The penultimate round of the NFL playoffs is imminent, and the matchups are phenomenal – especially if you like star quarterbacks. The four starting passers this weekend have a combined seven Super Bowl wins, and the only one who has never hoisted the Lombardi trophy – Matt Ryan — will likely be MVP following his shockingly productive 2016 campaign.
In picking angles to pursue in these conference championship game previews, I’ve tried to focus on the most significant unresolved questions: In the NFC, the burningest question is whether Ryan is really as good as his numbers would indicate. In the AFC, the two multi-Super Bowl-winning QBs are more of a known quantity, and the more significant question is who else is contributing and how.
Let’s get to previewing.
Packers vs. Falcons
FiveThirtyEight forecast: Atlanta wins 61 percent of the time.
Sure, there were some years when he didn’t have the best statistical portfolio, but the rest of the league was playing “best hand” against him every time. Manning was a sure thing – ridiculously good and ridiculously consistent. At no point did I feel like there was any player I would pick ahead of him (in the short run) if I were putting together a team.
When Manning finally retired, the Packers’ Aaron Rodgers became my new presumptive best player in football. Yes, I’ve ribbed him a lot for playing good efficient football down multiple scores instead of slinging it up, but I’ve long felt he was at or close to the best in the game. And this season, his first in the post-Manning era, he went ahead and threw for 4,428 yards, a career-high and league-leading 40 touchdowns and just seven interceptions. He has been even better – epic, even — in the playoffs.
Now, if the MVP-in-waiting were a typical leading contender — like previous winners Shaun Alexander, LaDainian Tomlinson, Adrian Peterson, or even Cam Newton, say — I’d be confident sticking with Rodgers as my pick. But Ryan’s season is not so typical.
Ryan hasn’t really accomplished that much more than Rodgers — counting the playoffs, both the Packers and Falcons have 12 wins. But from a stats perspective, Ryan put together one of the best years for a QB in recent memory. He led the league with a QBR of 83.3 (the highest posted by a QB since Rodgers’s 85.5 in 2011) and a passer rating of 117.1. The choice of metric doesn’t really matter — when you complete 69.9 percent of your passes (third in the league) and still get 13.3 yards per completion (best in the league), you’re going to come out looking pretty good.
Ryan’s stats this year aren’t just good, they’re somewhat mysterious. His QBR is over 20 points higher than he posted last year (61.79), and 11 points higher than his career-best of 72.27, which he posted in 2008 as a rookie. FiveThirtyEight contributor Ty Schalter looked at how the Falcons’ offense went from good to world-beating, and it seems the best explanations we could come up with were 1) the offseason addition of center Alex Mack, and 2) they got better at stuff.
The lack of an obvious explanation for Ryan’s extraordinary stats could have implications. A QB’s unusually good season often means they got lucky or benefited from dramatic team improvement or changes in style. For example, a team that gets better across the board can bring down their QBs interception rate since the QB no longer has to gamble for wins as often. Or a team that opens up its running game or revamps its offensive line can give the QB more time to pick apart the defense. Or sometimes a QB is called to do different things – like throwing high percentage shorter passes, or throwing deep more on first down – which can dramatically affect their box score contributions. Or sometimes a QB just gets lucky on a few deep balls, and it skews his overall stats (like winning the biggest pot of the night in poker).
So, in this case I went digging through the data to look for the usual explanations for a surge like Ryan’s, but came up pretty much empty-handed. For starters, here’s a summary of the difference between last year and this year:
|Seconds per pass||2.5||2.5|
|Yards per attempt||7.5||9.3|
Some things to note:
- His average time to pass hasn’t changed.
- He has been pressured a little bit more often. (Of course, this can mean that opponents are sending more pass rushers at him instead of dropping them into coverage, so it doesn’t necessarily make his life easier or harder.)
- His improvement can’t be explained by a reduction in interceptions alone (which is often a big driver in a QB’s improved stats year-to-year). If you turned all of his extra interceptions from last year into touchdowns, you still wouldn’t be close to his TD percentage for this year.
So I tried something different, breaking down his passing by how deep it was:
|PASSES SHORT OF 1ST DOWN||<10 YARDS PAST 1ST DOWN||10+ YARDS PAST 1ST DOWN|
|% of passes||57.2||55.4||32.1||31.5||10.7||13.1|
First thing to note is that the balance of pass types changed a little, but not much: He threw deep on 13.1 percent of passes in 2016, up from 10.7 percent in 2015, and threw inside the first down marker 55.4 percent of the time, down from 57.2 percent.
And, of course, Ryan made huge improvements in all types of passes, despite not having significantly more time or less pressure in the pocket. The most noticeable situational shift is in how often he was pressured: He faced more pressure on each type of pass, and by as much as 11-12 percentage points more often than the year before on his longer throws – and those are the throws on which his stats exploded the most. If there’s a lesson there, it may be “Don’t blitz Matt Ryan,” but it’s unlikely that his staggering across-the-board improvement is a result of how his opponents played defense.
Given all this, and the fact that Ryan’s overall statistical improvement doesn’t seem to be coming from any one thing, it makes it more likely than is typical for a one-year-wonder that this season reflects a genuine improvement in his skill level.
That said, it’s still one year — which is not very long in football — and the fact that it’s so unusual for him makes it more likely that he will regress toward the mean on Sunday, or in 2017 and beyond.
Meanwhile, Rodgers’ years of consistent performance suggest he is (at least) as good as he looks, and we should expect him to stay that way until something changes. It’s still only a dozen or so of Ryan’s passes going one way vs. the other that put his 2016 above a typical Rodgers season. If I had an MVP ballot or was forced to choose, I’d still take Rodgers as the likely best player on the field Sunday (and in the game in general), but without as much confidence as usual.
Patriots vs. Steelers
FiveThirtyEight forecast: New England wins 70 percent of the time.
I apologize to Patriots fans for not mentioning Tom Brady in my MVP discussion. Brady is obviously a great quarterback and continues to put up gaudy numbers — even after losing
Randy Moss Rob Gronkowski — but in my mind there is persistently too much uncertainty about to what extent his statistical accomplishments are a product of his own greatness or his team’s. (He is the second-highest rated QB on the Patriots this season, after all.)
Bill Belichick’s run as head coach is one of the great phenomena in the modern NFL – he has had winning years every season since going 5-11 in his first with the team in 2000. He has won at least 10 games every season since 2003, and has won 14+ games more often (five times) than he has won 10 or fewer (four times, most recently in 2009!). Of course, Belichick may not have been able to do all that without his Hall of Fame QB, but the Patriots have found so many different ways to win at this point that it’s easy to lose track. (Sure, Brady threw only 2 interceptions in 12 games this year, but the Pats also led the league in scoring defense.)
But Pittsburgh is no slouch. Counting the playoffs, they’ve now won nine games in a row, and can beat you in a lot of ways – including with their feet. Ben Roethlisberger has both statistical and playoff bona fides that will put him in Canton, though his numbers are down a bit from his past couple of years. And perhaps the most exciting player on the Pittsburgh offense is fourth-year dual-threat running back Le’Veon Bell.
Generally, I’m skeptical of running backs. Yes, all of them. (Aside from Jamaal Charles, who was, at least briefly, good at everything.) Running back stats are as affected by the player’s team performance as QB stats, yet are even more volatile and even more dependent on the particulars of the offensive system. For example, is the RB’s line opening up big holes for him, or is he constantly having to break tackles just to get back to the line of scrimmage? While figuring out the causes of a back’s success and his actual value may be nigh impossible, it is possible to get a better picture. For starters, is the back getting pretty far downfield before being hit, and how good is he at picking up extra yardage after the defense makes its first contact?
I wanted to know the answer to that question for Bell, so I created a model for expected yards for a run based on where the player was first contacted relative to the line of scrimmage and what yardline his team was on, and then compared his average actual results to what the model expected.)
Generally, being in the upper right quadrant of this chart, as Bell is, is a good sign. For any particular back, their placement along each of these axes can still represent a bunch of different things. Yards before contact can be a result of getting outside very quickly (speed), or the defense being knocked back (good run blocking), or just the types of plays that are being called (inside runners obviously don’t do as well). Yards after contact can also be affected by a back’s speed (how many defenders can they outrun after breaking a tackle) as well as their elusiveness (breaking tackles) or power (running guys over). Though of course both axes are still affected by the player’s team (how many extra blocks does the player get? etc), the y axis is probably a bit more player-driven than the x-axis. Bell does pretty well on both measures, which increases the chances of him being the real deal as a runner.
We can do the same kind of analysis for a run defense as a whole, so let’s look at Bell’s opposition Sunday, New England’s rush defense:
In this case, the yards a defense allows before first making contact is likely more a case of the teams’ schemes and ability to control the battle in the trenches, while the yards after contact is more a function of tackling ability, etc. The Patriots are in the top group as far as when their opponents first get hit, and are about average in how far they get after.
One interesting thing about this is that the Patriots are often playing with the lead, meaning they are facing passes more often than usual (about 60 percent of opposing team’s plays were passes, good for 4th-highest in the league), and likely playing defense accordingly. So their opponents probably get better rushing opportunities than usual. In other words, the fact that their defensive line has been so effective getting at backs is even more impressive than it looks.
So it’s not quite the unstoppable force against the immovable object, but who wins the battle between Bell and the Patriots defense will likely be a key to who wins on Sunday.