Shake, Rattle & Roll: Dumervil's Chase, The Importance of Sacks & Keeping The King in The Building

Published Dec. 5, 2009 at 8:30 p.m.
...

...More photos »...by Ed Andrieski - AP...

...

...3 months ago:...Chicago Bears quarterback Jay Cutler (6) scrambles out of the pocket to avoid the rush of Denver Broncos' Elvis Dumervil (92) during the first quarter of an NFL preseason football game in Denver on Sunday, Aug. 30, 2009. (AP Photo/Ed Andrieski )...

...

Browse more photos »

It's good to be the king--of sacks.

That's because sacks are critical in today's NFL.  Sacks have become such an important part of football, that today they are cause for orgasmic euphoria on the part of players and fans.  It seems that this single act of taking the opposing quarterback down two-to-ten yards behind the line of scrimmage is almost on par with touchdowns themselves. Gyrating, break-dancing, and in the case of Shawn Merriman, outright seizures disguised as celebrations are in order after just a single sack.

Deacon Jones might have fathered the sack.  Lawrence Taylor might have raised it through adolescence.  And Michael Strahan might have helped it pay for college.  But today, every roster is filled with a least one player trying to make a real man out of the sack.

That's because some sacks are hellaciously important.  Not because Ryan Clady makes a lot of money defending against them. Not because Michael Lewis says so.  But because a sack has real value.

So much value, in fact, that the Denver Broncos should resign Elvis Dumervil--before the king leaves the building.

 

The Temple of "Doom"

Let's start by stating the obvious.  Elvis Dumervil is taking a run at the record book in 2009.  He's going through left tackles like Al Davis goes through 1st-round draft picks.  Dumervil is not only flourishing under the Mike Nolan 3-4, he's getting fat on the hog.  With 14 sacks, there's an outside chance he could reach Strahan's single season record of 22.5.  

With 5 games remaining, Dumervil needs to average 1.7 sacks/game.  While this may appear like a lot, it's certainly within the realm of  possibility.  Dumervil is already averaging 1.2 sacks/game, so he'd need to amp up his current production slightly.  But he's not miles behind Strahan's pace.  Let's compare Strahan's 2001 season with Dumvervil's 2009 season through 12 weeks.    Keep in mind Strahan had a week-12 bye and Dumervil a week-7 bye.  

Strahan was at 15.5 with 5 games remaining, and averaged 1.4 sacks during that last stretch of games to set the bar at 22.5.  Can Elvis get on a hot streak?   The numbers suggest it's plausible.  Here are the sacks/game average of his remaining opponents:

• Week 13: Kansas City (3.45 sacks/game)
• Week 14: Indianapolis (.909 sacks/game)
• Week 15: Oakland (2.7 sacks/game)
• Week 16: Philadelphia (2.45 sacks/game)
• Week 17: Kansas city (3.45 sacks/game)

The Colts don't allow many sacks, but every other team on the schedule could generate a multi-sack game for Dumervil.  Given that Dumvervil is collecting 44% of his team's current sack total, we could take a overly-generalized and wildly simple stab at his potential sack number through the next 5 games:

• Week 13: 3.45 X 44% = 1.52 sacks
• Week 14: .909 X 44% =.04 sacks
• Week 15: 2.7 X 44% = 1.2 sacks
• Week 16: 2.45 X 44% = 1.08 sacks
• Week 17: 3.45 X 44% = 1.52 sacks
• Total Sacks = 5.72 + 14 (current) = 19.72 sacks

We might also simply take his average sacks per game and multiply them by 5 games.  Using this method, he would end up with 20.36.  Both methods are crude guesses, but no matter how you slice the numbers, it's clear that Dumervil is going to have to increase his production even more over the next five weeks and have.  And there's no better way that to increase a player's production than taking on the Chiefs (twice) and the Raiders.    

The Value of Sacks

Does all of this really matter or is it just a way for Dumervil to pump up his value for free-agency?   Are sacks really that important as Denver makes its playoff push?  Are sacks overrated?   Just how valuable is a sack?  And by default, how valuable is Dumervil?

Statistically, a team's sack total has a moderately-strong correlation to winning.  Over the last ten seasons 1999-2008, the correlation coefficient (measure of the strength of the relationship of two variables on a scale of -1 to 1) of a team's sack total to team wins has been .4956.  To give you an idea of how this type of correlation ranks with respect to other stats, consider the correlations of my weekly Stats That Don't Lie to team wins over the same time period:

Field Position Differential: .489

Time of Possession:.554

3rd - Down Efficiency: .511

Giveaways: -.486

For those that don't want to be bothered by what this means, simply know that there is a moderate to moderately strong relationship between sacks and wins (say three times fast: correlation is not causation).  For those that like this stuff, we know we can square this number and postulate that sacks can explain/account for about 25% of a team's wins over a season.   Either way, one would say that sacks have a moderate correlation to winning.  

Brian Burke (the best football-stats man going) over at Advanced NFL Stats actually did me one better (this is not exceedingly difficult).  Instead of calculating the correlation of stats to wins, he calculated the value of a sack itself.  His findings?  A sack has a expected point potential of 2 points:

If we average the expected points of all situations in which there wasn't a sack, and compare it with the average expected points following plays that did result in a sack, we get a difference of 2.0 points. In effect, a sack swings the balance of the game by an average of 2 points in favor of the defense, either by forcing a punt or a longer FG try, or even just putting a team in a predictable passing situation. That's a big swing for a single play. A turnover is generally worth 4 points, so a sack could be thought of as half as good as a fumble or interception.

Burke is right.  This is a huge swing for one play.  Two sacks in one game and you've got the equivalent of an interception or fumble.  Three-and-a-half sacks, and your team is getting the equivalent of a touchdown, which was the case when Dumervil had four sacks in one game against Cleveland earlier in the year.   It's little wonder then that there's a moderately-strong correlation between winning and sacks. 

This allows us to put an approximate point value on Dumervil's sack numbers thus far this year.  28 points.  That's quite a contribution from the 4th round draft pick and pint-sized Defensive End.  In a season in which the Broncos have scored 196 points, that's a potential whopping 14% of the team's total.

Let's be careful, however.  I'm not saying that Dumvervil has actually contributed 28 of Denver's 196 points.  We are dealing with averages over time and point potentials.  And many variables go into a single pass rush, specifically, coverage and other actions of all of the defensive linemen as a unit.  Just last week against the Giants, for example, Vonnie Holliday stunted and took what amounted to three blockers out of the play so that Dumervil could attack up the middle for the sack of Eli Manning.   However, the point is sound.  Sacks--and Elvis Dumervil--have high value.  

Burke's calculation is based the average for all sacks.  But we also know intuitively, and without bull-rushing the point with more stats, that not all sacks are created equal.   A few points of consideration:

• A sack on 3rd or 4th down is more valuable than a sack on 1st or 2nd down.   This makes sense if you think about each down as a missed opportunity.  On 1st down, if my quarterback is sacked and loses some yards, I've still got options--two to three more downs--to make up for the error. On 3rd (and occasionally 4th) down, a quarterback sack essentially ends the drive right there.  There are no more options.  I'm punting, or if I'm close enough to my opponent's goal line, I'm going for a field goal.   Thus, the sack on 3rd down puts and end to more chances.

• A sack is more valuable than a stop because of the loss of yards and the inherent potential for a turnover. Consider the turnover risk to an offense, if on 3rd-and-3, I run right up the middle for 1 yard.  I punt the ball and the other team gets a shot.  There is more risk involved in dropping back on 3rd-and-3 and having a guy like Elvis Dumervil strip the ball from behind.  And if that doesn't happen, at a minimum, my field  position will have declined to a greater magnitude that a simple no-gainer.

• Like turnovers, a sack later in the game is mover valuable than a sack early in the game.  A sack on the 1st play of the game is going to be less valuable to my team than a sack on the last play of the game when the other team has a chance for a game winning pass.  This is because in the first example, I've got dozens of drives and almost 4 full quarters to overcome the effect of the sack.  On the second, I've no opportunity.  

• A sack when the game is in question is more valuable than a sack win winning easily.   When my team is up by 30 points, a sack is going to be less meaningful than a sack when the game's outcome is still in question.  While compiling a couple of extra sacks in scrub time might pad my stats and kick in some incentives in my contract, they are not nearly as valuable to my team.  

• A sack in my (the defense's) territory is more valuable than a sack in my opponent's (the offense's) territory.  This has to do with the potential for points.  As my opponent gets closer to my goal line, the chances of my opponent scoring increase.  So a sack at my 40-yard line that takes my opponent out of field goal range is going to be more valuable than a sack when my opponent is at his own 30-yard line.  The notable exception to this is when my opponent is backed up against his own goal line.  In these situations, the possibility of a safety exists, and the point potential (for the defense) increases.

All five of these points may or may not seem logical to you and I, but the NFL doesn't see it this way.  A sack is a sack--all of them, created equal. 

Dumervil's Body of Work

If we accept our intuition as fact--and in this case it's wise to do so--we can begin to get our hands around the situational value of each of Dumervil's sacks. 

Sack # 1 - Cleveland Browns, Week 2

Dumervil's first sack of the season was interesting to say the least.  It came on the first play of the 3rd quarter.  I would classify it as a coverage sack.  However, Dumervil did make a nice power move to the inside after coming off the edge.  The sack came out of the a base 5-2 look and it occurred on 1st-and-10 at Cleveland's 23-yard line.  At the time, Denver was up only by 4 points, so the game was in question.

Sack # 2  - Cleveland Browns, Week 2

This sack came out of Denver's nickel package and Dumveril simply flew  past John Saint Clair for the punishing sack of Brady Quinn.  This was all Dumervil.  The sack came with 11:19 remaining in the 4th quarter on a 2nd-and-17 at the Cleveland 13-yard line.  Denver was up by 14 points

Sack # 3 - Cleveland Browns, Week 2

On the next play (3rd-and-22), Dumervil did the exact same thing to Saint Clair, going around him on the edge.  The coverage was also excellent.  Denver was again in its nickel package and the sack occurred at the 1-yard line and probably could have been called a safety.

Sack # 4 - Cleveland Browns, Week 2

This came with 7:29 remaining in the 4th quarter on a 3rd-and-1 at the Cleveland 29-yard line.  Dumervil again abused Saint Clair with the outside speed rush in the nickel package.  Denver was up by 21, however, so there was little doubt as to the outcome of the game.

Sack # 5 - Oakland, Week 3

This sack came with 7:16 in the 3rd quarter on a 3rd-and 10 at the Denver 45-yard line.  Dumvervil bull rushed and then spun to the inside to get the sack, but this also had to do a lot with Denver's excellent coverage.  Also, Denver was up by 17 points at the time.  The came was still in question, but barely.

Sack # 6 - Oakland, Week 3

The sack came with 9:50 remaining in the 4th quarter on a 3rd-and-7 at the Oakland 14-yard line.  Dumvervil starts on the outside and then stunts towards the middle and it untouched for the sack.  This play essentially pins Oakland against it's goal line and they are forced to punt.   With Denver up by 20, however, the game was not in question.

Sack # 7 - Dallas Cowboys, Week 4

The sack came with11:20 remaining in the 1st quarter on a 3rd-and-9 at the Dallas 40-yard line.  Dumervil does his classic edge-rush-stop-on-a-dime-then-spin-to-the-inside move against Flozell Adams.  Admittedly, the coverage helped him out.  The game was very much in question, however, since the score was 0-0 and the Cowboys were force to punt as they were about to enter Denver's territory.

Sack # 8 - Dallas Cowboys, Week 4

The sack came with 7:37 remaining in the 3rd quarter on a 1st-and-10 at the Dallas 31-yard line.  This is a coverage sack (out of a base 5-2 formation), but Dumvervil essentially gets under the pads of Adams and just pushes him back into the quarterback. Dallas was ahead by 3 at the time.

Sack # 9 -  San Diego Chargers, Week 6

The sack came with 1:52 remaining in the

Back | Read more at Mile High Report-Denver Blog