Bill Connelly takes a look at what we can learn from defensive box score stats and general rates of havoc.
18 Aug 2003
by Aaron Schatz
This is the third of a series of eight articles taking a closer look at every team in the NFL, division by division, using our new statistics such as line yards and DVOA (explained here). Don't be scared away by all the numbers -- the goal of these IN FOCUS articles is to go through all the numbers and translate the most interesting trends into actual English paragraphs for those allergic to endless tables of stats. You enjoy the stats separated by team, or just enjoy the insights of each team's commentary.
The format of these articles is explained at the beginning of the one on the AFC East. Remember, offensive numbers are better the more POSITIVE, defensive numbers are better the more NEGATIVE. Overall total is offense minus defense, so the more positive the better. Schedule strength is harder the more negative, with the hardest schedule ranked #1 and the easiest ranked #32. New players on a team are colored blue and players who have left the team are colored red.
Remember when I said that the AFC East is the most balanced division in sports? The AFC West is about the same, with three teams packed near the top of the total efficiency standings and a fourth a few spots back. Real bummer for you Chargers fans.
OUT: NFC West, AFC East
IN: NFC North, AFC North
HELPS: Passing defense in specific, all defense in general. Last year, the four hardest defensive schedules in football belonged to the four AFC West teams (New England was #5, in case you are curious). This year, it gets much easier. Because of the poor teams in the North divisions, the four AFC West teams start the year with a significant advantage in the wild card race.
HURTS: Running offense, though these will still be four of the best running games in the league.
|Plummer, Jake (ARI 02)||-22.0%||42||564||3027||-19.1%||36||-10.1||47|
|Griese, Brian (MIA 03)||7.0%||15||468||3043||11.6%||11||28.9||14|
|Plummer, Jake (ARI 02)||-8.5%||21||35||286||-10.8%||25||-4.4||25|
|Griese, Brian (MIA 03)||-5.3%||17||25||117||-0.8%||17||-2.0||19|
|Gary, Olandis (BUF 03)||-3.5%||37||147||-4.1%||-1.0|
|Gary, Olandis (BUF 03)||66.0%||21||148||51.7%||10.8|
Other important additions: S Lee Flowers (PIT), DT Daryl Gardner (WAS), OT George Foster (R1), LB Terry Pierce (R2),
Other important losses: DT Chester McGlockton (NYJ), DE Keith Washington (NYG), CB Tyrone Poole (NWE), LB Kavika Pittman (FA)
Denver clearly had one of the best running games in the league in 2002, and a well-rounded one as well. They're known for having a great offensive line and great run-blocking wide receivers. That being said, what stands out here is that 5.16 aly/carry (adjusted line yards per carry) from the left side -- that's the adjusted number, despite a penalty for the poor run defenses Denver faced in 2002. Neither LT Ephraim Salaam nor LG Ben Hamilton (who also played some center) made the Pro Bowl, and neither is really lauded as being any better than the rest of Denver's line.
The Denver running game, going left, was better than any other direction from any other team in the NFL by a large margin. Kansas City left was second, 4.72 aly/carry, and Washington right was third, 4.69 aly/carry.
Are Salaam and Hamilton really that good? Well, it's generally accepted that Clinton Portis is the best of the three Denver running backs at going to the side, and he did get by far the most carries in 2002. What happens if we break the adjusted line yard numbers down for all three backs?
|Adj. Line Yards||Left||Middle||Right|
Well, either every single Denver running back is exceptional at going to one side but not the other, or Salaam and Hamilton are the most underrated offensive linemen in the NFL. I'm guessing the latter.
Despite his great performance overall, Clinton Portis was below average on third downs. Mike Anderson also had this problem, though to a smaller extent:
|DVOA||1st Down||2nd Down||3rd Down|
Both Brian Griese and Steve Beuerlein had much more success throwing to the right side than the left. This trend showed with nearly every receiver, with Ed McCaffrey as the exception. New Denver QB Jake Plummer, on the other hand, was far superior throwing over the middle compared to the sides. This will be another interesting trend to watch in 2003; is this more a product of the team, or the quarterback?
Oddly enough, although Denver runs were evenly split between left and right, passes to running backs were thrown to the left much more often than the right.
Denver played the hardest defensive schedule in the league.
Denver's rush defense tightens up when they are losing by more than a touchdown:
Winning by 8 points or more +29% VOA
Winning or losing by less than 8 -4% VOA
Losing by 8 points or more -61% VOA
If you want to succeed passing the ball against Denver, pass on first down and stay away from the sidelines. The Denver pass defense allowed a +19% VOA on first down, but 0% VOA, or league average, on downs two through four.
Denver's passing defense also was miserable on passes in the middle of the field, with a +78% VOA, compared to a +19% VOA on passes to the left or right.*
*(Due to negative pass plays like sacks and batted passes that don't have a direction, passing ratings by direction tend more positive.)
|Cloud, Mike (NWE 03)||-22.3%||49||115||-15.4%||-11.5|
|Cloud, Mike (NWE 03)||-1.6%||9||48||-2.7%||-0.1|
Other important additions: RB Larry Johnson (R1), WLB Shawn Barber (PHI), CB Dexter McCleon (STL), DE Vonnie Holliday (GNB)
Other important losses: LB Marvcus Patton (RETIRED), DE Duane Clemons (FA)
I can throw numbers at you until the cows come home, and when it comes to the Kansas City Chiefs, they won't matter that much. The biggest question mark for the 2003 NFL season -- and the issue that will make or break the Kansas City playoff hopes -- is Priest Holmes' hip.
Priest Holmes was the #1 player in football last year, both by acclamation and by our statistics. Between rushing and receiving, he was 102 "success points" over an average back (DV+). The other Kansas City backs, combined, were 13.4 "success points" worse than an average back. And while Larry Johnson could be a capable backup, the track record of Penn State running backs isn't very good. Blair Thomas? Curtis Enis? Ki-Jana Carter?
Assuming Holmes is healthy, there are some interesting details from his amazing 2002 season. Surprisingly, Holmes was thrown only five red zone passes. That's a lot fewer than most top backs. He still had plenty of red zone touches, since his 65 carries in the red zone ranked fourth in the league behind Ricky Williams, Eddie George, and Travis Henry.
The Chiefs had two offensive linemen make the Pro Bowl, LT Willie Roaf and RG Wil Shields. Looking at the line yard numbers, you can see part of the reason Roaf is so well-regarded, but it makes you wonder about Shields and RT John Tait. Here are the adjusted line yards for Holmes and the other KC backs:
|Adj. Line Yards||Left||Middle||Right|
|Other KC backs||5.04||3.42||1.82|
Yes, DVOA really says that the Chiefs, despite going 8-8, were the fourth best team in the league last year. How the heck did they manage to miss the playoffs? Three possible issues:
The Chiefs had one of the most balanced passing attacks in the game. They passed left as often as they passed right and their DVOA rating was about the same left, right, and over the middle. They were one of the more set teams when it came to wide receiver formation, with Kennison usually on the left and Morton usually on the right.
33% of Kansas City's red zone passes went to Tony Gonzalez. Only three receivers were more commonly used by their teams as red zone targets: Moss, Moulds, and Ward.
Trent Green was much more efficient when the Chiefs were losing, with a +69% DVOA when down by more than a touchdown but +11% DVOA when winning by any amount or losing by only 7 or less.
Shocked by Green's rank as the #1 QB runner based on DVOA? We don't think of Green as mobile, but the guy can scramble when he has to, especially in long yardage situations. Green ran 19 times with 8 or more yards to go for a total of 186 yards -- almost 10 yards per carry!
Kansas City's pass defense was not strong in general, but they toughened up when it counted -- in the red zone, on third down, and in the fourth quarter.
VOA against Chiefs passing defense:
1st Down +16% VOA
2nd Down +34% VOA
3rd Down -1% VOA
Red Zone -33% VOA
Rest of Field +24% VOA
Q1 to Q3 +26% VOA
Q4 -9% VOA
The Chiefs' defensive improvements in the fourth quarter came primarily in games where they were losing or winning by a touchdown or more. In close games, the fourth quarter pass defense played closer to the Chiefs' normal level.
|Martin, Cecil (PHI 02)||-57.9%||22||126||-57.9%||-14.8|
|Ritchie, Jon (PHI 03)||-32.1%||21||66||-33.9%||-5.2|
Other important additions: DT Dana Stubblefield (SFO), CB Anthony Parker (SFO), DE Tyler Brayton (R1), WR Teyo Johnson (R2)
Other important losses: DT Sam Adams (BUF), DE Regan Upshaw (WAS), CB Tory James (CIN)
You wouldn't expect the older, veteran receivers on the Raiders to go threatening their safety over the middle much, would you? Believe it or not, the Raiders were much better on passes up the middle than on passes to the sidelines. Here are the numbers for the four main receivers on the team
Oddly enough, the receivers who don't have an advantage over the middle are the ones you would expect to be going there all the time -- the tight ends, Jolley and Williams.
Charlie Garner is the shifty one and Tyrone Wheatley is the straight-ahead guy, right? Then why is Garner better running up the middle, and Wheatley better to the sides?
|Adj. Line Yards||Left||Middle||Right|
Jerry Porter had the best rating of any player in football thrown at least 20 passes between the opponent's 40 and the end zone (+78% DVOA).
Oakland was much better defending the run to the right, usually the strong side, and not quite as good up the middle. The power of Bill Romanowski? The reason Oakland will miss Sam Adams? Part of the reason Dana Stubblefield was brought in? Probably all three. Here is VOA against the Oakland rush defense:
Left: -3% VOA
Middle: +3% VOA
Right: -24% VOA
Based on last year's statistics, the best strategy against the Oakland defense would seem to be passing on first down, running on second down, and then passing again on third down. Check out the VOA against the Oakland defense -- remember, negative numbers mean less scoring and thus better defense.
|VOA vs. OAK D||1st Down||2nd Down||3rd Down|
|Neal, Lorenzo (CIN 02)||10.2%||9||31||3.0%||0.9|
|Conway, Curtis (NYJ 03)||50.0%||7||53||71.9%||4.3|
|Conway, Curtis (NYJ 03)||-2.7%||55||94||955||1.5%||47||-3.5||57|
|Boston, David (ARI 02)||-2.4%||54||75||572||-5.9%||62||-2.3||54|
|McCrary, Fred (NWE 03)||-57.9%||49||31||96||-57.0%||49||-17.6||48|
|Neal, Lorenzo (CIN 02)||4.2%||29||133||1.3%||1.3|
Other important additions: RG Solomon Page (DAL), S Kwamie Lassiter (ARI), CB Sammy Davis (R1)
Other important losses: WLB Junior Seau (MIA), S Rodney Harrison (NWE), CB/S Ryan McNeil (FA), MLB Orlando Ruff (NOR)
San Diego's offense was the opposite of their rivals up north. San Diego threw only 14% of passes over the middle, the lowest percentage of all 32 teams. They ran only 46% of running plays over the middle, which ranks 28th out of all 32 teams.
Then again, this makes you wonder. Did San Diego not run plays up the middle? Or perhaps, could this be a product of whoever does play by play for games in San Diego? After all, the question of whether a play is "left" or "middle" can be arbitrary if it is somewhere in between. Chalk that up as another issue for further study.
So, San Diego ran most of its plays to the sides -- and depending on the player, they were definitely better going one way or the other.
You'll notice above, on the line yards chart, that the San Diego line was much better blocking on the right-hand side. That right-side advantage is also easy to discern in the DVOA numbers, and it is equally strong for Tomlinson in both rushing and receiving:
San Diego's coaches apparently realized that Tomlinson wasn't having much luck on left-side passes, since they threw to him on the right far more often. San Diego was the only team in the league to have a running back lead the team in passes.
Now here's the quirk: that right-side advantage for Tomlinson is completely switched when it comes to the rest of Brees' receivers. The only three quarterbacks with a more pronounced left-side advantage were Peete, Maddox, and Harrington. Here are the DVOA numbers for Brees passing and for the top four San Diego receivers:
|D. Brees total||+24%||+21%||-23%|
The closer Drew Brees gets to the end zone, the worse his performance:
|DEEP (Own 0-20)||+14% DVOA|
|BACK (Own 20-40)||+6% DVOA|
|MID (40-40)||+1% DVOA|
|FRONT (Opp. 40-20)||-23% DVOA|
|RED (Opp. 20-0)||-49% DVOA|
He wasn't here last year, but I have to comment on the really strange usage of David Boston by the Arizona coaches. The best receiver on the Cardinals had two -- count 'em, TWO -- passes thrown to him in the end zone. Yes, I know he missed half the season, but still, two passes? Either he had three guys hanging on him every time the Cards got past the 20, or they have no idea what plays to run out in Phoenix.
The San Diego defense was great against the run but lousy against the pass when winning, and the exact opposite when losing:
|VOA vs. Pass||VOA vs. Rush|
|San Diego up by 8 or more||+32%||-10%|
|Game within 8 points||+8%||+1%|
|San Diego losing by 8 or more||â€“16%||+24%|
Remember, these numbers are based on a comparison with similar situations, so the fact that teams that are losing tend to pass while teams that are winning tend to run is taken into account.