FanPost

Darrell Bevell and Constraint Plays: A Primer on Offensive Philosophy

Bob Donnan-USA TODAY Sports

I feel like as of late - the last month or so in particular - people have been getting on offensive coordinator Darrell Bevell for calling too many of one particular kind of play and saying that because of this the offensive scheme is poor and he ought to be canned. Aside from the fact that the offense has actually outplayed the defense this year (which is just plain true according to Football Outsider's metrics, no matter what narrative about the team you have in your head), there is this lingering sense of wrongness I felt the need to attempt to right.

The specific kinds of plays people second-guess Bevell on - particularly bubble screens and regular old running back screens - are what are known in the business as "constraint" plays. Generally speaking, constraint plays are plays that you run to get the defense to stop cheating and allow you to run your base offense against them. Essentially there are four different flavors - play-actions, draws, screens, and bubble screens, and I'll talk about why and when you call them.

Before I do so, though, there is one thing that needs to be pointed out in advance: the point of constraint plays is not always to immediately gain yardage. The point of constraint plays is to remind the defense of what you can do if they start cheating (what do I mean by cheating? I'll go into that in some more detail as I talk about the different types of constraints). Generally speaking the best way to remind them of this is to make a nice, yard-gaining, first-down-earning play, but depending on how the play worked out, a constraint play that ends in a 2 yard loss because one player sniffed out the play and stopped it while the rest of the defense bit can make the opposing defensive coordinator stop cheating just as much as if the play was traditionally successful.

Play Action

You probably see the Hawks call this particular constraint play more often than any other but there's generally not a lot of complaint about it because those plays work well, In fact, I often see people admonishing Bevell for not calling *more* of these. Personally, I think Bevell calls more than enough of them. I know that in years past the team has been at or near the top of play-fakes run, which is about right for a run-centric team like Seattle.

Basically, what the play-action is about is this: when a team like the Hawks are churning up yards on the ground, opposiing defenses will often counter this by bringing the strong safety into the box (the part of the field that the defensive linebackers and linebackers line up on a play) to add an extra defender to help out. You see the Hawks do this all the time with Kam Chancellor, actually. The thing is, cheating that safety forward (see? Told you I'd explain the cheating!) comes at a cost: that safety is now in a disadvantaged position to be able to drop back into a deep zone. On offense, the space that strong safety would have occupied is now wide open for a tight end, for example, to catch a nice ball.

So what you do is pretend that you're running the ball and then pull the ball back at the last minute. The safety's first couple steps are then towards the line of scrimmage instead of backwards (if the other side is running a Tampa 2 defense where the middle linebacker is also supposed to occupy a deep zone, the play action can be especially deadly if you can get him to bite on it). Meanwhile your tight end is running into that now-unoccupied zone, and now suddenly you have a man running free who you can either throw the ball to or who the other safety has to account for, leaving 1 on 1 opportunities for your wide receivers.

But here's the thing about the play-action: like all constraint plays, it only works when the D is cheating in towards the run. Calling play actions on passing downs (3rd and 8, for instance) is going to end up with nothing but give you a pass play that takes an extra second for the quarterback to finish his drop back. If there are only 7 guys in the box, even a well-executed play-fake will likely not make the safeties freeze for long enough to open up space for the passing game. Cornerbacks generally pay little attention to what's happening at the line anyway until it becomes obvious what kind of play it is, and so you can play-fake all you want and it will do very little against nickel and dime sets.

Successful play-actions make a defense think twice about lining up 8 men in the box and otherwise cheating up to defend the run. In turn, that makes non-constraint running plays gain more yards for you.

Draw

On the other hand, the Hawks very rarely run the draw, although I think this is misleading (and the Hawks definitely had no compunctions against running the draw when Hasselbeck and TJax were the starters). The draw play is basically the opposite of the play-action. It works the best when defenses completely stop respecting the run on passing downs, sending their pass rushers in on nothing but edge rushes or zone blitzes and dropping back their coverage into man to man or deep zones.

I know that the draw is one of those plays which maybe drive fans the craziest out of all of these constraint plays ("WHAT ARE THEY DOING CALLING A RUNNING PLAY ON 3RD AND 12 THEY ARE JUST GIVING UP!!!!!") but sometimes it has to be done. Even a slow guy like Hass can run 15 yards down an open field and slide if he needs to, and if the D is even disrespecting the run when you've got a running back in the backfield, that can be a huge gainer for you.

The reason why I say that it's misleading to say that the Hawks rarely call draw plays is that Russell Wilson is a threat to run on *every* passing play. I think people maybe give all the credit to Wilson himself when he decides to run 15 yards on 3rd and 8, but it comes from the same place. You'd better believe that Wilson is being told to take what the defense gives him on those plays, and yeah, he also does that naturally, but one thing that defenses have to account for with Seattle's offense is that practically every pass play they run is potential draw.

What the draw does, whether it works or it barely doesn't work, is it forces defenses to leave a guy or two in the center of the defense to rush the passer (unless you have a great defensive tackle, those guys very rarely get to the QB because the pocket is designed to not allow that) or, in extreme situations, leave a guy in the center of the field to "spy" the quarterback or running back, basically sitting there doing nothing until someone takes the ball upfield. That leaves you with fewer defenders in coverage or rushing around the edge and makes pass plays more successful.

Screens

Screen passes are also constraint plays. I know that they did not generally work last week because Kansas City's front seven are a disciplined lot that maintains their assignments (this is also, generally, a feature of Andy Reid coached defenses), but that also explains a big part of why the screen is such a good constraint play. On a classic running back screen, one side of the line drops back like they're setting up a pocket while the other half appears to just leave the play, running to one side to set up, literally, a screen of blockers for the running back to run behind. Meanwhile, the QB takes the snap, runs for dear life for 4 or 5 steps, and then, ideally, throws an easy quick toss over to that back on the side of the field.

On the defensive line, you're taught to follow the ball but also to keep track of what the guy across from you is doing. This is a very easy thing to forget because the play ends when you tackle the guy with the ball and people tend to talk you up if you were the person who made that play. The screen plays into the natural greed of the pass rusher - there's the quarterback, right in front of you, undefended, just asking to get hit. Then suddenly he doesn't have the ball anymore, and the play is suddenly on the other side of the field and you just got served.

As stated, this play fails when everybody keeps track of what their man is doing, but it's still a play it behooves you to run if you run it well, even against a team that is known for not blowing assignments a la Kansas City. Anybody can get greedy at any time. The screen is an especially good play to run against the blitz because, essentially, the blitzers are being told to be as greedy as possible on that play. Their job is to get to the quarterback as quickly as they can and generally don't even worry about that pulling guard and tackle.

If the screen works or even if one guy makes an amazing 1:1 tackle against the running back to hold the play to 3 yards instead of 30, it does two things. For one thing, it makes the defensive coordinator think twice about blitzing. The screen has a well-earned reputation as a blitz-killer. Back in the 60s and 70s the blitz was viewed as a high-risk, high-reward play that you generally beat by getting your QB to run around long enough to be able to find the inevitable open gap that the blitz created. Nowadays, there are different flavors of blitz (the zone blitz for instance) that maintain a level of conservatism even when you send more guys at a point of attack. The screen is still effective because it acts against the mindset of the blitz moreso than the actual playcall itself.

The other way a successful screen works is that it makes the front seven in particular pause in their pass rush to make sure that their man is staying in front of them. That extra half second can sometimes be all a pass play needs to work correctly.

Bubble Screens

The last of the 4 common constraint plays and perhaps the most controversial one for Seahawks fans is the bubble screen. It's basically not a lot more than a quick toss out to a wide receiver. Sometimes you'll pull a blocker or leave a second receiver out there to screen your pass-catcher, but the short, quick throw is the really big thing.

This is actually a staple of the college spread game and I'm sure we'll see plenty of these when the Hawks play Philly in a couple of weeks. At its most basic level, if you have 2 receivers split out wide and there's only 1 cornerback on that side of the field set to cover them, in a spread offense you call that quick bubble screen every time: that's got big gain written all over it.

Of course, defenses are generally not so stupid as to leave a man undefended ("without a hat on him" as the parlance goes), but the bubble screen will still work in a few different situations:

  • If the defender is playing off the ball, for example to defend against the receiver running a long route and getting past him, but also when it is, for instance, 3rd and 7 and he's sitting at the 7 yard marker to stop any yards gained after catching a short route.
  • If the cornerback is cheating inside, either to because he's going to blitz or to cut off a slant route.
  • The defense has 2 hats on a pair of receivers split wide, but a 3rd guy closer to the line of scrimmage is not being covered in the defense (you can run that guy in motion or just have him sprint outside when the ball is snapped in that case).
  • The defense has 2 hats on a pair of receivers split wide but the safety on that side of the play is cheating into the box for a run. This can be a trickier situation, because the pass-catcher will often have to shake off a defender as soon as he catches the ball, but if he gets free he's got nothing but open room in front of him.
  • When you're running a trips formation (3 WRs on one side of the field) and the defense counters by putting 2 cornerbacks near the line and a high safety deep. That safety isn't going to get to the bubble screen in time to stop you from getting yardage.

The key here is to, ideally, get more guys blocking than there are guys on that side of the field, but what works very nearly as well is to get a shifty wide-out in a 1 on 1 matchup with a defender, especially a cornerback. Cornerbacks are not known as the best tacklers in the world, and if you as an OC can set up a situation where a Golden Tate or a Percy Harvin has to beat just one of those guys to set up a 10 or 20 yard gain, that is a play you will run over and over again.

The bubble screen is based on some of the oldest concepts in the game but it's also one of the newest. As noted, the college spread is predicated on the threat of the bubble screen. At its finest, if you split 2 receivers wide to both sides of the field. The defense almost has to counter by sticking 2 cornerbacks each on those WRs and then sitting a safety high on each side of the field. Take a look at the numbers: that's 6 men who by definition are not in the box at all, leaving you with your 5 blockers plus your running back to gain yardage. You *never* get that kind of numerical advantage on a run play. There's a reason why pundits keep saying that at its heart the spread is a running offense: it's because it's true.

I think that, as much as any other constraint play out there, the mere threat of the bubble screen opens things up for your offense to a fantastic degree. Stacking eight men in the box is problematic because that safety will be in no position to close down on a bubble screen, so all by itself the mere threat of it opens up running lanes. On top of that, if you can prove you can gain 7 yards on 3rd and 6 by allowing your receivers to catch the ball behind the line and running forward, you will force teams to chuck them at the line, which in turn makes them a great deal more vulnerable to medium and long range passes. Also, if your offensive line is allowing defenders to get through to the QB, the bubble screen is alluring because there's essentially nothing they can do. The quarterback is literally taking the snap, looking, and throwing the ball - there's no time to get in for a sack.

Conclusion

I hope that helps folks understand why Bevell and other OCs call these plays so much. Here's the tealdeer:

  1. Constraint plays are not an offensive philosophy, they are something you run to make defenses play "base" schemes against your offensive philosophy. The play-action prevents teams run putting 8 men in the box against you on runs. The draw prevents teams from leaving large swathes of the field uncovered just because your men aren't there right now. Screens punish blitzing defenses. Bubble screens force teams to respect the formations that you run. If teams are doing all of that, constraint plays may not be effective - but then, if teams are doing all that, your base offense suddenly will be.
  2. The success of a constraint play is not always measured in the yards it gains. Just like how sometimes a long incompletion will force teams to say "oh crap, they almost hit that, we need to keep a deep safety more often", so, too, will an almost-successful constraint play make a team change what they're doing.
  3. The real value of constraint plays is what they allow the rest of your offense to do what they want to do. It's always fun to see a play-action work, but let's face it: the real reason the Chiefs or Seahawks run the play-action is so they can keep on running the football. The yards gained on the constraint play are nice, don't get me wrong, but it's the threat of the play that allows the Hawks to run Marshawn Lynch on zone blocking plays over and over and over again.

The modern game of football is at its heart extremely cerebral, especially at the NFL level. The worst run defense in the league can stop the best run offense if they know for a fact that they are going to be facing a run. The trick is to always keep the other side guessing, and that is what constraint plays do.