TCU has played much better in the second half of games this year. What other schools have seen dramatic shifts of play after halftime?
17 Jan 2014
by Aaron Schatz
Well, here it is, the game we've all been waiting for. OK, to be more accurate, the second game of the doubleheader we've all been waiting for. However, the feeling that this Seattle-San Francisco NFC Championship matchup was always destined somewhat obscures the fact that this matchup, based on Football Outsiders from the regular season, is not really that even. Seattle was one of the best teams we've ever measured in DVOA history. San Francisco was a very good team, tied for sixth in total DVOA this season, but when you add in the usual estimate of home-field advantage you end up with a gap of more than 40 percentage points worth of DVOA. That's even true after we consider the fact that San Francisco got healthier near the end of the season -- the 49ers had a higher weighted DVOA than total DVOA, but so did the Seahawks.
This game preview follows our usual playoff format of "when Team X has the ball, when Team Y has the ball," but before we get to that, I wanted to talk about one way in which it doesn't matter which team has the ball because these two teams are so ridiculously similar. Both defenses are very good when blitzing, but neither defense blitzes very much, which is fine, because both quarterbacks excel against the blitz. These stats all come courtesty of ESPN Stats & Information:
|2013 Regular-Season Performance by Pass Rush|
|Defensive Success Rate||Yards per Play||Frequency|
|Unit||3-4||5||6+||DB Blitz||3-4||5||6+||DB Blitz||3-4||5||6+||DB Blitz|
| "DB Blitz" includes any play with a defensive back rushing the passer, no matter the number of total pass rushers.
Success Rates here are from the defense's perspective.
Another item that doesn't really fall into the category of either offense vs. defense matchup, specifically: with all the discussion about Seattle's home-field advantage, I thought the actual home/road splits for these teams in 2013 were quite interesting. Research has generally shown that the actual size of home-field advantage tends to be inconsistent from year to year, but that overall, most teams really generally have the same home-field advantage: something around seven to eight percentage points of DVOA on both offense and defense. Over the last decade, at least, the home-field advantages in the NFC West have been a bit bigger than those of the rest of the league.
Now take a look at the numbers for 2013. Seattle does come out with a stronger home-field advantage than average, but it actually isn't that much stronger. The Seahawks were the top defense in the league both at home and on the road, and a top-ten offense both at home and on the road. Meanwhile, San Francisco was actually a better team on the road in 2013, even though these numbers include that 29-3 Seahawks stomping back in September. The 49ers had the same offensive DVOA at home and on the road, and a slightly better defensive DVOA.
The point of this digression is not that San Francisco can neutralize Seattle's home-field advantage, but just that it may not be as important (or as consistent) as the prevailing narrative suggests.
Throughout this preview, I may call back to look at how San Francisco and Seattle played each other six weeks ago, when the 49ers won 19-17 at Candlestick Park. It's my opinion that the Week 2 game is too long ago, and was too one-sided, to tell us much about what strategies these teams might use against each other on Sunday.
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. Any game charting data that appears with an asterisk appears courtesy of the ESPN Stats & Information Group and is complete through the end of the season. Other game charting data (such as defensive back coverage stats) is roughly 90 percent complete. Please remember that all stats represent regular season only except for WEIGHTED DVOA, which includes the playoffs, and where otherwise noted.
For this week and for the Super Bowl, we've got two different week-to-week charts for each team, one for offense and one for defense. Because defensive DVOA is opposite of offensive DVOA, the defensive charts are flipped upside-down; thus, the higher dots still represent better games.
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 San Francisco 49ers offense really only had three particularly poor games this season: Week 2 against Seattle, Week 10 against Carolina, and Week 11 in New Orleans. Of course, that's nothing compared to the consistency of the Seattle defense, which had just one bad game all year. Even in that game, Week 9 against Tampa Bay, the defense turned things around by the fourth quarter to help the team win in overtime.
The Seahawks ranked eighth in defensive DVOA against the run, which is pretty good, but that can't even compare to the ridiculous quality of the pass defense. The Seahawks had the fifth best defense in the history of DVOA, going back to 1989. They ranked seventh or better against all five categories of receivers, including third against tight ends and second against running backs. The pass defense gets even better in the most important situations, leading the league with -78.9% DVOA against the pass on third or fourth down and -104.1% DVOA agaisnt the pass in the red zone.
How intimidating is the Seattle pass defense? The Seahawks faced a shotgun formation on only 45 percent of plays this year, which was actually the lowest rate in the league, even though Seattle opponents were almost always playing from behind. That's okay with the Seahawks, though. They were the No. 1 defense in the league against shotgun plays and against non-shotgun plays. (In Week 14, San Francisco ran 18 of 64 plays, or 28 percent, from shotgun, significantly lower than the 45 percent they used shotgun for the entire season.)
As I noted in last week's Divisional round preview, Defensive Player of the Year candidate Earl Thomas really shuts down those deep seam plays. Seattle allowed a total of eight pass attempts all year listed as "deep middle." That's not eight receptions. That's eight passes, period, two of which were actually caught (Houston tight end Garrett Graham for 31 yards in Week 4 and Atlanta's Roddy White for 20 yards in Week 10). Every other team had at least 16 deep middle passes against them, with the NFL average being 25.5.
That defense against the seam plays is one reason the Seahawks have an excellent track record for shutting down Vernon Davis, at least as a receiver. Including the two playoff games, San Francisco has thrown 22 "deep middle" passes this season. Thirteen of those were to Vernon Davis. In his last four games against Seattle, Davis has six catches for 68 yards and a touchdown. That's a nice day for Davis. It's not a nice four days.
What about Vernon Davis on shorter passes? Last week when I previewed Seattle's game with New Orleans, I pointed out that "short middle" passes actually seemed to be a relative weakness for Seattle. That makes some sense when you consider the way Pete Carroll stacks guys at the line and combine that with the press corners on the outside/Cover-3 elsewhere that Chris Brown wrote about at Grantland yesterday. The Seahawks actually ranked just 13th in DVOA on passes to the short middle, 21st with 7.8 yards allowed per pass, and 26th with a 73 percent catch rate allowed.
This week I went to look a bit closer and the results are a bit shocking. The NFL play-by-play defines "short middle" passes as passes that go up to 15 yards through the air, but of course that's a pretty wide variety of passes. I split them up into passes of up to five yards, passes of 6-10 yards, and passes of 11-15 yards. And it turns out that while the Seahawks were money against deep middle passes, they got crushed by passes in the middle of the field that went 11-15 yards through the air. Of course, that's "crushed" with a nice sample size caveat. There were 11 of these passes. Carson Palmer threw one of them in Week 16 and got picked off in the end zone by Kam Chancellor. The other 10 were all completed, and then on top of that, Seattle gave up two first downs with pass interference on passes of this length.
That's crazy. That has to be a fluke, right?
Yes, of course it is. It looks like a one-year fluke. I checked 2012, and while Seattle once again didn't face many deep middle passes -- 15, which was the lowest figure in the league just as in 2013 -- the Seahawks allowed a 60 percent catch rate on both passes of 11-15 yards over the middle and passes of 16+ yards over the middle.
But while the extreme crazy record of giving up passes of 11-15 air yards was a fluke, it is still true that Seattle was also surprisingly average against passes of single-digit air yards that were thrown in the middle of the field. And it just so happens that San Francisco's passing game was excellent at throwing passes in the short middle of the field this season. The 49ers didn't do it that much -- only 15 percent of their passes were listed as short middle, one of the lowest rates in the NFL -- but they led the league in DVOA on these passes, about 50 percentage points ahead of the average offense. Perhaps Vernon Davis may finally have his big day against Seattle after all.
Seattle ranks just 12th against the pass on first down, then improves to rank as the best pass defense in the league on both second and third down. San Francisco's pass offense has particularly quirky down splits: fifth on first down, declining to 19th on second down, and then ranking as the best in the league on third down.
Perhaps you are wondering: What happens when the best third-down pass offense in the league faces the best third-down pass defense in the league in the playoffs? I tweeted about this a couple days ago, and folks on Twitter were wondering the same thing, so Scott Kacsmar went and looked it up for me. First of all, this is only the third playoff game since 1989 where the No. 1 pass offense on third down is taking on the No. 1 pass defense on third down. The other two: 2000 Saints (defense, won) vs. the 2000 Rams (offense) and the 2003 Ravens (defense) vs. the 2003 Titans (offense, won). There have been 40 playoff matchups where one offense and the other defense were both in the top five for third downs, and in those games the teams with the top third-down offense are 22-18, which basically teaches us nothing.
Although... Scott then looked at the actual difference in third-down DVOA between the top-five pass offense and the top-five pass defense. When the offense was at least 15 percentage points better than the defense, the offensive team was 13-7. When the defensive team was similarly better than the offensive team, the defensive team was 7-4. These are small samples, but at least they show in the more extreme cases that the superior unit on third downs wins out more often than not. The Seahawks' third-down defensive passing DVOA is 11.4% better than San Francisco's third-down offensive passing DVOA, which would be the fourth-largest difference in favor of the defense out of the 41 games being looked at here.
Seattle's cornerback charting stats this year are actually shockingly unimpressive, but that's more likely saying something about the imperfection of trying to chart cornerbacks more than its saying something about the Seattle corners being overrated. The best plays a cornerback makes are usually the ones we're not going to chart because he prevents a pass from even being thrown. Given that and the general year-to-year inconsistency of cornerback stats, the difference between being ranked 15th and being ranked 35th isn't really that significant. (What's significant is the difference between ranking 15th and ranking, say, 85th.) Charting that we've collected so far puts Richard Sherman in the 30s (out of 96 cornerbacks) in both Success Rate (57 percent) and yards per pass (6.9). We have Byron Maxwell listed with 6.8 yards per pass, which ranks 36th, but he does rank a bit lower with a 51 percent Success Rate (60th). Walter Thurmond comes out in the top ten for both Success Rate and yards per pass, but that's in part because he tends to cover such short passes as Seattle's nickelback. The average yards through the air on a pass listed with Thurmond in coverage was just 6.7, the lowest figure for any cornerback with at least 32 passes charted so far.
Given the quality of the Seattle defense, perhaps we're thinking that the 49ers should just concentrate on running the ball. Except, as we've pointed out numerous times this season, the 49ers aren't as good at running the ball as you might think. San Francisco and Seattle are the only two teams that ran on over half their offensive plays this season, but the 49ers' offensive line blocking really struggled. San Francisco finished just 29th in Adjusted Line Yards. They finished much higher in DVOA (14th) for two reasons: runs by quarterback Colin Kaepernick and a few long highlight runs by Frank Gore. The 49ers ranked fifth in Open Field Yards per Carry.
By the way, remember last year in the playoffs when the 49ers blew the Packers away with zone reads? The 49ers averaged just 3.5 yards per carry on zone reads this year, the lowest average for any team that used the play more than ten times.* For bonus points, the Seahawks led the league by allowing the fewest yards per carry on zone reads, although the sample size was quite small (57 yards on 22 carries, or 2.6 yards per). If there is a particular weakness in Seattle defensive front against the run, it comes on runs listed as right end, where the Seahawks rank 24th in Adjusted Line Yards. Perhaps the way to prevent 323-pound Red Bryant from destroying your running play is to just run wide around him.
Giving some hope to the underdog 49ers is that this matchup is one of inconsistency vs. inconsistency. As great as the Seattle defense was, it ranked 27th in variance. The San Francisco offense was 25th.
First of all, there's no reason here to dwell on what the Seahawks are missing with Percy Harvin ruled out for this game. The Seahawks only had Harvin for parts of two games this season anyway, so his presence isn't really reflected in any of the stats we'll be analyzing.
As with San Francisco, three specific games stand out as failures for the Seattle offense this year: Week 4 against Houston, the first St. Louis game in Week 8, and the upset by Arizona at Qwest Field in Week 16. When looking at those three games to try to figure out what tied them together, two ideas stood out:
One problem with the offensive line is that it produces a lot of penalty yardage. The Seahawks didn't just lead the league in penalties because their defense plays it close and physical. Only Oakland had more offensive penalties (including declined and offsetting). The Seahawks were seventh with 22 offensive holding calls (that's on offense only, not special teams) and tied for fourth with 21 false starts. The offensive line problems are also a good explanation for the biggest situational weakness of the Seattle offense: third-and-long, where the Seahawks ranked just 26th in offensive DVOA. You need to give the quarterback time to convert on third-and-long, and the Seattle offensive line just isn't very good at that. (And as we noted earlier on, you don't need to send extra pass rushers to overwhelm these blockers, which means you can leave extra guys in coverage to prevent Russell Wilson from hitting a hot read or scrambling all the way to the sticks.)
Although San Francisco's defense wasn't particularly weak on third-and-long, it wasn't particularly strong either, ranking 15th. That's just the start of a nice long list of ways in which the San Francisco defense was not as strong as you think it was this year. The 49ers are promoted as a defensive juggernaut, and they are not one. They have six-sevenths of an awfully sweet front seven, and a couple of nice safeties, but the cornerbacks are really no better than average and nose tackle is a gaping hole. Even when we look only at games with Aldon Smith in the lineup, the 49ers were nowhere near the powerful defense that the Seahawks were. Strange as it is to say, the overall production seems to be a little less than the sum of the parts.
One example of this is that San Francisco was horrendous against pass plays on first downs. The 49ers were third in pass defense on second down and seventh on third/fourth down, but 27th on first downs. The good news is that the first-down gains San Francisco allowed tended to not be of the double-digit variety, and their first-down pass defense did improve over the course of the season. Here's a look at San Francisco's pass defense on first down before and after their Week 9 bye, with the "after" section including the two playoff games. (The table says Week 1-9 even though San Francisco didn't play in Week 9 because the ranks given are based on what other teams did in those nine weeks.)
|San Francisco Pass Defense on First Downs, 2013|
One part of the problem for San Francisco was a lack of takeaways on first-down passing plays. The average NFL team had 1.14 turnovers per game on first-down pass plays. (For "turnovers" here, I'm counting both fumbles kept and lost but not Hail Mary throws.) The 49ers had only six in 18 games: five picks, including Newton last week, and an Aaron Rodgers sack-fumble that the Packers themselves recovered in the Wild Card game. Another part of the issue seems to be that the 49ers are perfectly fine giving up yardage chunks on first down when they are playing with a lead, which would seem to dovetail with Danny Tuccitto's criticism of Jim Harbaugh's philosophy when it comes to playing with a lead. They tend to stay in their 3-4 base defense and give up a lot of mid-length gains, which is why their yards allowed per play (especially in the first half of the season) ranks so much worse than first downs allowed.
Seattle's pass offense, by the way, ranked ninth in DVOA on first downs. It was comparatively at its best on second down, where it ranked third. Despite all the use of the read option, the Seahawks are in many ways a very old-school offense: they run a lot and set up play-action for downfield pass plays. Seattle ranked third in the NFL with 24 percent of its pass plays (not counting those without an intended receiver) going 16 or more yards through the air. And Seattle used play-action on 34 percent of pass plays, which led the league according to ESPN Stats & Information.* They gained 8.5 yards per pass on these plays, fourth in the NFL and first in the NFC. Continuing in our theme of San Francisco's defense being good but not spectcular, the 49ers defense allowed 6.6 yards per pass on these plays, tenth in the NFL. (The NFL average was 7.3 yards per pass.)*
When the Seahawks do go deep, they go deep down the sidelines. Remember the stat from the first section of this preview about Seattle opponents only throwing eight "deep middle" passes all year? The Seattle offense has a stat even more extreme. The Seahawks threw only two deep middle passes all season. TWO. One was to Golden Tate in Week 15 on the road, and the other to Doug Baldwin in Week 16 at home, and both were incomplete. However, when it came to deep passes marked left or right, the Seahawks had the best offensive DVOA in the entire NFL. When Sidney Rice went down, they didn't miss a beat, adjusting by sending Doug Baldwin and Jermaine "stop confusing me with former NBA star Jerome Kersey" Kearse deep more often.
The FO game charting stats suggest that if you want to go deep, you should go after Tramaine Brock and Carlos Rogers more often than Tarell Brown. All three 49ers corners have the same Success Rate in charting collected so far, 52 percent, but Brown has 6.4 yards per pass allowed (25th out of 92 corners) compared to 7.4 for Rogers (49th) and 7.6 for Brock (55th).
We hear a lot of blah-blah about teams that use the run to set up the pass, but with the Seahawks, this really is how the offense works. And the Seahawks also use the run for its own sake, and because it's extra fun to watch Marshawn Lynch knock people over. As with passing splits, the San Francisco defense was not as impressive against the run as you might think, ranking just 14th in run defense DVOA.
Seattle's running game, however, was also not as impressive as you might think, especially in the most important situations. For example, the Seahawks ranked 17th in run offense DVOA on third and fourth down. They were 10th in run offense DVOA in the red zone, but had problems when they got down to the goal line. Despite the presence of Sir Beast Mode, the Seahawks were dead last in short-yardage running, converting just 49 percent of such runs. That's very likely an issue related to the offensive line problems. It makes you think it will be almost impossible for the Seahawks to get short-yardage pickups against the 49ers after considering what they did to Carolina last week... but what the 49ers did last week wasn't what you think they did, and it bears no resemblence to how they played most of the year.
During the regular season, the 49ers ranked 18th allowing conversions on 66 percent of short-yardage runs (1-2 yards to go on either third down, fourth down, or at the goal line). And if you look at the red zone as a whole, the 49ers run defense was horrible allowing a league-worst run defense DVOA of 44.4%. Even last week, San Francisco's huge red-zone stands against Carolina were based on a couple of huge game-turning plays right at the goal line, but they gave up yardage that put the Panthers into position to not convert those runs. Near the end of the first quarter, Carolina had three straight successful runs in the red zone -- 11 yards on first-and-10 from the 17, three yards on first-and-goal from the 6, and two yards on second-and-goal from the 3 -- before they got stuffed from the 1 on two straight attempts. Then when they got the ball near the goal line again in the second quarter, Carolina again had two successful runs -- four yards on second-and-1 at the 11, then six yards on first-and-goal from the 7 -- but couldn't get in the end zone because Newton got zero yards trying to scramble on second down from the 1, and Mike Tolbert lost a yard on third down from the 1.
Yes, the 49ers made an important goal-line adjustment after the Wild Card win, with defensive coordinator Vic Fangio moving from a 6-2 defense at the goal line to a 5-3 that added Ahmad Brooks playing betwen NaVorro Bowman and Patrick Willis -- but it is still really, really, really unlikely that San Francisco would stop four different plays from the 1-yard line in consecutive weeks when they had not stopped a play from the 1-yard line during the entire 2013 regular season.
Looking at the running game in general, rather than in the red zone specifically, we can see that the Seattle running game should be able to strike well up the middle:
|Adjusted Line Yards, Seattle O vs. San Francisco D, 2013|
|Unit||L End||Rk||L Tkl||Rk||Mid/Gd||Rk||R Tkl||Rk||R End||Rk|
If they're true to their usual tendencies, Seattle's runs will often be going right at Justin Smith. Although this is in some way an issue of how the official scorers at Qwest Field mark things, Seattle led the league with 25 percent of their running back carries being listed as left tackle. Even if you look at road games only, Seattle would lead the league with 20.2 percent of running back carries coming as "left tackle." (New England was the only other team at 20 percent.) However, the Seahawks had only 37 percent of running back carries listed as middle or guard, 30th in the NFL. Only Dallas and Washington were lower.
One problem with Adjusted Line Yards, I will note, is that I created the statistic long before the arrival of the zone-read option to the NFL, and so those numbers only include running back carries. When the Seahawks run outside, they're probably going zone read, and that should be another advantage. According to ESPN Stats & Information, the Seahawks ran 101 zone reads during the regular season, fourth in the NFL, and gained 4.79 yards per carry.* The San Francisco defense allowed an average of 4.55 yards on zone reads, 11th in the league.*
This is an area where both teams are strong, and in similar ways. Both teams are were excellent this season on field goals, kickoffs and punts. The difference is in returns, and while Seattle was the better team during the season, San Francisco might actually have a small advantage based on who is getting used as a returner at this point. Seattle got very strong punt returns from Golden Tate, but used a mess of people on kickoff returns, none of whom were very good. We'll have to see if Percy Harvin is healthy enough to play, and if Seattle would then use him on kick returns in this game. Most of San Francisco's negative value on kickoff and punt returns comes from Kyle Williams, who was cut at midseason. LaMichael James has taken over both roles and has positive value in both.
I've been driving the Seahawks bandwagon for two years, so I'm not stopping now. Seattle was the No. 1 team in DVOA in 2012. I picked them to win the Super Bowl in the preseason, and then they went out and led the league in DVOA again. They get the home-field advantage, and one Arizona win a few weeks ago doesn't suddenly neutralize the fact that Seattle has had the strongest home-field advantage in the NFL over the last decade. The difference between these two teams is on defense; the 49ers may have better-known names but the Seahawks give the better all-around defensive performance. Maybe their players are less known because Richard Sherman is so busy sucking up all the media oxygen on that side of the locker room and it leaves nothing for guys like Bobby Wagner and Red Bryant.
This isn't to say that the 49ers can't win this game. The Seattle defense seems well-constructed to neutralize the San Francisco offense, but Sherman can't cover both Anquan Boldin and Michael Crabtree. Both the receivers and tight ends should be able to get open in the middle of the field at times if they can stay in front of Earl Thomas. The 49ers running game can get yardage against the Seattle defense if they block like they did last week against Carolina and not like they did for most of the regular season. On the other side of the ball, there's certainly reason to believe that the 49ers' pass rush can overwhelm the Seattle blockers, which just leaves it up to the safeties and linebackers to track down Russell Wilson if he gets out of the pocket. You can see the storyline for a San Francisco upset here. It's just not as likely as a Seattle win. And for all the talk about that other game and legacies, it's very likely that whichever team wins this game will be the favorite for this year's Super Bowl.
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 rolling average of the last five games. For the Conference championships and the Super Bowl, there are two charts for each team, one for offense and one for defense. Because defensive DVOA is opposite of offensive DVOA, the defensive charts are flipped upside-down; thus, the higher dots still represent better games.
45 comments, Last at 22 Jan 2014, 3:58pm by LionInAZ