Thanks a lot, Dak Prescott. Now more people will think the fourth round is still a gold mine for quarterbacks, but the data says otherwise. The update to our quarterback draft study for 1994-2016 shows little has changed: finding a good QB is really hard.
16 Jan 2015
by Aaron Schatz
The NFL season, or at least it's NFC half, is ending exactly where it started: with the Green Bay Packers heading to Seattle to face the defending champion Seahawks. Seattle pulled away from Green Bay in the second half of that game to win 36-16, and by the end of the night it looked like the Seahawks would be completely unstoppable in 2014. It didn't quite work out that way. Seattle's 78.6% DVOA against Green Bay in that game ended up as their best single-game performance of the entire season. They faltered a bit over the next few weeks, starting the season just 3-3 with their defense looking almost human at times. But the Seahawks turned things around later in the season. After allowing at least 20 points in six of their first ten games, the Seahawks finished the regular season on a six-game winning streak and allowed a touchdown or less in five of those games. By the end of the season, Seattle had risen back to the No. 1 spot in our DVOA ratings for the third straight season.
Meanwhile, the Packers were certainly better than they looked that first night in Seattle. They finished the season 12-4, just like Seattle, and ranked No. 1 in offensive DVOA. But while Seattle is known for its extra-strong home-field advantage, it was Green Bay that truly built its numbers on its eight home games. At home, the Packers were 8-0 with 38.3% offensive DVOA. On the road, they were 4-4 with 11.4% offensive DVOA. No team had a bigger gap between its offensive performance at home and on the road. And now, if the Packers want to make it to Super Bowl XLIX, Aaron Rodgers will need his best road performance of the season despite a lingering calf injury. The Packers need to somehow win in Seattle, where the Seahawks defense matched the Packers offense with the highest gap betwen defensive performance at home and on the road -- although Seattle actually had a better offensive DVOA on the road (20.1%) than at home (13.2%).
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. Please remember that all stats represent regular season only, except for weighted DVOA and anything else specifically noted. All game charting for these two teams is now complete; any game charting data that appears with an asterisk appears courtesy of the ESPN Stats & Information Group. This preview has 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.
OK, kids, go on and bring out your strength vs. strength clichés.
This game brings us the No. 1 offense according to DVOA, which improved in the final two months of the season, against the No. 1 defense according to DVOA, which also improved in the final two months of the season. Green Bay isn't just the No. 1 offense overall, but also the No. 1 offense on first down and on third-and-long. Then again, the Seattle defense also led the league in DVOA on first down and on third-and-long.
The best time for the Packers to have success might be when passing the ball on second down, where they ranked second in DVOA but the Seattle defense was somehow only 19th. It was the only down/play combination where the Seahawks did not rank in the top five.
And the best place for the Packers to throw the ball if they want to have success is, of course, to the left side of the field, away from Richard Sherman. You'll notice very little difference in the Seahawks' defensive DVOA against left-side passes compared to right-side passes, but that's because of all the success of all the other players on their defense. Our charting stats on defensive coverage suggest two things: first, that Sherman is clearly better than Byron Maxwell, and second, that Sherman is not infallible.
|Seattle Cornerbacks, 2014|
Maxwell primarily lines up on the offense's left, and Sherman on the offense's right, although the Seahawks did finally start moving things around at times this year. Sherman switched sides to follow the other team's best receiver against three opponents: Dallas (Dez Bryant), Carolina (Kelvin Benjamin), and Kansas City (Dwayne Bowe, in which case we're talking about "only" receiver rather than "best"). As for the other cornerbacks, Burley was in the slot for the first half of the year but hasn't played much since Lane and Simon returned from injury at midseason. Lane is now the primary slot guy, while Simon tends to play outside when Maxwell is injured (for example, Week 8 against Carolina) or ill (for example, last week against Carolina; apparently, Maxwell is allergic to Jerricho Cotchery).
These stats certainly suggest who to go after in the Seattle secondary. In the Week 1 game, the Packers kept Jordy Nelson almost exclusively on the left side of the field and he caught 9 passes for 83 yards. The Seahawks could change their strategy and use Sherman to track Nelson around the field, but Rodgers certainly won't mind going after Maxwell with passes to Davante Adams if Adams is playing as well as he did against the Cowboys last week. Cobb had six catches for 58 yards and a touchdown in the Week 1 game, but he only caught one pass when covered by Jeremy Lane before Lane got hurt in the middle of the game. Seattle had -8.0% DVOA with 61 percent catch rate allowed against "other receivers" (usually slot guys) in the first half of the season, then -28.2% DVOA and 51 percent catch rate allowed after Lane's return in Week 10.
It's very unlikely that Green Bay can go another whole game without targeting Sherman even once; in fact, our charting says that was the only game all year where Sherman was never thrown at. Looking at who had success against Sherman, the best strategy seems to be either a) physical receivers with height (Kelvin Benjamin, Dez Bryant) or b) routes that cross guys to mess up Seattle's Cover-3 (Keenan Allen). Jordy Nelson is actually taller than Bryant, but he's not really that type of receiver, and neither is Adams, so it seems like (b) might be the better option here for the Packers.
Of course, Earl Thomas is behind everyone at free safety, basically updating the old joke about how two-thirds of the earth is covered by water and the other third by Garry Maddox. Seattle has allowed a league-low four touchdown passes that gained at least 10 yards. That's the fewest since the 2009 Jets allowed three in Rex Ryan's rookie year. (Green Bay allowed 14, which also happened to be the league average.) Thomas tied Charles Woodson for the league lead among safeties by making 57 tackles (or assisted tackles) after receptions
The Seahawks are comparatively weaker when covering tight ends and running backs, and rank 18th in DVOA against each, but even there we see the effect of some of their early-season defensive struggles. The Seahawks went from 4.5% DVOA and 70 percent catch rate for tight ends in Weeks 1 to 9, to -8.5% DVOA and 55 percent catch rate in Weeks 10 to 17. And they went from 10.9% DVOA and 79 percent catch rate for running backs in Weeks 1-9 to -12.3% DVOA and 69 percent catch rate in Weeks 10-17. Eddie Lacy was one of the best receiving backs in football this year, finishing fourth in receiving DYAR and ninth in DVOA with 42 catches for 427 yards and four touchdowns. But the Packers tight ends are decidedly unexciting. They catch some of the passes that need to be caught when it is time to go to the tight end, such as in the red zone, but nobody is going to game-plan with them in mind. Andrew Quarless had -0.3% DVOA (25th out of 50 tight ends) and rookie Richard Rodgers had -11.3% DVOA (36th). Quarless is more of a red-zone guy (15 passes in the red zone, average pass length of just 5.7 yards), while Rodgers runs the seam routes (only six passes in the red zone, and average pass length of 7.9 yards).
The issue of Rodgers throwing to his tight ends in the red zone gets us to the biggest, clearest weakness of Seattle's defense in 2014, and the biggest difference between this year's Seahawks (best defense of the year) and last year's Seahawks (one of the greatest defenses of all-time). The Seahawks' defense wasn't good in the red zone this season. I don't mean that they were average, as opposed to being fantastic elsewhere on the field. I mean, they were actually very bad. The Seahawks ranked 28th in DVOA in the red zone. They allowed 5.2 points per red zone visit by opponents; only New Orleans and Indianapolis were worse.
The good news for Seattle is that this is the weakness that got fixed when the Seahawks went on that dominating six-game winning streak to close the season. This table appeared in my ESPN Insider piece earlier this week, but it's updated here with some of Jim Armstrong's drive stats as well:
|Seattle Defense in Red Zone, 2014|
It will be interesting to see if Seattle might choose to blitz Aaron Rodgers more because his is less mobile due to the calf injury. One of the problems with the Seattle defense in the first half of the year was that they brought less pressure than last season. Last year, Seattle put the quarterback under duress on a league-high 34.4 percent of pass plays. This year, it was 24.9 percent, barely above the NFL average of 24.4 percent. But in that six-game winning streak to finish the season, the Seahawks brought pressure on 30.8 percent of pass plays.
The Seahawks' percentages for sending four, five, or six pass rushers are roughly equal to the NFL average, with the standard four pass rushers about two-thirds of the time. But the difference in yards per play as they sent more and more pass rushers was extreme. Seattle this year allowed 7.5 yards per pass with three pass rushers, 5.8 with the standard four, 4.7 with five, and 3.7 with six.* (Not six or more, as ESPN Stats & Information never recorded a 7-man pressure from the Seahawks.) And although opponents didn't often blitz against Aaron Rodgers this year -- only four starting quarterbacks faced more than four pass rushers less often -- he got successively worse with more pass rushers, with 8.2 yards per pass against four but 6.7 against five and 6.4 against six or more. Rodgers had no problem against DB blitzes (8.7 yards per pass) but that's a strategy Seattle rarely uses (5.5 percent of passes, 30th in the NFL).*
The Packers will mix things up with the run from Eddie Lacy and James Starks, and they are usually efficient when they run the ball, but the Seahawks' front is stalwart against opposing running games. Green Bay was sixth in run offense DVOA, including second on first downs (behind only the Seahawks), but Seattle was second in run defense DVOA (including third on first downs) and fifth in Adjusted Line Yards. In particular, there's a danger that the Packers will see a lot of first-and-10 runs turn into second-and-10s, as the Seahawks stuffed opposing runners for a loss or no gain 23.7 percent of the time (sixth in NFL) while Packers running backs were stuffed 21.0 percent of the time (22nd). And the Packers' "power" success rate on offense matched what Seattle allowed on defense at 59 percent; that's lower than the league average of 65 percent, although not by very much.
Directionally, the Packers' running game doesn't stand out anywhere, but the Seattle defense sure does. The Seahawks are far more susceptible to runs on the left side than the right side, especially right end. Running backs only carried the ball 24 times during the regular season on runs listed as "right end." They averaged 2.0 yards per carry with -24.4% DVOA. Five of these runs lost yardage and another five were stuffed at the line of scrimmage. Only three of these runs gained more than five yards and only one, an 18-yarder by Jonathan Stewart of Carolina, gained more than eight.
Seattle's offense is not only very good, but also very consistent. You may have noticed in the graph to the left that the Seahawks had one horrendous game when they lost to Dallas back in Week 6, but otherwise they had only one game all season where the offensive DVOA was even slightly negative. You also might notice that Seattle's second-best offensive game of the entire season came way back in Week 1, against the Green Bay Packers.
In fact, the Seattle offense is part of the secret of the greatness of the Seattle defense, and why they look even better in standard stats than they do in our advanced stats. By sustaining drives and limiting turnovers, the Seattle offense limited the number of times it forced the Seattle defense to play a short field. Therefore, the Seattle defense began its average drive with the opposing offense needing a league-high 74.8 yards to go for a touchdown.
If you have read a lot of my NFL analysis over the last few years, you have gotten used to my constant insistance that the passing game is more important than the running game. Here's the exception you've all been waiting for. A Seattle offense built around one of the best running games of the last 25 years is about to take on an opponent whose main strength against the run is the ability to stop backs after they have already gained 10 yards.
The Seahawks had 29.9% rushing offense DVOA this season, the fifth highest figure going back to 1989, and were one of only two teams (Houston was the other) with more runs than passes (after adjusting for kneeldowns and spikes). The Seahawks were only fourth in Adjusted Line Yards, in part because ALY doesn't account for all those Russell Wilson keepers, but they were also fourth in Second-Level Yards (5 to 10 yards past the line of scrimmage) and 11th in Open-Field Yards (11-plus yards past the line of scrimmage). The Seahawks converted runs on 81 percent of short-yardage tries, tied with Philadelphia for the best figure in the league. And while we haven't finished our count of broken tackles for the whole league, we can remind you that the Seahawks led the league in broken tackles last year, and we can show you
this play where Marshawn Lynch basically pushes the entire Raiders defense into the end zone for a 3-yard score.
The Packers, on the other hand, were 24th in run defense DVOA, though that improves to 20th if we only look at Weeks 10 to 17. They were even worse in Adjusted Line Yards, ranking 26th. ALY numbers suggest that the Packers are better stopping runs to the right side or up the middle, so they maybe might be a little bit better at stopping Seattle's running backs, who are also best when running to the right side or up the middle. Unfortunately, "better" in this case means "close to league average" in Adjusted Line Yards, as opposed to "good." Meanwhile, the Seahawks led the league in ALY on runs around right end and were fifth in runs up the middle.
And any suggestion that run direction makes a difference sort of presumes that the Packers know who has the ball and where the run is going. The Seahawks, of course, love to use the read option on running plays. ESPN Stats & Information recorded the Seahawks with 177 read-option runs this year, third in the NFL, and they led the league with 5.55 yards per carry on these plays. And Dom Capers and the Packers defense famously have a massive problem with the read option. The 49ers destroyed Green Bay in the 2012 Divisional Round with 323 rushing yards, including 176 on the read option. When the Seahawks and Packers played in Week 1, the Seahawks gained 49 yards with a touchdown on seven read-option runs. The Packers allowed 4.88 yards per carry on read options this season,* 20th in the NFL, although because of how their schedule worked out they haven't faced a team that uses the read option since playing Philadelphia in Week 11.
The hope might be that the tackling in the Green Bay secondary can literally keep the Seahawks from running away with this game. The one big strength of the Green Bay run defense was stopping long runs, and the Packers led the league by allowing just 0.42 Open-Field Yards per carry. Lynch gained 21 yards on a carry back in Week 1, but that was one of just six runs of 20-plus yards against the Packers during the regular season. None of them have come since Week 8, although DeMarco Murray broke that streak with runs of 26 and 30 yards last week.
All this running should set Seattle up for the play-action pass, but the Seahawks aren't as good on play-action as you might think. Seattle used play-action on 31 percent of passes, second in the NFL behind Philadelphia, but averaged only 7.4 yards per pass compared to 6.6 yards per pass the rest of the time. That gap of 0.8 yards is actually smaller than the NFL average of 1.3 yards.* (Apologies, but we don't have DVOA set up with the game charting data yet, just yardage.)
And the Seahawks somehow ranked 31st in DVOA on second-and-short (1 or 2 yards to go) this season. Only Chicago was worse. The Seahawks only gained 3.9 yards per carry on these downs, and, very oddly, the Seahawks never threw a single pass on second-and-short until Week 13. In six plays over the final five weeks they had a 12-yard completion (and fumble), a 22-yard completion, three incomplete passes, and a sack. The Seahawks make up for this by ranking third in the NFL in DVOA on third-and-short.
(In case you are curious, the Seahawks had two second-and-2s last week and Lynch ran for 1 yard and then 3 yards; Green Bay's defense ranked 18th on second-and-short, then 15th on third-and-short, and the Packers were average on defense against play-action passes.)
The Seahawks are going to need to pass the ball, and the best hope for the Packers defense is to get the Seahawks into a third-and-long situation. Seattle's passing DVOA drops from seventh on first down to 11th on second down to 22nd on third down. It seems like Seattle's running game should set them up for a lot of short third downs, but actually the Seahawks averaged 7.47 yards to go on third-down plays this year, longer than the NFL average of 7.19 yards. (These figures are for third downs only, with fourth downs removed.) The problem is that Seattle's running game gets sort of counteracted by the Seahawks' biggest offensive weakness: their penchant for flags. Seattle led the league with 130 accepted penalties this year, far ahead of the NFL average of 106 or the Green Bay total of 92. Half of those flags came on offense (only Tampa Bay had more accepted offensive penalties), making this four straight years the Seahawks have ranked first or second in offensive penalties. The biggest problem, as it seems to be every year, was false starts. Seahawks led the league with 29 false starts on offense, with no other team over 23.
When the Packers get the Seahawks into third-and-very-long situations, 10 or more yards to go, they probably want to bring the house as much as possible. In 2013, Wilson excelled against the blitz to an absurd extent, including 8.2 yards per pass against five pass rushers, 8.2 yards per pass against six or more pass rushers, and 9.9 yards per pass (second only to Philip Rivers) against DB blitzes. This year, Wilson was still excellent against those five-man pressures (8.6 yards per pass), but he wasn't so great against the big blitzes (only 5.0 yards per pass) or the DB blitzes (6.2 yards per pass).* The problem seems to be more pressure allowed by the offensive line. In 2013, ESPN Stats & Info recorded duress (or a sack) on 37 percent of pass plays with five rushers and 40 percent with six or more. This year, that went up to a league-leading 50 percent of pass plays with five rushers and 68 percent with six or more.*
In addition, the Packers' pass defense was much better when it brought more pressure. The Packers allowed just 3.5 yards per pass when they brought a big blitz (six or more) and just 4.8 yards per pass with a DB blitz.* (Oddly, the Packers were excellent on DB blitzes without actually pressuring quarterbacks -- ESPN Stats & Information has them putting the quarterback under duress on only 30 percent of these plays, the second-lowest figure in the NFL.)
I wrote in last week's preview about the weird left/right dichotomy of the Green Bay pass defense, where the Packers had the best defense in the league against passes on the offensive right even though left cornerback Tramon Williams had poor charting stats compared to the other Green Bay cornerbacks. Last week's game against Dallas certainly backed up the conclusion that Williams' poor charting stats are a more accurate representation of his performance than the "DVOA by sides" numbers, as Williams got dinged for two pass interference flags and gave up a 38-yard touchdown pass when he completely whiffed on an attempt to tackle Terrance Williams after a short 5-yard hitch. This issue becomes even more important this week because Seattle once again this year had the most right-handed passing game in the NFL, throwing a league-leading 46 percent of passes to the right side of the field -- although unlike last year, Wilson was just as good throwing to the left and middle as he was throwing to the right.
It will be interesting to see how the Seahawks react to losing Paul Richardson to a torn ACL. Richardson ran a lot of the shorter patterns that used to be Percy Harvin's after Harvin was dealt to the Jets. Richardson (-15.1%) and Harvin (-32.2%) were the only Seattle players with negative receiving DVOA and at least 20 targets this season, which is perhaps indicative of the design of those routes in the offense -- they seem to set up other things Seattle wants to do rather than being terribly efficient on their own. Ricardo Lockette (50.1% DVOA on 15 targets) might be the replacement in the lineup, but he's really much more of a deep threat, not a versatile slot player.
Both Green Bay and Seattle were generally mediocre on special teams this year, but a longer-term view suggests that Seattle does have an advantage here.
The Seahawks had ranked in the top five for special teams DVOA in both 2012 and 2013, and not because of Percy Harvin. Although the Seahawks employed the same kicker (Steven Hauschka) and punter (Jon Ryan) as years past, they declined from recent years in net punting value and, to a smaller extent, both net kickoff value and efficiency on field goals. It's the punts that will be most interesting, if the Green Bay defense can force Seattle to punt. Seattle almost never allows a punt return. That's a little different than saying that they are great on punts, because Ryan had plenty of punts this year where the returner called for a fair catch because the punt was just too short to be worth the risk of a return. However, 44 percent of punts were returned this year once we filter out blocked and aborted punt attempts. Against Seattle, that number was a league-low 28 percent, or only 17 punts.
The Packers have one strength on special teams: Micah Hyde. He may not be the fastest return man in the game, but he has exceptional start-stop speed and sees his blockers well to get to holes instead of just running into piles. Hyde tends to share return duties with Randall Cobb, who has never had the success returning punts that he has enjoyed returning kickoffs. Though Hyde only returned 14 punts this year, two of those were for touchdowns and four others went for 15 yards or more. Seattle only allowed four punt returns longer than 10 yards all season, though one was a touchdown by Stedman Bailey of the Rams. Hyde and Cobb also returned kickoffs last week rather than the usual regular-season kick returner, third-string running back DuJuan Harris.
When they are kicking or punting instead of returning, the Packers' special teams are decidedly poor. Mason Crosby has been average or worse for years, on both kickoffs and field goals. Tim Masthay was worth an estimated 9.7 points worth of field position below average in our gross punt measures, the third-worst figure in the NFL behind Michael Koenen of Tampa Bay and Drew Butler of Arizona. The punt coverage, led by Jay Elliott and Sean Richardson, is less of an issue.
Seattle's overall mediocrity on special teams this season had more ups and downs than Green Bay's, as the Seahawks ranked 29th in variance. Part of that came from Hauschka having one particularly bad day when he missed three field goals of 47, 50, and 52 yards indoors against Arizona in Week 16. (If that week never happened, Hauschka would have finished sixth in weather-adjusted field-goal value this year, not 13th.) Another part of the issue was Pete Carroll's weird punt return experiments. Earl Thomas and Richard Sherman were worth a combined -1.75 points of value on three punt returns, but the main punt returner, Bryan Walters, was worth -0.61 points of value on 27 punt returns.
Nobody is sure who will return kickoffs for the Seahawks after rookie Paul Richardson tore his ACL against the Panthers last week. Doug Baldwin returned a few punts this year when Walters was inactive and could do it; Carroll even said that third quarterback B.J. Daniels could return kickoffs after the Seahawks activated him off the practice squad.
Here's one thing I didn't mention above, because it applies to both teams on both sides of the ball. The Seahawks this year, on both offense and defense, played better after halftime. The Packers, on both offense and defense, played worse after halftime.
|Green Bay and Seattle DVOA by Quarter, 2014|
|Team||Q1||Rk||Q2||Rk||Q3||Rk||Q4/OT||Rk||1st Half||Rk||2nd Half||Rk||Late and Close||Rk|
|Green Bay offense||36.3%||1||40.7%||1||3.4%||15||10.2%||6||38.5%||1||6.7%||10||12.0%||8|
|Green Bay defense||-12.8%||4||-5.3%||12||-0.7%||16||13.6%||30||-8.6%||9||6.5%||24||5.3%||24|
This trend certainly didn't hold up in every game this year. In fact, it didn't hold up at all in last week's playoff games. The Seattle offense had almost exactly the same DVOA before and after halftime, while the Seattle defense went from -19.2% DVOA and 4.2 yards allowed per play before halftime to 20.5% DVOA and 6.9 yards allowed per play after halftime. Meanwhile, the Packers offense went from 0.2% DVOA and 4.6 yards per play before halftime to 76.0% DVOA and 8.9 yards per play after halftime.
There's a reasonable theory that perhaps Aaron Rodgers had a Toradol shot at halftime of last week's game, which is why he seemed more mobile and his throws more steady in the last 30 minutes. If he does the same thing this week, then the Packers offense will be better after halftime, just as it was last week. But it's also more likely that the Seahawks will get better later in the game, closer to how they played during most of the regular season (and in the Week 1 game between these two teams). It's something to watch for.
The Seattle defense is fantastic, but so is the Green Bay offense. Aaron Rodgers had a reasonable game against the Seahawks when these teams first played in Week 1, especially once we adjust his stats for the strength of the opponent. But he's going to have to overcome his calf injury and play even better than that in Sunday's game, because the Seattle offense is very likely going to manhandle the Green Bay defense on the ground. The Packers are going to need Rodgers to play one of his best games of the year, and a couple of big plays from Lacy. There will be a couple of drives where the Seahawks inevitably false start, hold, get stuck in third-and-15, and then are forced to punt. When the Packers get the good field position after those -- and a nice long Micah Hyde return would be an added bonus -- they need a touchdown every time. But even with that, they probably need a couple of big defensive plays, big turnovers or stops even after the Seahawks get into third-and-short or down into the red zone. I just don't see those plays as very likely.
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.
Team DVOA numbers incorporate all plays; since passing is generally more efficient than rushing, the average for passing is actually above 0% while the average for rushing is below 0%.
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).
122 comments, Last at 19 Jan 2015, 2:01pm by Sid