≡ Menu

Grouping Players Into Attackers and Mitigators: Part I

I’ve been reading Chris Brown’s excellent new book, The Art of Smart Football. One of the passages in Brown’s book about legendary head coach Sid Gillman stood out to me:

Realizing that a football field is nothing more than a 53⅓-yard-wide geometric plane, Gillman designed his pass patterns to stretch defenses past their breaking points. His favorite method was to divide the field into five passing lanes and then allocate five receivers horizontally in each one. Against most zones, at least one receiver would be open.

Below is an example of what Gillman was referring to:1 you can see that, horizontally, one target will end up in each fifth of the field:

When it comes to pass patterns, the receivers are the players on the attack, and there’s a relatively wide variance in how effective a receiver can be (i.e., he can get open all the time, some of the, none of the time, etc.). But the players in pass coverage should be viewed in a different way: all they can do is mitigate the player in front of them.

If a cornerback is going against Odell Beckham Jr., then the skill level of that cornerback is really important. But if the cornerback is going against the Browns, the skill level of that cornerback isn’t nearly as important — or as valuable. During Darrelle Revis’ first stint with the Jets, there was a period of time where the Patriots top targets were Wes Welker, a slot receiver, and two tight ends: Rob Gronkowski and Aaron Hernandez. As a result, the Jets weren’t getting quite as much value out of having the game’s top shutdown cover man when playing the team’s biggest rival. That’s because Revis was stuck covering Deion Branch. Revis was a very valuable for the Jets to have on their roster when playing say, the Lions, but his skill level was in some ways “wasted” when facing the Patriots.

That’s a key difference between a player whose main job is to attack, and a player whose main job is to prevent the other player from attacking. The former is always valuable; the latter can only mitigate the other player’s value.

There’s another element at play here: the “mitigators” are also reliant on each other. Darrelle Revis is valuable on two sets of plays: those pass attempts when the opposition throws directly at him, and pass plays where the opposition is forced to make a tougher throw because the quarterback doesn’t want to throw at Revis. The Jets were wise to add Antonio Cromartie and Buster Skrine to the team this offseason along with Revis, because opponents are going to test the other Jets cornerbacks. Revis — or any other mitigator — becomes less valuable if another mitigator is bad at his job.

Think back to Nnamdi Asomugha’s tenure with the Raiders. He was a Pro Bowl cornerback in each year from ’08 to ’10, and was a first-team All-Pro choice by the AP in ’08 and ’10. But Oakland’s defense ranked 22nd, 27th, and 10th those three seasons in Net Yards per Pass. The problem: over that three-year period, Asomugha was targeted far, far less often than every other cornerback in football. Stanford Routt and Chris Johnson were not great players, and the Raiders safeties and linebackers weren’t standouts, either. Teams would be forced to throw away from Asomugha, but that wasn’t such a big deal because the other mitigators were below average.

So a cornerback — or any mitigator — is only as valuable as the person he’s trying to mitigate and the other mitigators on his team. A pass defense is only as good as its weakest link. If that line sounds familiar, it’s because it’s the same way I’ve described offensive linemen when it comes to pass blocking.

The attacker/mitigator description is more precise than the offense/defense tags, because offensive linemen (on pass plays) are a lot like defensive backs (on pass plays). Joe Thomas may be the best offensive linemen in the NFL, but well, who knows? Thomas’ value is actually limited by three things:

  • Like the Revis and Branch scenario, Thomas’ value is limited by the ability of the opposing pass rusher he’s facing. It’s great to roster Joe Thomas if you’re playing the Chiefs, but not quite as important if you’re playing a team with a below-average pass rusher at RDE/ROLB.
  • And, like Revis, Thomas’ value is limited by the ability of his fellow offensive linemen. If Cleveland has a terrible right guard, it doesn’t matter how long Thomas can mitigate the opposing right edge rusher if another defender is getting to the quarterback in two seconds.
  • There’s a third factor that limits Thomas’ value: the ability of the Browns quarterback and receivers. Call this the Jonathan Ogden problem. It doesn’t matter how good the left tackle is if his quarterback is that much of a liability. That’s why Ogden caused a glitch in the Approximate Value calculation: how much value was he really adding if the Ravens passing attack wasn’t taking advantage of its stud left tackle?2

When discussing the value of an elite left tackle, I wrote that he’s only truly valuable when:

The All-Pro left tackle does his job, and the other four, five or six blockers do their job, and the quarterback makes the right read and an accurate throw, and the receiver makes the catch, and on this particularly play, the player(s) that was (were) blocked by the All-Pro left tackle would have gotten to the quarterback in time to prevent him from throwing and completing said pass had he (they) been blocked by a replacement-level tackle.

That’s a lot of ands.  If we want to translate that to cornerbacks, you could write that an elite cornerback is only truly valuable when:

The All-Pro cornerback does his job, and the other cornerbacks and players in coverage do their job, and the opposing quarterback did his job correctly3 and the receiver the cornerback was covering did his job correctly,4 and on this particular play, the receiver would have made the catch had he been covered by a replacement-level cornerback.

Again, a lot of ands.  Football is a team sport, and there are a lot of ands when we look at the attacker side of things, too.  But the attackers bring about a wider variance in results, and are always going to be valuable.5 The mitigators?  They do seem to be more situation dependent.

That’s enough for Part I.  I’d love to hear your thoughts.

  1. While I couldn’t find the exact picture Chris used in the book, this one illustrates the same concept. []
  2. In a way, the mirror image of this applies to Revis, too. If the opposing quarterback is terrible, then Revis becomes less valuable. Revis accumulates the most value over replacement when the opposing team is more likely to complete passes against only bad players. But if the opposing team isn’t even going to do that, well, then Revis’ value over replacement isn’t quite as significant. []
  3. After all, if he throws the ball at an open receiver’s feet, no harm no foul to the defense. []
  4. If he drops the ball, no harm no foul. []
  5. Well, maybe not “always.”  How valuable is a wide receiver if the team is always winning and run-heavy? How valuable is a great clock-chewing back if his team is terrible? []
  • I think this is a pretty interesting discussion. Because of Revis’ high asking price back in the first stint with the Jets my opinion was always that he made a best fit in the NFC South because of the passing games there and I was on board with a move to Tampa long before it was really discussed (the Tampa choice of the 4 was simply cap related and the way they approached their deals being something that could work for Revis.

    Back in the Jetscap days I had worked on some stuff with corners basically putting some numbers to what you are doing here. Im far too lazy to look it up but the gist of it was the shutdown player really just siphons targets from WR A to WR B. I think the average catch rate when I did that was around 59% for an outside WR, so essentially playing a mediocre receiver just shifts a 59% chance to another 59% chance. If you prevent a star maybe its something like a 66% chance to a 59% chance, but either way the “shutdown” corner cant really shut down a passing game unless they have the ability to bait a team. Asomugha had the best “shutdown” ratings in the NFL by preventing opportunities to one player but it didnt necessarily mean the team was any better off. At the end of it all I concluded that the Jets did the right thing by investing so heavily in Revis, Cromartie and Wilson (who of course stunk, but he was a 1st rounder) even if the positional spend was so high. Basically you could only get value from Revis if he partnered with players who had similar success rates, basically funnelling things so far down the chain that you eliminate the WR attack. The only exception would be Revis in 2009 who somehow got tested a bazillion times and those passes werent going anywhere else. I can only imagine that was due to Revis being unique in coverage staying exclusively on one player and that threw teams off in their adjustments.

    That said I dont think the NFL really believes in this for the corner position when you look at the way teams are constructed. Most teams pay one guy and then let the chips fall for their other two primary players. However I think they have considered it on the tackle subject you mention. LT has been a non-moving market in part because pass rushers now line up all over the field plus the increase of 34 defenses had kind of changed some of the approach as well.

    • As always, good thoughts. Thanks.

  • Mark O’Donnell

    The intricacy that might be important here (and this might apply better to safeties than corners, but the idea is the same) is that a corner is able to add value by creating an interception. Yes, they’re still dependent on the quarterback throwing the ball, but the same can be said of wide receivers, and they are clearly attackers. I agree with the overall point of corners being mitigators, but I think that’s an important differentiation between them and offensive tackles, who don’t really have an analogous ability to add value

    • That’s a good point. Although I suppose there are some passing TDs that you could ‘credit’ to an elite LT?

  • Dr__P

    Excellent discussion of how to use individual talents in a team scheme. There are a few coaches who have made huge strides in the NFL by creating mismatches and other coaches who have countered. One key is to have the players who fit your scheme. This is harder in the salary cap era and puts a premium on scouting and evaluation.

  • Ben Fitzgerald

    Another factor is that even though u can throw away from elite CB’s, even if the players being thrown at aren’t great, shifting of safeties can give them a numbers advantage. Similiar things happen on o-line ie shift protections away from elite LT.

  • Clint

    This exact scenario happened to the Browns last year. Alex Mack went down and the offense was miserable. Couldn’t run with any of our backs (not just Ben Tate), and Hoyer was doing his best Jay Cutler impression by throwing off his back foot way too often. Went from being one of the top offenses, being 1st place in the division, to losing the last five games of the season.
    Going into the year, PFF had the Browns O-Line as the best in football, and I don’t think many people argued. Losing Mack killed them though.

  • James

    Receivers value fluctuates as well, just not as much. After all, the value of Megatron went down when Golden Tate joined the team, because then someone else could get open easily. However, that’s offset by Tate’s value going up because he was no longer the #1 WR (at least when Johnson was healthy) and was playing against worse DBs.

    Thinking back, you could probably make a similar argument about Warner/Fitz/Boldin. Having two great receivers helps when playing against one or two great DBs (you either win or draw, which is better than losing), but they are less valuable when playing against medicore or bad DBs where winning twice on the same play is only a marginal improvement for the team. You could probably make a similar argument that a good QB like Warner makes a great WR less necessary. Or is it that a good QB “unlocks” a great WR like Chase has said before? I can definitely see both sides of it, but I’m not sure how it evens out over the course of 400+ pass plays.

    • LightsOut85

      Very interesting idea(s). While I would agree there’s definitely some “line” where you’re wasting talent/money on a super-elite WR group, I think the offensive line example comes into play here. Even if the worst OL player surrenders pressure only 11% of the time (that’s in the ball-park for worst in any given season, per PFF), because there’s 5 guys working together – and no one can 100% shut down the rush – it increases the chances he’ll allow pressure when one of the other 4 is also giving up pressure.

      So, similarly (or you could say, inversely), increasing WR “talent” decreases the chance all your targets’ not-open snaps happen at the same time (and the QB is….screwed). Teams probably would rather risk “overpaying” & have *someone* open much more often than other teams, than have 1 stud WR & the rest clowns (and I think most would agree that’s the best plan, providing you have a QB who can distribute the ball efficiently).

      Speaking of QB – IMO, it’s mostly a good/great QB unlocking/”filling up” a great WR (like pouring completions into an empty container). More to the point, I think great WR *production* comes from a great QB – because no matter how often a WR gets open, or how great they are after the catch, they can’t make the impact on the game if the QB doesn’t throw an accurate pass (often)**. That said, I still think there’s some (unknown) level of WR ability (ie: how often they can “get open”) needed to be a sufficient target (/piece of a passing attack). (Or rather, a level for any given ability of QB. ie: Even for those with the best vision, they still need a guy to get open ___% of the time, for them realistically find him in their reads, etc).

      **bit of a tangent: I often think about how great it would be to have receiving splits for past decades, so we could attempt to equalize the QB factor for WRs, and see who was possibly a much better producer than we originally thought. To elaborate – if we had depth-of-target splits (including receptions, yards, targets & drops) for receiving/passing yards, we could find the accuracy (drop+rec/target) of, say, the top 10 most accurate QBs (at each level) & taking each WR’s YPC & drop-rate to find their “accuracy adjusted” receiving yards (assuming they were targeted the same number of times) – and if you have routes-run, adjusted YPRR. It obviously doesn’t account for poor QB awareness (failing to notice to even target them), or “touch”/placement (in setting up easy YAC), but it does help highlight guys who are especially hurt/helped by QBs with poor/great accuracy. The few times I’ve done this with PFF’s data, it’s often WRs whose QBs lack deep-accuracy that have the most radical change (because obviously your YPC on targets past 20yds will be high, and any pass your QB “should have made” would add significant yards to your adjusted total) – AJ Green is a great example.