Previous post:

Next post:

The greatest pass rushers of the last 30 years

by Chase Stuart on July 11, 2012

in Defense, G.O.A.T., Statistics

Thirty years ago, the NFL began officially recording defensive player sacks. Prior to 1982, all teams kept their own individual sack data, but those records (with few exceptions) have never been verified. As a result, it’s an unfortunate reality that for much of NFL history, we simply do not have reliable sack data for individual defensive players.

Three times, Deacon Jones produced 20+ unofficial sacks in the 1960s.1 In 1967, Raiders defensive end Ben Davidson had 17 sacks2 in the AFL. Jack Youngblood and Jim Katcavage both led the league in sacks on two different occasions in the pre-1982 era.3 Cincinnati Bengal Coy Bacon has been credited with 21.5 unofficial sacks during in 1976. The first team to record 60 sacks in a season was the ’57 Bears, and we can be sure that Doug Atkins recorded more than his fair share of that number. For players like Gino Marchetti, Norm Willey, and Len Ford, even unofficial records weren’t kept during their time, leaving us unsure as to who is the true sack king.

It’s important to remember that just because we don’t have official sack data before 1982 doesn’t mean there were great sack artists before then. But that’s a topic for another today. So while we can’t precisely measure how the forefathers of the game played, we do have official data for the last 30 years. So who has been the best pass rusher of the last three decades?

Brett, are you SURE you're okay?

Using total sacks isn’t a fair method to current players, or to those players who chose to retire instead of sticking around to compile six-sack seasons. So if we want to measure sack dominance, we can’t simply look at total sacks any more than we can grade running backs by looking at career rushing yards. One method I like that I’ve used before is sacks over one-half sack per game. This makes 8 sacks in a 16-game season the bar; a player only gets credit for their production over that level. This means that 12 sacks in a 16-game season brings a value of +4.00, while 16 sacks is twice as valuable at +8.00.There’s no great reason to choose 8 over 6 or 10 or any other number. I chose 8 because it feels right, but I don’t claim that it’s based on anything other than my personal, subjective preference.

In addition, there are two adjustments that I think help level the playing field. The first is an era adjustment, not unlike what you see when discussing wide receivers — except this one goes the other way. In 1984, the league sack rate was at 9.2%, but in 2009, it was just 6.5%. In other words, it was significantly easier for defenders to record sacks — on a per pass attempt basis — in 1984 than it was 25 years later. The other is to take into account the number of pass attempts each team faced. In 1984, Rulon Jones recorded 11 sacks for the Broncos, but Denver faced an incredible 688 pass attempts that season. Compare that to Jason Taylor, who had 11 sacks in 2007 for the Dolphins. That season, Miami faced just 439 pass attempts. I don’t think it’s controversial to say that Taylor’s sack total was much more impressive. As a result, I’ve adjusted each player’s “sacks over 0.5 sacks/game” metric for era and based on the number of attempts his team faced.

Here are the single-season leaders for each of the last 30 years. In addition to listing the player’s sack totals, I’ve listed the number of pass attempts their teams faced per game, the league sack rate that season, the number of sacks over 0.5 games played, and then an adjusted score, which is how the table is sorted.

Year
Team
Player
G
SK
PA/G
LgSkRt
SkOv0.5
ADJ Score
2001NYGMichael Strahan1622.535.47.414.514.41
2011MINJared Allen162236.86.81414.27
1987PHIReggie White122141.28.21513.17
2008DALDeMarcus Ware162035.46.31213.11
1989MINChris Doleman162134.97.71312.7
1984NYJMark Gastineau162234.79.21412.14
1989GNBTim Harris1619.531.97.711.511.91
1990KANDerrick Thomas152035.87.912.511.81
2011DALDeMarcus Ware1619.536.76.811.511.79
1990BUFBruce Smith161931.17.91111.41
2003NYGMichael Strahan1618.535.36.610.511.3
2002MIAJason Taylor1618.535.46.810.511.06
2006SDGShawne Merriman121737.47.11110.85
2011PHIJason Babin161835.56.81010.54
1983NYJMark Gastineau161931.98.61110.53
2008MIAJoey Porter1617.536.96.39.510.32
1986NYGLawrence Taylor1620.540.48.312.510.14
2009DENElvis Dumervil161734.36.5910.11
1995BUFBryce Paup1517.539.46.41010.2
2007KANJared Allen1415.531.26.58.510.02
1989MINKeith Millard161834.97.7109.7
1992PHIClyde Simmons161935.88.5119.6
1991NORPat Swilling161733.87.199.59
2008ATLJohn Abraham1616.536.46.38.59.42
2008PITJames Harrison151636.56.38.59.34
1988PHIReggie White161838.87.3109.25
1986WASDexter Manley1618.536.78.310.59.09
1988NYGLawrence Taylor1215.538.67.39.58.97
1999ARISimeon Rice1616.532.97.58.58.88
2000NORLa'Roi Glover161734.67.598.98
2005OAKDerrick Burgess161632.67.288.75
1992SDGLeslie O'Neal151733.98.59.58.66
1986PHIReggie White161836.68.3108.62
1984CHIRichard Dent1617.531.79.29.58.49
2003TAMSimeon Rice161531.96.678.5
1984NWEAndre Tippett1618.535.59.210.58.41
2002TAMSimeon Rice1615.534.66.87.58.23
2010DALDeMarcus Ware1615.535.96.57.58.29
2011NYGJason Pierre-Paul1616.539.86.88.58.14
2000MIATrace Armstrong1616.536.17.58.58.14
1991PHIReggie White161532.67.177.85
2006GNBAaron Kampman1615.535.17.17.57.8
1983SFOFred Dean1617.536.48.69.57.83
2001STLLeonard Little1314.536.67.487.7
2003MIAAdewale Ogunleye161535.86.677.7
1988RAMKevin Greene1616.539.27.38.57.66
1999STLKevin Carter161740.87.597.55
1992DENSimon Fletcher1616328.587.6
2008MINJared Allen1614.535.96.36.57.51
2004INDDwight Freeney161637.67.387.52

Lawrence Taylor in ’86 and Elvis Dumervil in 2009 provide good examples of how this system works. On the surface, Taylor’s 20.5 sacks looks well, 3.5 sacks better than Dumervil’s 17 quarterback takedowns. But we need to consider that from ’82 to ’11, the average sack rate was 7.43% and the average defense faced 34.84 pass attempts per game. Taylor in ’86 was fortunate enough to face 16% more pass attempts than the average player and he played in an era that was 11% more sack-friendly than the average player. Combined, that put him in an environment that was 29% more favorable for achieving sacks than the average player. Therefore, instead of calculating sacks over 0.50 games played for Taylor in ’86, we calculate his sacks over 0.65 games played, which represents a 29% increase in the size of the denominator. His 20.5 sacks therefore translates to 10.1 sacks over 0.65 games played.

Dumverill, however, played in an era where sack rates were just 87% of what they have been, on average, over the last three decades. Considering he also faced slightly fewer pass attempts per game, too, he will get credit for sacks over 0.43 games played. His 17 sacks means he also gets credited with a score of 10.1, because he gained 10.1 sacks over 0.43 games played.

We can also create career lists based on how each player performed in each season since 1982. I’ve eliminated all seasons where a player fell below the baseline, i.e., a season of fewer than 0.5 (adjusted) sacks per game played. But if we total all other seasons, we can get their career total of sacks over X games played.

Rank
Player
Year
Sk
Gms
Grade
1Reggie White1985--200019823274.9
2Bruce Smith1985--200320027967.6
3Kevin Greene1986--199916021356.3
4Simeon Rice1996--200712216849.4
5Jared Allen2004--201110512547.5
6DeMarcus Ware2005--201199.511247.3
7Michael Strahan1993--2007141.521646.2
8Derrick Thomas1989--1999126.516945.1
9Chris Doleman1985--1999150.523244.9
10Jason Taylor1997--2011139.523343.4
11Lawrence Taylor1982--1993132.516842.5
12Leslie O'Neal1986--1999132.519641
13Richard Dent1983--1997137.520040.9
14John Abraham2000--201111215940.6
15Dwight Freeney2002--2011102.514938.4
16John Randle1990--2003137.521934.4
17Leonard Little1998--200987.514131.8
18Julius Peppers2002--201110015431.5
19Mark Gastineau1982--1988748931
20Pat Swilling1986--1998107.518529.8
21Dexter Manley1982--199197.512327.7
22Simon Fletcher1985--199597.517227.6
23Andre Tippett1983--199310014227.3
24Robert Mathis2003--201183.513527.3
25Clyde Simmons1986--2000121.523627.2
26Tim Harris1986--19958111826.6
27Hugh Douglas1995--20048013825.8
28Neil Smith1988--1999104.518125.7
29William Fuller1986--1998100.519425.5
30Jacob Green1982--199197.514825
31Charles Haley1986--1999100.516925
32Rickey Jackson1982--199512821124.2
33Greg Townsend1983--1994109.518623.2
34Sean Jones1984--199611320122.1
35Elvis Dumervil2006--201152.57521.5
36Trace Armstrong1989--200310621121.5
37Kabeer Gbaja-Biamila2000--200874.512421.4
38Patrick Kerney1999--200982.515921.2
39Robert Porcher1992--200395.518721.1
40Osi Umenyiora2003--20116911320.8
41Aaron Schobel2001--20097813320.7
42Wayne Martin1989--199982.517120.6
43Warren Sapp1995--200796.519820.5
44Joey Porter1999--20119818820.4
45Trent Cole2005--20116810820
46Terrell Suggs2003--201182.514120
47Mario Williams2006--2011538219.8
48James Harrison2004--20115810619.6
49Lee Williams1984--199382.514019.5
50Michael Sinclair1992--200173.514419.5

Kevin Greene teaches Clay Matthews how to make those golden locks shine

It’s not too surprising to see Reggie White and Bruce Smith on top of the list. They’re the best two defensive ends of the official sack era, and also rank 1-2 in total sacks. Kevin Greene was a star 3-4 outside linebacker in Los Angeles, Pittsburgh, Carolina and San Francisco; how many people know that he’s third on the official sack list? Greene wasn’t really a “compiler” either, although he did seem to play forever. He finished his career with double digit sacks in each of his last four seasons, and recorded 12 sacks for the Panthers in 1999 at the age of 37; in fact, he’s responsible for two of the five seasons of 12+ sacks by a player over the age of 35. Greene had a reputation for being extremely one dimensional, but few excelled at one dimension for as long as he did.

The next names have a more modern flair, and I’ll be honest, caused some head scratching on my part. Rice is an interesting guy: he was basically done after his tenth season, recording only three more sacks after turning 32. And he spent the first five seasons on miserable Cardinals teams, that had two borderline Hall of Famers on defense and little else on the roster. But Rice had a dominant prime: three times from ’99 to ’04 he ranked 2nd in the league in sacks. From ’98 to ’05, he was the only player to record 100 sacks. But what’s interesting about Rice’s production is that among elite pass rushers, no player faced fewer pass attempts.

Simeon Rice got to the quarterback on the rare occasion when his opponents passed.

From ’01 to ’06, the Bucs ranked in the bottom 8 in pass attempts faced each season. During his Cardinals days, Arizona ranked in the bottom five in pass attempts faced three different times. Rice doesn’t have extremely impressive career totals, but he’s one of just 8 players to record double digit sacks in 8 seasons since ’82. Considering the low number of pass attempts he faced, that production was enough to vault him into the fourth spot on the list. In that

Jared Allen and DeMarcus Ware slot behind Rice, and both just completed dominant seasons in 2011. I wrote a lot about Ware yesterday; as for Allen, he’s become the best pass rushing defensive end of this era. Both are at a “sack per game” pace over the last five years, and no one else is particularly close. Because of the lower sack rates in today’s NFL, Ware and Allen have vaulted ahead of some players who produced more sacks in more sack friendly environments. I guess that makes him the pass rushing version of a Paul Warfield, a wide receiver who played on run-oriented teams during a run-heavy era.

Comparing Derrick Thomas and Lawrence Taylor is interesting. Remember, the above table ignores the 9.5 unofficial sacks Taylor had his rookie season. Leaving out Taylor’s rookie year, the two played in a nearly identical number of games, and Taylor holds a slight edge in sacks. But Taylor faced more pass attempts and played the prime of his career during a more sack-friendly era; as a result, Thomas vaults him on the list.

I’ll close today with one more table, showing the top 50 pass rushers from the above table. The below focuses just on the adjustments you might want to make for each player. For example, let’s look at the second row of the final table. Mario Williams has a rank of 47, which is where he ranked on the top 50 pass rushers list above. He had 53 sacks from 2006 to 2011, and has a career total of 15.5 sacks over 0.5 games played (this is calculated per season). He has a weighted average in the “Attempts” column of 1.00, which means that he’s faced a perfectly average number of pass attempts (Simeon Rice is at 0.94, which means he’s faced only 94% as many pass attempts as the average pass rusher). Williams has a 0.87 in the ERA column, which means that in his era, sacks were only 87% as common as they were over the average of the last 30 years. The final column is the combined adjustment, which simply multiplies the “Attempts” and “Era” adjustments together. On the bottom of the list, you see players who dominated in the mid-1980s, when sack rates were north of eight percent. This table is a handy guide to understanding which players were helped and hurt most by factors largely out of their control.

Rank
Master
Year
SKs
SkOv0.5
ATT
ERA
COMB
35Elvis Dumervil2006--201152.5180.970.890.86
47Mario Williams2006--20115315.510.870.87
24Robert Mathis2003--201183.520.50.990.910.9
4Simeon Rice1996--2007122430.940.960.9
5Jared Allen2004--2011105431.010.890.91
41Aaron Schobel2001--200978160.980.930.91
14John Abraham2000--20111123510.910.92
48James Harrison2004--201158171.060.870.92
17Leonard Little1998--200987.5260.970.950.92
15Dwight Freeney2002--2011102.531.510.920.92
6DeMarcus Ware2005--201199.543.51.050.880.92
45Trent Cole2005--20116816.51.050.880.93
29William Fuller1986--1998100.522.51.010.930.93
10Jason Taylor1997--2011139.537.50.990.940.94
44Joey Porter1999--201198191.040.90.94
18Julius Peppers2002--201110028.51.030.910.94
46Terrell Suggs2003--201182.5171.040.920.96
38Patrick Kerney1999--200982.519.51.040.930.96
27Hugh Douglas1995--2004802410.980.97
28Neil Smith1988--1999104.52510.990.99
40Osi Umenyiora2003--201169201.090.921
8Derrick Thomas1989--1999126.5430.971.031
37Kabeer Gbaja-Biamila2000--200874.519.51.060.951.01
22Simon Fletcher1985--199597.5270.971.041.01
36Trace Armstrong1989--200310620.51.0111.02
42Wayne Martin1989--199982.5201.0211.02
33Greg Townsend1983--1994109.5260.971.061.03
2Bruce Smith1985--20032006911.031.03
26Tim Harris1986--199581270.961.061.03
7Michael Strahan1993--2007141.5471.050.981.03
12Leslie O'Neal1986--1999132.5421.011.031.04
39Robert Porcher1992--200395.5211.021.021.04
31Charles Haley1986--1999100.5271.050.991.04
16John Randle1990--2003137.537.51.060.991.04
20Pat Swilling1986--1998107.532.51.031.011.04
3Kevin Greene1986--199916059.51.0511.05
25Clyde Simmons1986--2000121.52711.051.05
34Sean Jones1984--199611324.51.031.031.06
9Chris Doleman1985--1999150.5481.021.051.07
49Lee Williams1984--199382.5220.991.081.08
43Warren Sapp1995--200796.5201.071.011.08
32Rickey Jackson1982--1995128300.991.11.09
50Michael Sinclair1992--200173.5211.071.011.09
13Richard Dent1983--1997137.5471.011.111.12
1Reggie White1985--200019884.51.071.061.13
19Mark Gastineau1982--19887435.50.971.181.14
30Jacob Green1982--199197.531.511.151.15
11Lawrence Taylor1982--1993132.550.51.061.11.16
23Andre Tippett1983--1993100321.021.171.19
21Dexter Manley1982--199197.536.51.041.151.2
  1. According to research done by John Turney. []
  2. According to Nate Webster. []
  3. Source: Turney/Webster []

{ 11 comments… read them below or add one }

Danish August 23, 2012 at 7:58 am

Lovely excersize in era-adjustment.

Reply

sam mandonald February 12, 2013 at 12:46 am

To the author – In all due respect, this analysis is garbage. Your era adjustment tinkering assumes that the higher sack rates in the 1980′s were the product of poor offensive lines, ignoring the possibility that maybe sack rates in the 1980′s were higher because there were better pass rushers back then.
But the real failure in your analysis is the “pass attempt adjustment” you apply to all defensive players. The stat would make sense if you were evaluating just down linemen, but your list includes linebackers.
Linebackers don’t rush the passer on every pass attempt.
You willfully ignore the fact that todays pass rushing outside linebackers rush the QB far more often than OLBs of the past.
You don’t bother to do a pass rush to coverage breakdown for the linebackers on this list. Instead you assume they all rushed the QB the same percent of the time! Why?
Take Demarcus Ware and Lawrence Taylor.
In LT’s first 5 years (1981 -1985) he rushed the passer just 55% of the pass attempts he was on the field for.
In 1986 (the year he recorded 20.5 sacks) Taylor dropped into coverage on 242 passing downs. That’s over 1/3 of all QB drop backs!
You know how many times Demarcus Ware dropped into coverage in 2012? Answer – 64 times…. on 518 passing plays (12%)
In other words, Lawrence Taylor dropped in coverage on passing plays 3X more than Demarcus Ware…. yet your analysis makes him seem like the more effective pass rusher. Sloppy work my friend. I expect better from this site

Reply

Chase Stuart February 12, 2013 at 9:57 am

Sam,

Where did you get those numbers for Taylor?

Reply

sam mandonald February 12, 2013 at 3:38 pm

I have every NY Giants game Lawrence Taylor played in (with the exception of 5 games from the ’92 season) on disc.
I’m sure you’re skeptical of this claim…. and I don’t blame you.
Do a youtube search using the keywords “lawrence taylor forced fumbles”. That’s my work.
Regarding his defensive snap breakdowns from each season, I have that on spreadsheet.
What can I say, I’m a stats geek

Reply

Chase Stuart February 12, 2013 at 4:10 pm

I agree that ideally, we would want to credit pass rushers for their sacks per actual pass rush, and exclude pass plays in coverage form pass attempts. There are other factors to look at (for example, if in Dallas’ 3-4 defense, Ware was always rushing but Dallas generally still has 4-man rushes, while in NY’s defense, Taylor generally rushed with another linebacker, meaning the Giants had a five-man rush), but all else being equal, I agree with you.

That said, any idea how to analyze not just Lawrence Taylor, but every pass rusher over the last 30 years?

Reply

sam mandonald February 12, 2013 at 9:06 pm

The giants rarely rushed more than 4 guys during LT’s career. Parcells’ 3-4 defense was very conservative.
Belichik rarely dialed up blitzes (I define a blitz as sending more than 4)
By my count, Taylor recorded 101 sacks (playoffs included) on 4 man pass rushes.
He recorded an additional 6 sacks on 3 man pass rushes.
Of the 101 sacks on 4 man pass rushes, 36 were against 5 man pass protection schemes, 51 against 6 man pass protection schemes, and 14 against max protection (7 or 8 pass blockers)
Happy to upload the footage if you don’t believe me.
Re: how to analyze a players pass rush productivity – I don’t think you can for retired players. There was no Pro Football Focus or Advanced NFL Stats back then. Data like QB hits, pressures, the time from snap to sack, the pass rushing and blocking schemes for each sack, etc etc…. none of it was recorded.
I suspect that had all that data been recorded, Reggie White would be atop the list. I’ve never seen a pass rusher like him.

Reply

Chase Stuart February 12, 2013 at 10:33 pm

Thanks sam. Good stuff. I agree that it’s close to hopeless to really analyze older defensive players, sadly.

Reply

sam mandonald February 12, 2013 at 9:11 pm

ps – The most under appreciated defensive player on those 1980′s Giant teams was big Leonard Marshall. He was a big help to LT.

Reply

Chase Stuart February 13, 2013 at 11:41 am

Agreed — I talked about Marshall in the post immediately before this one. http://www.footballperspective.com/random-perspective-on-the-2012-dallas-cowboys/

Reply

BLACKnBLUEnGOLD March 14, 2013 at 2:24 pm

Regarding era adjustment — the difference in sacks per pass attempt are likely related in some part to offensive skill. The rules about pass blocking changed in 1978. While pass blocking to the letter of the post-1978 rules with minimal experience is still more effective than pass blocking to the letter of the pre-1978 rules with a lot of experience, pass blocking today is undoubtedly better. The coaches aren’t teaching brand-new technique, and the linemen haven’t learned a completely new technique in mid-career.

However, I think the biggest factors are scheme-related. Lawrence Taylor ushered in an era of pin-your-ears-back blind side speed rushers, who were different from the Youngbloods and Deacon Joneses of the world. The old guard, and their old-school coaches, wouldn’t play you if you weren’t a strong run defender. Back then, there was no “play the run on your way to the quarterback” like LT and Bruce Smith did. Offenses had not yet invented schemes to minimize the impact of this new type of player. Pass protection schemes were rudimentary, and passing plays took a lot longer to develop.

Statheads tend to forget about blocking, and scheme, and particularly blocking scheme. It’s hard to blame them, because it’s not as conducive to statistical analysis. However, those things are real, and in this case, they had a statistical effect that was measurable on the large scale.

Reply

sam mandonald March 14, 2013 at 6:52 pm

Re your first point about the 1978 blocking changes – I don’t doubt there was a learning curve for O-Linemen in mastering new techniques. However the same would apply for defensive players who also had to learn new techniques to combat the more lenient blocking rules. Also, since the rule change was implemented to increase scoring by making blocking easier, I think the transition was likely more difficult for the defensive players than the offensive ones, Yes?

On your 2nd point – I can’t speak for Bruce Smith’s run defending capabilities (don’t have enough game tape of him to have an informed opinion), but I can speak to Taylor’s run defense- it was phenomenal. He played the run first & the pass second. I’ve always believed the most under appreciated aspect of LT’s game was his ability to defend the run. None of todays 3-4 pass rushing OLB come close to recording 100 tackles a year. LT did it on a yearly basis from 1981-88.

On your third point – I completely agree. Football stats don’t provide much insight into a player, especially for the defensive guys. There’s just too many variables at work. e.g. …
Player A gets chipped by the TE off the snap, double teamed by the tackle & running back, and records a sack for a 7 yrd loss.
Player B is untouched off the snap & has a free path to the QB and records a sack for a 7 yrd loss.
The stats make no distinction b/w the 2 plays…. but they should

Reply

Leave a Comment

You can use these HTML tags and attributes: <strong> <em> <pre> <code> <a href="" title="">

{ 1 trackback }