2021 Slot vs. Wide: Joe Burrow and Matthew Stafford Lead

Cincinnati Bengals QB Joe Burrow
Cincinnati Bengals QB Joe Burrow
Photo: USA Today Sports Images

NFL Preseason Week 1 - Our annual look at slot/wide splits in the passing game continues thanks to the charting efforts from our friends at Sports Info Solutions. Last time, we looked at wide receivers, where we saw Cooper Kupp dominate in the slot and Ja'Marr Chase stack up big numbers out wide. It will not shock you to see Matthew Stafford and Joe Burrow atop those same categories as we shift to the passers. That isn't always the case, but with Kupp and Chase so far ahead of the rest of the pack, their quarterbacks join them atop the pile.

That's not to say there's nothing unexpected about these results, so we can dive in a little deeper and see what we can find.

The following table shows the data for the 34 qualified passers from 2021. Each player's DYAR, DVOA, and number of targets are shown on passes to receivers both in the slot and split wide. The table is sorted by descending Slot%, which is passes thrown to players who were lined up in the slot as a percentage of passes thrown to players at wide receiver positions (i.e., slot and wide are included, but not at tight end or in the backfield). That does include passes to tight ends and running backs if they lined up in traditional wideout positions. Note that the charting labels come from players' locations on the field regardless of the positioning of their teammates. A receiver on one side of the formation who was a few feet away from the offensive line was considered to be in the slot even if he was the widest receiver on that side.

Quarterbacks, Slot vs. Wide, 2021
    Slot Wide    
Player Team DYAR DVOA Pass DYAR DVOA Pass Slot% DVOA Dif
Derek Carr LV 539 12.1% 294 148 5.7% 107 73.3% +6.4%
Matthew Stafford LAR 705 14.7% 338 213 9.3% 127 72.7% +5.4%
Jimmy Garoppolo SF 454 17.0% 214 72 -1.1% 84 71.8% +18.2%
Patrick Mahomes KC 677 19.4% 302 128 -0.4% 136 68.9% +19.8%
Zach Wilson NYJ 61 -7.8% 177 -34 -18.6% 80 68.9% +10.8%
Andy Dalton CHI -35 -15.7% 115 46 -1.7% 54 68.0% -14.0%
Mac Jones NE 230 1.5% 228 29 -9.1% 109 67.7% +10.6%
Aaron Rodgers GB 630 21.3% 244 79 -4.5% 124 66.3% +25.7%
Ryan Tannehill TEN 167 -2.6% 246 129 0.5% 126 66.1% -3.1%
Baker Mayfield CLE 273 11.0% 175 56 -3.8% 95 64.8% +14.8%
Lamar Jackson BAL 213 5.0% 188 94 -0.7% 104 64.4% +5.7%
Matt Ryan ATL 362 8.7% 257 125 0.4% 150 63.1% +8.4%
Justin Fields CHI 139 4.4% 121 -49 -21.4% 74 62.1% +25.8%
Taylor Heinicke WAS 123 -3.7% 204 168 5.1% 132 60.7% -8.8%
Tua Tagovailoa MIA 97 -3.6% 174 51 -6.8% 113 60.6% +3.2%
Jacoby Brissett MIA -67 -18.9% 101 41 -4.2% 66 60.5% -14.7%
Jared Goff DET 167 0.5% 188 10 -11.7% 124 60.3% +12.2%
Dak Prescott DAL 394 7.8% 270 155 -1.9% 180 60.0% +9.7%
Carson Wentz IND 65 -7.6% 213 214 7.0% 146 59.3% -14.6%
Justin Herbert LAC 312 2.9% 268 139 -2.7% 185 59.2% +5.6%
Davis Mills HOU 90 -4.9% 158 138 3.6% 113 58.3% -8.5%
Josh Allen BUF 284 1.4% 283 171 -1.7% 204 58.1% +3.1%
Tom Brady TB 520 13.2% 273 288 6.2% 197 58.1% +7.0%
Trevor Lawrence JAX -21 -13.2% 250 -36 -15.1% 183 57.7% +1.9%
Teddy Bridgewater DEN 146 1.6% 142 126 3.2% 104 57.7% -1.6%
Sam Darnold CAR 22 -9.7% 145 -100 -25.0% 107 57.5% +15.3%
Tyler Huntley BAL 119 13.3% 78 -96 -31.2% 64 54.9% +44.5%
Kirk Cousins MIN 516 21.9% 200 204 3.0% 175 53.3% +18.9%
Jalen Hurts PHI 295 16.9% 140 -4 -12.9% 124 53.0% +29.8%
Kyler Murray ARI 383 16.5% 186 268 8.3% 165 53.0% +8.2%
Daniel Jones NYG 27 -8.8% 138 -14 -14.2% 128 51.9% +5.4%
Russell Wilson SEA 321 17.4% 147 14 -11.3% 143 50.7% +28.7%
Joe Burrow CIN 359 12.6% 191 482 19.5% 196 49.4% -6.9%
Ben Roethlisberger PIT 30 -9.1% 190 80 -8.3% 250 43.2% -0.8%

Going Wide in the AFC North

In this article last year, we noted that 2020 was the first time we saw every single quarterback throw to the slot more often than they threw wide. Slot passing has become more and more prevalent and, generally speaking, has been more efficient on a per-play basis; it's the direction the league has been sliding in over the past half-decade.

Don't tell that to the AFC North, however. We had two passers throw wide more often than to the slot last year in Ben Roethlisberger and Joe Burrow. Burrow just squeaked over the 50% line as he led the Bengals to the Super Bowl, but Roethlisberger had the lowest slot percentage since Deshaun Watson in 2017. The Steelers quarterback of record, be it Roethlisberger or Mason Rudolph, has had the lowest slot percentage in the league in each of the last three seasons, but this was a new low point for them. Both passers saw significant drops in their slot percentage from 2020. Roethlisberger had an 8.7% drop, going from 51.9% to 43.2%, while Burrow's 9.4% drop was the second-largest in the league.

In both cases, personnel was probably the deciding factor. You don't have to dig hard to figure out why Burrow's numbers shifted—Ja'Marr Chase is really good, Ja'Marr Chase doesn't line up in the slot, Joe Burrow liked to throw the ball to Ja'Marr Chase. Slot percentage is really about scheme and receiver talent, as opposed to a quarterback's personal preferences. It's not that Burrow suddenly got better at looking slightly further to his right. It's that when he looked slightly further afield in 2020, he saw the ghost of A.J. Green, and when he looked there in 2021, he saw Ja'Marr Chase. If Burrow's wide percentage hadn't increased, we would be having serious talks about his decision-making skills. Similarly, we talked in the receivers article about how the worst wide receiver in the slot last season was Pittsburgh's Ray-Ray McCloud, as the Steelers never really figured out what to do after JuJu Smith-Schuster went down with a shoulder injury.

We'll get back to Burrow shortly, but other passers saw their slot percentage change significantly in 2021. There's no stat out there that says "this is a scheme stat" more than the fact that Jared Goff saw his slot percentage decrease the most in 2021 (-12.6%) moving from Los Angeles to Detroit, while Matthew Stafford saw his slot percentage increase the most (+8.9%) moving from Detroit to Los Angeles. Pretty much every other significant change for returning passers came from either a change in location (Teddy Bridgewater, Andy Dalton) or a change in playcaller (Kirk Cousins, Matt Ryan).

The one sort of exception to that is Daniel Jones, who saw his slot percentage drop -7.9% from 2020. Jones did get a new playcaller when Freddie Kitchens replaced Jason Garrett in November, but the drop started well before that happened. Some of it is Kenny Golladay coming in to take some of those wide targets, but I think it's also fair to say that trying to analyze the 2021 Giants offense for signs of logic or coherence is a fool's errand.

Same Team, Different Results

While most of an individual quarterback's slot percentage is based on scheme and receivers, that doesn't mean that two players in the same situation will perform identically. Two teams had a pair of qualified quarterbacks with substantially different target distributions.

One was the Baltimore Ravens, were Lamar Jackson (64.4%) threw to the slot far more often than Tyler Huntley did (54.9%). One of the knocks on Jackson throughout his career to this point is his lack of either desire or ability to throw the ball outside. Jackson had 32.4% of his targets marked as over the middle, second to only Jimmy Garoppolo among qualified passers last season. That's going to correlate with slot percentage, as Jackson was less likely than your average quarterback to, say, look for the deep go routes on the outside. The extent to which this is a problem is probably overstated, and Jackson deserve a lot of credit for finding any targets considering the Ravens' injury woes in 2021. But it's worth noting that, given the same sets of receivers and the same coaching, Jackson looked inside far more than Huntley did. And with both Marquise Brown and Sammy Watkins out of town, Jackson will be missing his top two wide targets in 2021. Can you build an offense solely out of players in the slot? Maybe they need to find someone who had a cup of coffee with Sean McVay to short these things out.

The other team with notable splits was the Chicago Bears, where Andy Dalton targeted the slot 68.0% of the time to Justin Fields' 62.1%. Fields was the more aggressive passer, and Matt Nagy and Bill Lazor responded by giving Fields looks downfield and near the boundaries that Dalton simply wasn't going to make considering the gap in arm strength between the two. Fields was third from the bottom in DVOA to wide targets, but jumps to 12th best if you only look at deep shots to wide players—and that qualifies as a highlight, considering how bad Fields' numbers were as a rookie. Interestingly, the lion's share of Allen Robinson's slot targets came with Fields and not Dalton, even after adjusting for opportunities—58.3% of Robinson's targets from Fields were from the slot, compared to 42.9% from Dalton. Explaining Robinson's 2021 season is extraordinarily difficult.

Wide Leaders

And now we come back to Burrow. We only have data going back to 2016, but the previous leader in wide DYAR was Aaron Rodgers with 460 in 2016. Burrow's 482 sets the new standard there—thank you, 17th game. Burrow had a 194-DYAR lead over second-place Tom Brady, the biggest gap between first and second place by a wide margin. Burrow was the only qualified passer in 2021 to hit a double-digit DVOA going out wide; he was just leaps and bounds more effective looking to the edges of the field than any other passer.

And Burrow did all of this basically only targeting Ja'Marr Chase and Tee Higgins. They account for 170 of his 196 wide targets, with no one else hitting double digits. That's unusual—Brady, for instance, had Mike Evans, Antonio Brown, Chris Godwin, Tyler Johnson, and Breshad Perriman all hit double-digit targets out wide; third-place Kyler Murray found double-digit targets for A.J. Green, DeAndre Hopkins, Antoine Wesley, Christian Kirk, and Rondale Moore. The Bengals stayed healthy at receiver all year, which helps, but it's still unusual for a passer to rely on two guys to that extreme. We'll see if Burrow and the Bengals spread the ball around more as their young talent continues to develop, or if it's going to be a relatively predictable scheme once again.

Burrow was basically the only good quarterback to have a better DVOA throwing wide than throwing to the slot. You could make an argument for Teddy Bridgewater, the only other passer to have positive DVOA in both splits and a higher DVOA out wide, but if that's where we're scrounging for big names, we have a problem. Most of the players who were better out wide simply aren't good—Jacoby Brissett, Carson Wentz, Andy Dalton, and Taylor Heinicke are your top four, and none of them will be starting for the same team in 2022.

Slot Leaders

Matthew Stafford has always been good throwing to the slot—he had 413 DYAR and an 11.7% DVOA in Detroit in 2020—so we can't credit all of his success last season to Cooper Kupp or Sean McVay. But, well, it's a lot easier to hit 700 DYAR and top our tables when you're throwing roughly one zillion routes to 2021's slot king. Stafford's DVOA drops down to -0.4% if you exclude Kupp's slot targets; the Stafford-to-Tyler Higbee connection never really took off. But it turns out that a lot of quarterbacks look worse when you remove a record-setting season from their numbers, so we'll just admire Stafford's numbers for what they were. Stafford did not set the record for most DYAR on throws to the slot despite the 17th game; that still belongs to 2016 MVP Matt Ryan with 789. Stafford joins Ryan and 2018 Jared Goff as the only quarterbacks to top 650 DYAR on throws to the slot—or, at least, they were the only ones until Patrick Mahomes did it this year as well, breaking up the McVay/Shanahan club thanks to having an extra game to work with. And also being very good, which often helps these sorts of things.

Stafford ends up atop the DYAR list because of volume, but he falls down to eighth in DVOA. Kirk Cousins takes the top spot there at 21.9%, though I'd argue that Aaron Rodgers and Mahomes finishing in the 20.0% neighborhood with significantly more attempts than Cousins are the more impressive results last season.

Rodgers and Mahomes are also interesting because they finished the year with negative DVOA out wide, despite being, well, Aaron Rodgers and Patrick Mahomes. Nearly all of their success in 2021 came with players lined up inside. And for Rodgers, most of that wide success went directly to DeVante Adams—he had a 21.5% DVOA targeting Adams wide, but a -32.6% DVOA targeting anyone else. Rodgers' top returning wide target from 2021 is Aaron Jones with nine targets. Suffice it to say, there are questions about the Packers' receiver corps this season.

Those questions exist for the Chiefs some, too, but the Mahomes-to-Tyreek Hill connection wasn't always firing on all cylinders in 2021. Mahomes had a -2.2% DVOA targeting Hill split wide, mostly because the deep ball wasn't there. Mahomes targeted Hill eight times on deep shots split wide, and Hill only caught two of them. This was something that had to get fixed between them even if Hill had stayed (they had a 32.3% DVOA hooking up wide in 2020), so perhaps it's less concerning than you might think at first blush that Hill is gone now—at least, when it comes to those wide targets. Replacing the 265 DYAR that Mahomes-to-Hill generated out of the slot is a different matter. Mahomes' slot DVOA only drops from 19.4% to 19.1% if you take Hill out of the equation, but finding someone that good on that much volume is a tough ask.

Comments

18 comments, Last at 11 Aug 2022, 11:18am

1 One thing is not adding up for me

I'm looking at Herbert's numbers and for both out wide and the slot his numbers are very so-so, both for DYAR and DVOA. He doesn't rank top 10 in DVOA or DYAR for slot or wide. Burrow, Stafford, Murray, Brady, Carr, Cousins, Ryan, Mahomes, Lamar, Jimmy and Dak all rank ahead of him for DVOA in both categories. 

And yet, by total DVOA Herbert ranks 6th. Is he really getting that big of a bump by throwing to RBs and TEs or what is going here? 

4 Similar question is: should…

Similar question is: should we be able to just infer that "QB total DYAR = slot DYAR + wide DYAR + TE/RB DYAR"? They should add, right? (Note I know "TE/RB DYAR" here isn't actually DYAR in the TE/RB pages, it's QB DYAR throwing to TE/RBs).

Doesn't seem to quite work that way: I mean, Zach Wilson's mildly positive in slot + wide (+27 DYAR) but massively negative overall (-569 DYAR) so that'd be a pretty massive split.

 

6 Don't forget that total DYAR…

Don't forget that total DYAR also includes throwaways/sacks/fumbles/intentional groundings.

So, for Justin Herbert:

  • Total DYAR: 1,341, 6th
  • Throws to WRs: 1,373, 6th
  • Throws to RBs: 293, 3rd
  • Throws to TEs: 253, 13th
  • All other plays: -578, 27th-fewest. (For comparison's sake, Joe Burrow was in last place here with -1,000 DYAR. Yes, exactly.)

While I'm here, here's the same thing for Zach Wilson:

  • Total DYAR: -569, worst among 34 qualifiers
  • Throws to WRs: 294, 29th
  • Throws to RBs: -68, 31st
  • Throws to TEs: 95, 31st
  • All other plays: -890, fourth-fewest

7 Oh, duh, forgot the fact…

Oh, duh, forgot the fact that Wilson's could be dragged down by "no receiver" plays. Negative's easy.

I still don't understand the "slot/wide" split categorization here though. Looking at the article more I guess it also includes passes to RBs/TEs if they line up in traditional WR positions (so split end/slot/flanker, presumably).

But for Herbert it just seems super-weird: "slot+wide" consists of 453 out of Herbert's 672 plays (67%), but only 34% of the DYAR? And obviously while RBs/TEs lining up in WR positions could drag down the total how do you get 451 DYAR slot+wide, but 1373 to WRs? Assuming WRs basically only lined up in "traditional WR positions" something like ~390/453 of those passes are WRs, so the other ~60-odd passes to non-WRs would have to be epically terrible on those passes to make that math work.

10 Confused on maths

Apologies if I'm being really dumb here, but you list Herbert's DYAR throwing to WRs as 1,373 which ranked 6th, but in the article his slot DYAR to WRs is 312, and his wide DYAR to WRs is 139 which is 451 WR to DYAR.

Am I missing WR screens maybe, but even so I don't understand how that would make up the 922 DYAR differential, maybe DPIs but I'd have figured that was already counted in DYAR wide/slot considering WRs already get credit for that and thus could be accounted as slot/wide.

Sorry again if I'm missing something very obvious.

3 By the numbers

Top 2 in slot DVOA > any wide DVOA

Bottom 3 in wide DVOA < any slot DVOA

Conclusion: slot higher floor and ceiling? More space to work with and get open makes sense with other data too.

5 If you sort by slot DVOA, it…

If you sort by slot DVOA, it does correlate pretty well to QBs you think of as good and QBs you think of as bad.

9 This is one of those I…

This is one of those I really wish went back farther even if we could only get back to the late 90s.

I'd imagine that several QBs would smash the existing "wide" DYAR record (and likely the slot record too). We have nothing for Manning and with so many massive DYAR seasons some of them would have been historic in this stat. Even if only 30% of his 2317 DYAR in 2006 came from wide that would still be nearly 700 DYAR and with how Harrison and Wayne were used I'm guessing way more than 30% of that DYAR was from wide targets. We don't have anything from the Brady with Moss season, again with 2674 DYAR total and Moss getting 568 on a team where Welker was going to be most of the slot targets both the wide and slot numbers are likely huge. Greatest Show on Turf did a lot out wide too with Bruce and Holt most wide with Hakim/etc in the slot. Plenty more examples too but I just went with the obvious.

Heck even Rodgers in 2016, which was the bar till this year at 460, was likely one of his lower numbers in the McCarthy offenses while healthy. 2011 when Nelson had 520 DYAR, and wasn't used in the slot much at all, with Jennings, Jones, and Driver adding another 650 DYAR that had to have at least some additional wide value would be very surprising to not be over 600 wide DYAR. That offense had so many back shoulder sideline highlights. Sure some of those came from routes out of the slot but McCarthy used 4 and 5 WR sets a fair bit with those teams and spread wide a lot. Nelson got more slot work each season as I recall and was actually 213 slot DYAR vs 169 wide DYAR in 2016 (Adams was -3 slot 261 wide for the majority of Rodgers wide DYAR that year) but even with a ratio like 2016 his 2011 season would have contributed more to volume.

Slot machines dominating the league is a more recent trend. There have been some historically great offenses prior to 2016 that we just don't have the data for sadly. The trends through eras would be amazing too. I would think the west coast offense was the real trend shifter where slot starts to really soar. What would those early Marino Dolphins offensive splits have broken down as vs what the Montana 49ers WCO splits were. The differences in league years when you had some very different trends in offenses would have been cool. Modern offenses all feel a lot more similar right now. The NFL cycles of innovation and convergence are currently at a convergence it feels like.

Part of why I tell clients even if you don't know what you want to do with it, if you have the means to collect the data, collect it. You can't analyze what you don't have. Just wish we had more of this or someone had the resources to get more of this data. If you have the film you can generate it so it's likely possible to go all the way back to the early 2000's if someone has the time (though AI video processing is getting to the point where it could do a lot of the heavy lifting and just mark the handful of plays where it didn't have the angle, couldn't read the players jersey, etc)

Personally I think this particular filter on historical trends in the league would be really fun to see. Ah well.

12 I just have a lot of…

I just have a lot of difficulty interpreting this stuff due to the ambiguity in slot/wide. I mean, we talk a lot about "slot machines dominating the league" but there, they're talking about an off-the-line receiver lined up with another receiver outside them, and a lot of that is just because, well, you're adding another receiver and hey guess what most defenses didn't have 3 high-quality corners.

But "slot" here's just a different thing, so I have no idea. Marvin Harrison for instance lined up rarely in a traditional slot (although he did do that, regardless of Dungy's platitudes saying that Harrison never moved around and just beat you) but he very typically lined up in the Z (flanker) position. And generally if you look at the alignments the Colts tended to use the flanker/slot receivers were practically equidistant from the line, and I don't know where that falls into here.

If you look here at the 0:52 (or 3:16) mark you can see what I mean. Both of those are three-receiver formations with Harrison lined up equidistant to the slot receiver on the opposite side of the line, but I'd never call him a slot receiver in both of those cases.

15 I wouldn't call him slot…

I wouldn't call him slot either. So I read over the methodology again and your comment in another thread that this isn't really slot vs wide and that he might indeed be slot for this analysis which does dampen my enthusiasm for it a bit. My brain kept going with the traditional definition of a slot receiver who is, as you mention right off the bat in your reply, off the line between the linemen and the outside receiver. But you're right that isn't what this charting is actually saying. This method seems to say that you could have a 3 WR set where all of them were "slot" receivers. But it also doesn't seem it's just tight vs wide. So yeah, I'm guessing that several of Adams "slot" targets last year are like your example of Harrison. I'm guessing that gets charted how this analysis is misusing slot.

So yeah how you defend an actual slot receiver vs how you defend the outside receiver who isn't split wide changes. If this charting is counting your Harrison examples as slot, and as you mention we aren't actually sure, then it doesn't have the same value I thought. Sure a tight split vs a wide split changes things but not as much as having another WR on the same side of the line as you does. You can't play as many games with routes and such to clear spaces and all that other fun stuff.

Thanks for the reply because it just hadn't sunk in as to what this measure actually was.

I still would love more history on it, because from that you could probably figure out what it was good for telling you about QB's, schemes, receivers, etc. Give it 5 more years. Or maybe get more granular with the charting like you mentioned in the other thread.  Tight/Wide, On/Off, Inside/Outide. That would cover pretty much all the combos, wide/off/inside might look weird but it could show up in a 5 WR set where you are still wide, just not widest. Of course then you have smaller samples for everything with 8 buckets instead of 2 and all those issues.

I'll just have to think about it more. Or poke them more for how far from the line the charting needs to be to mark someone as wide.

17 Yeah, the history really…

Yeah, the history really interests me, too, but my instinct is that the "rise of the (traditional) slot WR" is more just equalizing the performance of all three positions, rather than 'traditional slot' taking over. As in, previously the Y/Z receivers were dominant because three-receiver sets weren't even necessarily dominant so obviously the F can't be there. Then as three receiver sets become more common and the value of that third receiver grows, the relative contribution of F compared to Y/Z grows, but I'd bet that from a formational standpoint it'll just even out.

That being said, as I mentioned in a previous article, there's also a bias when you look at performance of each receiver type: if you run more mid-field routes (which true slots do) you'd expect the average performance of those plays to be better than sideline routes because you do not throw questionable balls over the middle. So, for instance, if your split end's even on the corner and, well, there aren't other options available, sure, take a shot downfield, because worst-case it'll probably just be an incomplete. Obviously you're not going to "take a shot" on a crossing route that's well covered.

Stated another way, NFL passing plays aren't just "pass to slot" or "pass to wide" so grouping them based on result is a false comparison. You need to group them based on intent.

Really, my guess is that if you looked at performance on routes rather than alignment there'd probably be only a negligible difference between average performance of Y/Z/F.

13 For a QB who is supposedly …

For a QB who is supposedly 'not good' with 27 TDs and 7 INTs last year, Wentz is near the top of the Wide tables along with Burrow and Stafford, and ahead of Brady and Carr.

Maybe his poor 'Slot' rating had to do with the many WR injuries in IND last year? Just a thought. Zach Pascal was Indy's 2nd leading WR with 38 catches, Hilton missed half the year and was a shell of himself with 23 catches. 

By contrast, PHL which was 1st in rushing had Quez Watkins - no future star - their WR2 with more catches than Pascal and that stiff Reagor with 10 more catches than Hilton.

Wentz is gone because egomaniac Irsay wanted him gone. Wentz was 9th in QBR to Ryan's 21st. EPA Wentz was 11th. Ryan's 20 TD-12 INT ratio not very good, I think Colts fans who think Ryan will be an improvement are going to be sorely disappointed. We shall see.

14 Mrs.Wentz!!

I know you won't respond but you're right Wentz diminishing with Indy injuries. It's like he's a product of something...

If Irsay is an egomanic, what does that mean Roseman is?

Comparing to the much worse situation of ATL is something else. 

Save 10%
& Support Bryan
Support Football Outsiders' independent media and . Use promo code KNOWLES to save 10% on any FO+ membership and give half the cost of your membership to tip Bryan.