still a conservative team defensively and play a lot of zone coverage. A look at our game charting stats shows the tell-tale signs of zone coverage, as the Colts cornerbacks rank much higher in yards per play than in Success Rate, and they allowed fewer yards per carry than any team in the league.
|Indianapolis Colts Cornerbacks in FO Game Charting Stats as of 2/3/10|
The Saints peddle in speed up and down the offensive roster, but the Colts can counter that with the league's fastest defense. While Colston against a linebacker is almost always a mismatch, Clint Session might prove to be the reason why the qualifier "almost" exists. The Colts have the team speed to run down Bush in the alleys and maybe even get pressure on Brees with a four-man rush.
They might be better off blitzing, though, and one of the reasons we suggest that has to do with the team on the Colts' 2009 schedule that had the most similar offense to the Saints: the New England Patriots. Before fourth-and-2, Indy had mostly been sliced apart by Tom Brady while rushing their front four and dropping into intermediate zones across the field, with one and often two safeties deep in a Cover-1 or Cover-2 alignment.
Tom Brady exploited these looks in a variety of ways, and it provides a blueprint for how the Saints are likely to attack the Colts. The Colts alternated between a 4-3 and a 4-2-5 alignment, with cornerback Jerraud Powers lining up across from Wes Welker in the slot, but dropping into a shallow zone on most plays. While the Colts are already undersized in their standard alignment, the presence of the nickelback gave the Patriots even more of an impetus to run the ball, which they did effectively. While Session's unquestionably a talented player, he gets lost at times in the running game and either gets caught in trash or misses on lunging tackles. Don't be surprised if Brees audibles into run plays against the nickel early on in the contest.
(A quick digression on the subject of missed tackles: Although the numbers are very preliminary, we should note that our game charters listed Colts safety Melvin Bullitt with 15 broken tackles, which led the league, while Session and the Eagles' Asante Samuel were tied for second with 13.)
Of course, the running game sets up play action, and that's where the Colts were really hit hard. The Patriots had five plays of 15 yards or more, and four of them came off of play action. Game charting stats show the Colts allowing 7.5 yards per pass against play-action fakes, but just 5.6 average yards on other pass plays. (This gap ranked tenth in the NFL.) With the Colts content to play soft zone coverage, play action forced their linebackers to rush towards the line, which created gaping holes in between the two deep safeties and the intermediate zones. That's the perfect opportunity for Colston to get open running down the seam.
The Patriots also used the threat of play action to isolate Randy Moss against defenders one-on-one, with differing levels of success. A long touchdown pass to Moss came when he was released by a cornerback into the zone of safety Antoine Bethea, who was concerned about the play action in the backfield; by the time Bethea realized it was a play-fake, he was backpedalling and trying to catch up with Moss, who ran right by him for the score. Bethea made up for it with an interception on a very strange play; Moss got open on a deep post against a cornerback that appeared to be Tim Jennings (hard to tell from even the HD feed on TV), but as he got his step, the two deep safeties were actually swapping sides in mid-play and running to opposite sides of the field. Bethea ran from a deep zone on the right side of the offense all the way to the left hashmark, where he undercut Brady's slightly underthrown pass for an interception.
Plays like that aren't how the Colts turned things around, though; their success came with pressure. They clearly changed their scheme around after Moss caught his second touchdown pass of the day, early in the fourth quarter. Up to that point, Tom Brady had dropped back 32 times, and the Colts had blitzed five defenders or more five times. They weren't even threatening to blitz on most plays; they gave Brady easy pre-snap reads, and he took advantage.
After that touchdown, things changed. Brady dropped back 12 more times, and the Colts rushed five or more seven times. The Patriots picked up two first downs on those 12 plays after nabbing 15 first downs or touchdowns on the preceding 32. It was a small sample, but the Colts clearly changed their style and it helped things. If you want a larger sample, a season's worth of game charting says that the Colts allowed 5.0 yards per pass with five pass rushers, compared to 5.9 yards per pass with the standard four and 5.5 yards per pass on the rare occasion they sent a big blitz of six or more. If they don't switch up the pass rush against Drew Brees, he will pick them apart at the line and manipulate them with his eyes after the snap. The likelihood of a limited Dwight Freeney only makes that more of a fact.
If the Colts do manage to harass Brees, they aren't going to get a lot of sacks; both of the quarterbacks in this game get rid of the ball too quickly to take many sacks. But defenses can force Brees into more dumpoffs and incomplete passes. Both quarterbacks have good numbers when hurried, but Manning was better than Brees this season. Looking at plays with a hurry, not counting sacks, Brees dropped to ninth in the NFL with 5.8 yards per play. Only 33 percent of these plays met our guidelines for success, which put Brees 15th in the league. (By comparison, Manning was third with 6.6 yards per play and sixth with 40 percent of plays successful when hurried.)
The Saints will respond to the Colts' Cover-2 with formations designed to beat zone coverage. Expect to see them line up twin wide receivers on one side of the ball on most snaps; if they want a fullback, they'll play David Thomas there and use him as their safety valve while Colston lines up in the slot, and if they're going with three wide receivers, Colston will play in what essentially amounts to the role of a tight end while Robert Meachem and another receiver -- a wideout, Bush, Shockey, even perhaps one of the Thomases -- operate on the other side. When the Colts start figuring that out, the Saints will move to their Offset-I with David Thomas as the fullback split directly behind Shockey as the tight end, and Colston lined up tight next to Shockey, giving them a look similar to the Trips Bunch sets that the Steelers used with great success in last year's Super Bowl.
The Saints are set up well to follow the Patriots' strategy of running the ball against the Colts' speedy and undersized defenders, and then taking shots deep if the Colts move a safety up to combat the run. Although the Saints offense is based around the pass, they run a lot more than most people realize. Only four teams had more running back carries during the regular season, and the Saints led the league in rushing DVOA. Although the Colts did a reasonable job of preventing long open-field runs, the Saints should be able to get steady gains on them, especially when they run up the middle or to the right side.
|NO Offense vs. IND Defense in Adjusted Line Yards, 2009|
|ALY||Rank||L End||Rank||L Tackle||Rank||Mid||Rank||R Tackle||Rank||R End||Rank|
The Saints also ran play-action more than any other offense in the league this year. They were one of the few teams that averaged more yards per pass without play action (8.2) than with (7.6), but 7.6 yards per pass is nothing to sneeze at. And no team in the NFL this year went deep better than the Saints did. Across the NFL, deep passes (more than 15 yards in length) gained an average of 11.8 yards. They were caught 40 percent of the time, and the league-wide DVOA on these passes was 22.5%. When the Saints threw deep, they gained an average of 17.0 yards with 77.4% DVOA and a 55 percent catch rate. The first two numbers were the best in the league, while the last was second to Houston -- but the Saints and Texans were the only two offenses to complete over 50 percent of deep passes. The Saints have faced stronger pass rushes since midseason, which means they haven't gone deep quite as often, but they're still very successful when they do it. And as Braylon Edwards showed two weeks ago, the Colts cornerbacks can get beaten deep if you have the speed to get behind them and they've switched their coverage out of the usual Cover-2. The best time to run the play-action deep pass is probably in a second-and-short situation where a run is likely, as the Colts had the league's worst defensive DVOA on second down.
However, if they want to have success running against the Colts, the Saints may want to get out of their usual patterns of which formations lead to runs and which lead to passes. It's surprising for a defense known for speed and not strength, but Colts opponents averaged 4.7 yards per carry when they spread things out with only one running back, and only 3.6 yards per carry with two backs in the backfield. That's an important split because during the regular season, 74 percent of the Saints' carries came with two players in the backfield, the second-highest percentage in the league behind Miami.
The Colts defensive line also shows strength when strength against the run is most important: short yardage. The Saints converted 67 percent of runs in power situations (seventh in the NFL) but the Colts only allowed opponents to convert 56 percent of the time (sixth in the NFL).
In many ways, the Saints defense is a second-class version of the Jets defense that Indianapolis had to beat to get to the Super Bowl. Both teams feature one cornerback who is significantly better than his teammates, with Jabari Greer set to play the role of Darrelle Revis. Both teams like to blitz -- the Saints sent more than four pass rushers on 46 percent of passes, making them the only defense that blitzed more often than the Jets during the regular season. The big difference between the two defenses comes against the run, where the Jets stayed strong despite the midseason loss of nose tackle Kris Jenkins... and the Saints are very, very weak.
The similarity between Revis and Greer suggests another big game for the Colts' second receiver, Pierre Garcon -- but upon further inspection, this may not be the case. Most teams leave their starting cornerbacks on specific sides of the field rather than moving them around so the better corner is covering the other team's top receiver, and the Saints are no different. Jabari Greer is generally playing on the left side, which is the offensive right. Offenses generally move receivers from side to side more than defenses move cornerbacks, but this is yet another way in which the Colts rarely vary their offensive game plan. 77 percent of passes to Wayne this season found him on the left side of the field. 86 percent of passes to Garcon found him on the right side. Unless somebody moves around on purpose, the Saints are going to end up with Greer on Garcon for most of the game. We're likely going to see coverage similar to the NFC Divisional Round game against Arizona, when Greer was often one-on-one against Steve Breaston while Tracy Porter and a safety (usually Roman Harper) doubled Larry Fitzgerald on the other side.
The biggest problem for the Saints will almost definitely be what happens in the slot. Against third and fourth wide receivers, most of whom normally line up in the slot, the Saints were 20th in DVOA. The game charting stats concur, as nickelbacks Randall Gay and Malcolm Jenkins came out much worse than Greer and Porter. Our "defense vs. types of receivers" stats show that once you adjust for the tendencies of each offense, Saints opponents threw to "other wide receivers" more often than against any other defense, and the Saints ranked third in yardage allowed per game to these receivers.
|New Orleans Saints Cornerbacks in FO Game Charting Stats as of 2/3/10|
Given the importance of Dallas Clark in the Colts offense, the Saints' defense against tight ends shows some interesting trends. Overall, the Saints finished fifth in DVOA against tight ends, thanks in large part to the play of middle linebacker Jonathan Vilma. However, if we break this down game-by-game, we find that the Saints struggled against the best tight ends on their schedule. Even after we do opponent adjustments, we see the Saints shutting down bad tight ends but giving up yardage to good ones. Brent Celek was the only tight end to have a 100-yard game against the Saints, but Tony Gonzalez had two good games against them, and Visanthe Shiancoe had four catches for 83 yards two weeks ago. (In fairness, not all good tight ends played well against the Saints; they kept Jason Witten to mostly inconsequential short completions, and Kellen Winslow played well against them in Week 16 but poorly in Week 11.)
Of course, Dallas Clark isn't just a good tight end; he's a good tight end who often isn't a tight end. As longtime FO readers know, Clark is a tight end because of his jersey number and his fantasy football classification; he spends more time flexed out in the slot than any other tight end in the league. Clark won't draw Vilma on most plays on Sunday; he'll draw Randall Gay, the Saints' nickelback, or outside linebackers Scott Shanle and Scott Fujita. He's a mismatch for any of them, especially Gay, on whom he has four inches and about 50 pounds.
When Clark does line up as a tight end, the Colts can opt to move Garcon into the slot or bring in Austin Collie, who is a better matchup for Gay, but not by much. The Saints could double-team the slot receiver with bracket coverage from Gay on the outside and either Shanle or Fujita on the inside, but you can't double-team everyone.
The pre-snap game between Manning and Vilma, who is entrusted with modifying coverages pre-snap by defensive coordinator Gregg Williams, is going to be fascinating to watch. On Vilma's side, he's going to have to do his best to convince Manning that the coverage he's seeing is right, only for it to change after the snap; in the meantime, when he sees Manning change the play at the line and his brain screams that the Saints need to quickly sprint into a new coverage scheme, he has to do it subtly enough that he doesn't reveal this new hand to Manning.
The Saints will blitz most on third downs, but that's also when they have the hardest time getting to the quarterback. The Saints send a big blitz on 15 percent of first downs, 22 percent of second downs, and 29 percent of third downs. Yet they rank eighth in Adjusted Sack Rate on first and second down, but 29th on third and fourth down. The Colts have the lowest Adjusted Sack Rate in the league, but that ASR gets even better on third down: a miniscule 1.4 percent. Peyton Manning only took three sacks on third down all year.
We know the Colts will have success throwing the ball, but they're bound to have even more success if the Saints defense isn't expecting pass on every play. Indianapolis doesn't run the ball very much, ranking 27th in running back carries. (14 straight wins didn't even lead to a lot of carries running out the clock, since so many of those wins were late-game comebacks.) Do the Colts try to battle the Saints' weakness with their slightly-less-weak weakness? It's probably a good idea, particularly since the Colts' best direction in Adjusted Line Yards is up the middle, where they rank eighth. That matches the weakness of the Saints' run defense, whose two defensive tackles -- Sedrick Ellis and Anthony Hargrove -- are much better pass rushers than they are run defenders. Apologies for repeating these numbers from past playoff previews, but... Sedrick Ellis had a 53 percent Stop Rate on runs, the worst of any NFL defensive tackle with at least 20 run tackles, and he made his average tackle after a gain of 3.5 yards, the second-worst figure among defensive tackles. Anthony Hargrove wasn't much better at 62 percent and 3.2 yards. Remi Ayodele actually had the best numbers of the three main Saints tackles (65 percent, 2.5 yards).
Expect to see a lot more Joseph Addai than Donald Brown in this game, considering that Addai has a better DVOA rating (7.7% vs. -12.4%), a better Success Rate (53 percent vs. 46 percent), and, goddammit, a much better grasp of blitz pickup responsibilities.
Perhaps the most important place for the Colts to get away from their pass-first mentality is the red zone. The Saints defense improves dramatically in the red zone, but that's primarily because of excellent pass defense -- so good that the Saints are first overall in red zone defensive DVOA despite ranking 21st against the run in the red zone.
Despite their poor running attack, the Colts still use a lot of play action. Charters recorded the Colts using a play fake on 23 percent of plays, fifth in the league and not too far behind New Orleans. However, unlike the Colts, the Saints had an excellent defense against play fakes, one of only seven teams that actually allowed fewer yards per pass when the opponent ran play-action before the pass.
Indianapolis has struggled on special teams for years, and "Mr. Clutch" Adam Vinatieri has been injured at midseason. Meanwhile, special teams performance was a huge part of the Saints' win in the NFC title game, and Reggie Bush is a threat to score every time you punt to him. So, this is the one place where the Saints have a clear advantage over the Colts, right? Actually, no. The Colts (20th) actually ranked higher than the Saints (28th) in the Football Outsiders special teams ratings during the regular season. The Saints' poor value on field goals was all compiled by John Carney, and replacement Garrett Hartley has been average -- but considering that just makes the two teams even.
The Saints' special teams have been great the last two weeks -- but really only the last two weeks. The Saints have only had three games this year where the special teams were worth at least five points of field position, and two of those three games have come in the postseason. There's no reason to expect that this level of performance is more "real" than what the Saints showed in the regular season.
The difference between the two teams is not the overall quality of special teams, but rather the likelihood of a big play on special teams. The Colts got a 93-yard kickoff return touchdown from Chad Simpson against the Jaguars in Week 15, but that was their only kickoff return over 40 yards all season. The Saints, by comparison, had five kickoff returns over 40 yards in the regular season, plus another two against Minnesota two weeks ago. (The Saints also have more big plays on punt returns, although the gap is smaller: including the playoffs, the Saints have nine punt returns over 10 yards while the Colts have six.) To neutralize Saints kick returner Courtney Roby, the Colts will need a big game from the rookie with the big leg, punter/kickoff specialist Pat McAfee. Based on FO measures, McAfee was the best kickoff man in the league this season. Take out squibs and onside kicks, and McAfee leads the league with 70.8 yards per kickoff. The problem is that the Colts were just average in net kickoff value because their coverage team gave almost all that value back by allowing long returns. McAfee needs to keep the game out of the hands of his coverage team by kicking as many touchbacks as he can.
Scott Green will be the referee for Super Bowl XLIV. Stats on Green from the last three seasons show two tendencies. First, his crew calls offensive holding much less often than average. This is less of an issue for the Super Bowl because the Super Bowl uses an "all-star" crew. The umpire will be the awesomely-named Undrey Wash, and he's called offensive holding at an average rate over the last couple seasons. However, the second tendency is that Green almost never calls roughing the passer. He's thrown only seven flags for roughing the passer over the past three seasons, including zero during the entire 2008 season. You may remember him missing a couple of possible roughing the passer situations in the Green Bay-Arizona wild card game, including that facemask on Aaron Rodgers during the fumble that ended the game. This could be bad news for the Colts given Gregg Williams' promise that the Saints defense will deliver "remember me shots" to Peyton Manning.
Also of note: The three officials in the secondary, Rob Bernatchi, Greg Meyer, and Greg Steed, all worked for crews that were below average in calling defensive pass interference this season.
The Saints (98 penalties, 24th) and Colts (87 penalties, 31st) were both among the least penalized teams in the league this season.
We've covered seven Super Bowls since Football Outsiders launched in 2003, and out of those seven games, Super Bowl XLIV looks like the closest matchup before the fact. Why does conventional wisdom -- and the Las Vegas line of Colts by six -- underestimate the Saints? Perhaps it is because the last game each team played is strongest in our minds, and that was the worst offensive game for the Saints since October. Perhaps it is because we think the Colts' regular season losses "don't really count." That may be true, but while the Saints' losses to Dallas and Tampa Bay are more "real," so is the fact that their regular-season victories were generally bigger than those of the Colts. In our regular-season ratings, the Saints (23.4% DVOA) ranked higher than the Colts (17.1% DVOA). In our weighted ratings, including the postseason and discounting the Curtis Painter games, the Saints (26.7% DVOA) still rank slightly higher than the Colts (25.6% DVOA) -- but the gap is minimal.
The Colts have a good chance to get out to a quick early lead, as the Saints ranked dead last in defensive DVOA in the first quarter, then eighth from the second quarter on. (Just ask the Arizona Cardinals about how easy it can be to score on the Saints early.) A big early lead that makes the Saints give up on the run is a problem, but a small early lead is easily made up, and the most likely scenario in this game is for it to stay close to the very end. Both the Colts and Saints ranked among the best teams in the league in late and close situations, on both sides of the ball.
The Saints may be a little bit overdependent on big plays, whether they be deep bombs, kick returns, or the kind of turnovers that Peyton Manning rarely makes. On the other hand, the Colts may be a little bit overdependent on the health of Dwight Freeney's ankle. Put it all in the big blender of matchups, and the result still comes out 50-50. This should be, and hopefully will be, an exciting, high-scoring, and very close game.
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.
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 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." There are two charts for each team: one for offense, one for defense. Because defensive DVOA is opposite of offensive DVOA, the defensive charts are flipped upside-down -- the higher dots still represent better games.
In the tables with cornerback stats, there are 70 cornerbacks ranked with a minimum of 40 charted passes. Passes marked Hail Mary, Hit in Motion, Tipped at Line, and Thrown Away are removed, as are wide receiver screens (not really a good measure of coverage ability because the corner is immediately blocked out of the play). These numbers are not yet corrected for "half coverage" on zone coverage plays (which of course affects the Colts numbers).
All stats except for WEIGHTED DVOA are regular season only, unless noted.
70 comments, Last at 13 Feb 2010, 11:39pm by M Styborski