The Dark Side of Tampa

As Colts Fans, many of us have a fairly rudimentary knowledge of the Tampa-2 defense, which was championed by the beloved Tony Dungy.  Even this basic knowledge, though, tends to be more than what most laymen of other franchises know, particularly ones who do not share the love of zone defense, and the small quick defenders that have come to characterize the Indianapolis Colts.

For those who may not have an intimate knowledge, or simply have never been given a real explanation, allow me a minute to briefly describe this heralded scheme.

Tampa-2 Defensive Scheme as displayed by ESPN (ESPN via Stampede Blue)

The Tampa-2 defensive scheme was first developed by Monte Kiffin under the direction of Tony Dungy in Tampa Bay in the late 90s — although according to Tony Dungy, it was first used by Chuck Noll of the 1970′s Pittsburgh Steelers.  This style of defense is characterized by zone defensive schemes with a general lack of blitzing.

The safeties generally drop back, providing deep coverage between the hash-marks and the sidelines, which is where the “2″ in “Tampa-2″ comes from.  This zone tends to extend from about 15 yards beyond the Line of Scrimmage (LOS) on back.

The MLB provides coverage between the hash-marks, partially filling the gap left by the safeties, and tends to play a slightly deeper zone than the other linebackers — with his zone extending from about 10 yards beyond the LOS all the way back to nearly 20 yards deep.  The other LBs tend to cover most of the middle 50% of the field from the LOS back to 10 yards deep.

This leaves the CBs to cover the sidelines from about five to 15 yards beyond the LOS.  As with all zone defenses though, it is susceptible to being picked apart by a competent QB with time, which is where the defensive line steps in.

It is the defensive line’s job to provide harrying pressure on the QB, forcing either quick dump-off passes resulting in small yardage plays, or forcing rushed passes which are more prone to being picked off.

All-in-all, though, a uniform characteristic for every position is the necessity for speed.  Whether it is the DEs needing a quick step to get to the QB, or one of the safeties, such as Bob Sanders, quickly moving up from coverage to take down an intruding RB, speed is an essential quality for all defenders in a Tampa-2 style defense to possess.  Without speed, a smart QB can thread passes into cracks in the zone and get considerable yardage after the catch, or a breakaway RB can get through the linebackers and pick up 15 yards instead of 7.

This need for speed results in a “swarm,” effect where on any given play 4 or 5 defenders can be in position to finish a play, or provide extra support.

This style of defense has been imitated by other teams, but by and large, most players in this type of a defensive scheme can come relatively cheap, because they tend to come out of college with major question marks about their lack of, “prototypical size.”  Now, as the cases of Robert Mathis and Gary Brackett have shown, opposing teams can pay heavily for carrying that same stigma today.

That has not stopped teams from passing over these players in favor of bigger, more burly defensive cogs which better fit their schemes — as our first round of selection of Jerry Hughes would attest — but this plethora of very talented Tampa-2 style players has a dark side.  It is one of the major aspects that has failed to really be mentioned thus far, and ends up being a major crack in this style of defense.  I am talking about, of course, the poor results we tend to see in run defense.

This flaw ends up not being so much an inherent trait of the Tampa-2, but more a cruel convergence of circumstance, happenstance and just pure dumb luck.  While this unique flaw in the Colts defense cannot be attributed to any single player, or single position, there are some very unexpected accomplices which may very well get passed over in just such a discussion.

For the Colts, this position features probably the best quality starters of any defensive position, and while being a doppelganger for many opposing offenses, this position ends up having an unexpected double edge.  I am talking, of course, about our starting defensive ends, Dwight Freeney and Robert Mathis.

Sam Riche of the Indianapolis Star

Dwight Freeney and Robert Mathis in a game versus the Jacksonville Jaguars (Sam Riche)

Before I continue, I should add a note to say that I am not asserting that all the ills the Colts have experienced against opposing running attacks are the byproduct of the defensive ends.  The assertion I intend to make is that through a cruel twist of fate, the overwhelming success of our defensive ends, combined with our general defensive philosophy has created a situation whereby our generally sturdy defensive scheme is faced with a unique situation it is not designed to automatically compensate for.

As before, this will take some explaining to fully convey, but bear with me.  The inconsistency of the Tampa-2 against the run is not due to the poor quality of our defensive ends as one may think, but instead it is due to the brilliance of the defensive ends!

The fact that both Dwight Freeney and Robert Mathis are so uniquely effective in rushing the passer has made even highly regarded quarterbacks, such as Tom Brady, willing to admit that they are the scariest things on the field.  The presence of Dwight Freeney and Robert Mathis automatically sets a QB on edge, and makes him slightly paranoid because as soon he lets his guard down even a little and thinks he is safe, Freeney and Mathis will be all over the ball.

They are the single best set of defensive ends in the NFL, and this is only strengthened by their unparalleled productivity together.  But, it is because of their greatness at rushing the passer that ultimately leads to the failure of the normal run stopping potential of the rest of the unit.

A little while ago Pro-Football Focus did and expose on both Freeney and Mathis, rating Freeney as the best overall DE, while bestowing the title of best outside pass rusher to Mathis.  With nearly all of Mathis’ pass rushing stats coming from outside of the RT, and with the extreme effectiveness of Dwight Freeney’s spin move, the Colt’s DEs easily prove their worth against nearly every set of offensive tackles in the league.  What we end up seeing, though, is when the QB delays the hand-off it’s the defensive unit’s greatest weakness.

When the QB snaps the ball, both Freeney and Mathis are off the blocks about as quickly as could be ever asked of anyone, and they automatically start working around the outside edge of the tackles.  As the QB either drops back, or adjusts the ball in his hands (depending if he is under center or not), Freeney and Mathis quickly make the right and left tackles scramble to keep pace with them as they move around the edge.

It is at this point that a pre-determined running play starts to unravel the defensive fabric of the Colts.  Between one and two seconds after the ball is snapped, the RB gets the hand-off and powers towards the LOS, but instead of facing an ever shifting pocket, the RB finds himself in a 5 on 2 match up.  As the Tampa-2 defense basically bans LB blitzing, it is the job of the defensive line to put pressure on the offensive line, and to generally plug gaps.

Due to the effectiveness of Freeney and Mathis though, by this point in the play they are five to seven yards deep into the opposing backfield with another five yards or more laterally displaced from the point where the ball was snapped.  In essence, this means that they are too far flung to be of much help in stopping a running play.

When a run play is decided prior to the snap, this puts the odds of the play being successful even more in favor of the opposing offense.  What tends to happen in those instances is the OTs push Freeney and Mathis of their normal direct line, making them go wide, but instead of continuing to push them further, they stop blocking and return to their pre-snap positions, in essence turning a 3 on 2 slug fest into a 5 on 2 demolition.

While our DTs (Daniel Muir, Eric Foster, Fili Moala, and formerly Raheem Brock) aren’t horrible, they generally do not force a double team, meaning that they can be controlled by the 2 OGs, leaving the center, and the slightly delayed tackles to push forward and provide lead blocking, easily controlling the LBs and giving the RB an easy four and a half to five yard pickup before finally being picked off by one of the safeties or Brackett moving forward from a deeper zone.

This all sounds fairly hypothetical, but in reality a number of stats help illuminate this phenomenon.

According to Football Outsiders, while the Indianapolis Colts were universally poor against the run across the board, the Colts were a staggering +1.28 combined yards (0.7 at left tackle, and 0.58 at right tackle) worse at runs between the OG and OT as compared to the league average.  The only teams to have a greater disparity from the league average in this gap (Miami at 1.62, New Orleans at 1.32 and Denver at 1.31) all had fewer adjusted line yards overall than the Colts (Colts had 4.58, Miami had 4.33, New Orleans had 4.28, and Denver had 4.13).  And, this ends up not being a simple statistical fluke due to small sample sizes.

As it turns out, teams tended to rush that gap (between the OT and OG) 31% of the time compared to a league average of 28%.  To top it off, opposing teams ran fewer stretch plays against the Colts (17%) than the league average (22%).  While statistically the Colts were worse than league average against stretch plays, they were more rigid, and more intimidating in those situations than when they were defending rushes down the tackle gap.

While all this may be coincidental, it is possible to infer that there is something more than meets the eye with this specific case.  In fact, by looking at stats over the past five years, we see that when looking at the difference in production between the Colts defensive line and the league average, the Colts tend to give up an extra +0.71 combined yards per year on running plays between the OT and OG, while having a startling -0.5 combined yards per year on stretch plays.  And on top of all of this, we also end up finding out that on a year by year basis, the Colts on average have run plays focused on the OT/OG gap about -0.5% from league average.

This doesn’t seem very dramatic until that number is compared to the whopping -4.8% difference from league average on stretch plays run.  While a number of inferences can be made about this particular tidbit, the one I want to highlight is that while being rushed at basically an average amount, the Colts have given up consistently larger chunks of yards on a yearly basis in the OT/OG gap, while not only consistently shutting down stretch plays for less yardage than the league average, but did so while intimidating opposing offenses into not even trying!  Plenty of factors go into these numbers, but one which should not be overlooked is the presence of Freeney and Mathis as threats outside of the tackles.

Still, all this is merely anecdotal and is by no means a “nail in the coffin,” but it is something to pay attention too.  As a byproduct of a project I am working on, I hope to shed more light on this curious case, and hopefully provide greater empirical evidence of this type of occurrence actually taking place.

For now though, it serves as simply a philosophical oddity worth examining further, but regardless, I must provide a simple disclaimer.  While I, as a writer, may be attempting to deliver an argument as unbiased from personal feelings as possible, eventually opinion finds a way of bleeding through.

It is not my opinion, though, that the detriment to the defensive unit as a whole as a result of this fluke is great enough to warrant such drastic actions as to replace such amazing players as Freeney and Mathis for lesser pass rushing talents.  If, indeed, this particular happenstance ends up being more widely valid than a hypothetical instance and some basic anecdotal evidence, other solutions may very well end up being significantly more beneficial than replacing Freeney and Mathis.

One such solution may be to increase the frequency of blitzing on suspected run plays, especially when both Freeney and Mathis are on the field.  This increase in blitzing, though, would have to pass the risk/reward test whereby the benefit for plugging holes at the LOS would be balanced with the risk created by opening larger gaps in the Zone scheme if the play ends up being a pass instead of a run.  With Freeney and Mathis providing an unrivaled pass rush combination, though, odds are the reward would well be worth the risk.

Regardless, it will be interesting to see which potentially innovative solutions arise from this look at one particular aspect of the signature defensive scheme of the Indianapolis Colts.  While this one specific type of wrinkle in the system is not to blame for the overall poor performance against the run by the Colts, it will be interesting to see if second year defensive coordinator Larry Coyer adds a new twist on the relatively pure Dungy/Kiffin/Noll Tampa-2 defense.