RETRACTION: Why Hasn't Our Defense Improved?

[TheItalianStallion]

I agree with that. When he was hired I thought to myself "Why?" They had a guy in Gregg Blatche that had the defense playing good and it was a top 10 defense. They made the change because of turnovers. Not enough of them and we still don't get enough of them. So again I ask "Why?"

 

Our 2009 D was only top 10 in yards. In points they were 18th, and in turnovers they were dead last. They also tended to give up late scoring drives. Furthermore how many players from that D are still productive in the NFL? Aside from Hall, Carter, Rak, Fletch, Landry, and Rogers, I can't think of any. You can argue that the switch to 3-4 wasn't necessary and/or that Haz wasn't the right man for the job, but the D Shanahan inherited was in decline and on its last legs; a rebuild of some sort was in order.

[Oldfan]

The problem is your assumption that points per drive is a divine stat.

 

My problem must go deeper than that. I wasn't even aware there was such a thing as a divine stat.

[Oldfan]

I don't think Mike Shanahan or the people he has consistently placed around him is a bad judge of defensive talent, and I think he's added good young players through the draft in his time here for the most part.

 

Jenkins stepped into a starting role only because of injury. I think Kerrigan is the only other drafted starter in three years. That's not good.

[Pedro]

Talent evaluation is fine, from my perspective.  We used our really good draft picks on LT (when we failed to trade up for Bradford) and QB.  Outside of these hard to fill positions we've had a roster to transform.  We started off with all Team Captains/stand out character/low risk guys and this year transitioned to higher risk/reward as our roster has allowed.  

 

I like the attention DL and pass rush has had.  Cofield/Bowen/Carriker a long with Kerrigan is 4 out 5 positions filled by decent starters brought in via FA, a throwaway trade and a traded down 1st round draft pick.  Perry Riley has made big strides every year so far, and if he does it this year will be among the best in the league for his position.

 

Where we haven't done so well is our DBs.  Injuries, Mara and the pressing needs elsewhere that were addresses (like QB, LT, DL, Pass Rush, OL in general, WR) means that group is currently unknown, but we have added a bunch of talent.

 

We've used FA dollars and good draft picks on D.  That we went all in on a QB after previously getting a franchise LT when we had premium 1st round draft picks is what we should have done.  Where I think we can make a large and instant upgrade is the DC.  I thought we were whiffing for Griffin as a franchise because we started so many games going down multiple scores, then we'd step up and start stopping teams.  Turns out it's our DC is a coach not a coordinator.  Kyle has consistently had open pass catchers since he got here, gives me confidence he is productive in his game planning/calling role.  I'm still waiting to be impressed by Haz on those points, and we should have a guy doing that every week not trying to show something this far in (though it is consistent with his history).

 

Our Offense should make any DC look competent this season, which will be nice for Haz as he rides off into the sunset.

[garyclarkforhall]

You have to remember that the cap penalty came down minutes before free agency last year.  Who knows who they were targeting to fill the team's voids after dumping high-priced, overage players the year before.  I'm sure safety was on their radar screen last year.

 

The cap has led to a lot of risk-reward signings - Tanard Jackson (risk, we lost), Merriweather (risk, so far it hasn't panned out).

 

Couple that with Carriker's injury, Jenkins' injury that has slowed his progress, Rak's injuries... you can sort of understand why the defense isn't what they had hoped for yet.

 

There will be hell to pay next year when we get our cap space back!

[goldbean]

The average points-per-drive (PPD) is my choice for a quick-and-dirty stat to rank offenses, defenses, and teams. I use the rankings not the stat itself.

 

The question in your title doesn't match the analysis of your OP.  Your title asks why our defense hasn't improved, but you haven't established that it has not improved, only that the raw defensive PPD stat has not improved.  The title of the OP more accurately would be "Why hasn't our defense's PPD stat improved?"  That's a much narrower, and much different, question.

 

 

Drive stats can be found at the Football Outsiders site. The rankings for the units, offense and defense, are not as reliable as the "net" stat which combines the two. The site Two-Minute Warning has the net PPD stat at a .92 correlation to wins. That's the best stat by their calculations. A weakness in the PPD is that it does not adjust for strength of schedule. For example, here are the rankings for a few of Mike Shanahan's old teams:

 

The Football Outsiders drive stats for units (e.g., the defensive PPD stat that serves as the sole foundation of your argument re: the Redskins) aren't just "not as reliable" as another stat but unfit for year-to-year comparison because, by their terms, the PPD stats "are not adjusted for strength of schedule or situation."  

 

The import of the first qualification is obvious and potentially great; the import of the second may be greater.  If John Beck loses a fumble at the Redskins' 5 and the defense surrenders a TD, that is the same in PPD's eyes as the defense surrendering a 95-yard TD drive after a near-perfect, coffin-corner punt.  I don't see how one can use as the best proxy for performance a stat that treats as equal wholly different performances.

 

Net=Team Ranking

 

1997 Broncos: O=1, D=6, Net=1

1998 Broncos: O=2, D=13, Net=2

1999 Broncos: O=18, D=17, Net=21

 

 

[Football Outsiders] Before 2011, the worst one-year increase in strength of schedule belonged to the 1999 Broncos. Denver had ridden the third-easiest schedule (in a 30-team league) to a Lombardi Trophy in 1998, only to fall apart the next season under the weight of John Elway's retirement, Terrell Davis' Week 4 injury, and -- oh, by the way -- the toughest schedule in the league.

 

 

Given that PPD is unadjusted for strength of schedule and scenario (i.e., offensive performance that leads to the drive), this comment explains well much of the drop in defensive PPD in 1999.

 

 

Note that the 1997 defense ranked sixth. After taking over full control of the Broncos in 1999, Mike would have just two defenses ranked in the top ten over the next 13 seasons (Skins years included): 2005 (6) and 2006 (9). In contrast, he would have eight offenses ranked in the top ten over the same span, including 2012 (O=6). We already knew that Mike's teams have been unbalanced. These stats just confirm that and offer a little more perspective on the degree.

 

Here are the PPD rankings on Mike's Skins years:

 

2010: O=30, D=17, Net=27

2011: O=22, D=25, Net=27

2012: O=6, D=25, Net=12

 

How do we explain that the defense did better in 2010 than in 2011 and 2012? Why has there been no improvement? Jim Haslett told us that the Pittsburgh 3 - 4 would take three years to learn. Are the players not learning? Has the talent not improved? 

 

How would you explain this?

 

Again, your question is why the defense did better in unadjusted PPD--which ignores strength of schedule and in-game scenarios--in 2010 vs. 2011 and 2012.  Look first to the most obvious answers--different strength of schedule and the scenarios that produced the drives--which aren't quantified in the stat.

 

My question is why one would believe defensive PPD, which is admittedly tainted by variables the defense cannot control, is a better proxy for defensive performance than FO's defensive DVOA, a stat that does a better job controlling for those variables, particularly when FO prefers DVOA as the best reflection of a defensive unit's performance.

 

When we look at the weighted DVOA ranking of the Redskins' defense, it doesn't paint the same picture as the PPD stat you chose:

 

2010:  26th unweighted; 27th weighted

2011:  14th unweighted; 18th weighted

2012:  17th unweighted; 13th weighted

 

So the more appropriate question might be whether the defense is improving in light of stats that, when adjusted for strength of schedule, scenario, and baseline league-wide performance, suggest an upward trend, even though unadjusted defensive PPD is not improving.

[PapaLazarou]

Week 1- New Eagles offense. Will be interesting to see how long it takes Haz to adjust. (The Rams game still sticks in my mind- Danny ******* Amendola.)

[Voice_of_Reason]

If we start this season the way we

Finished last we will be fine.

Five of our seven wins to close the season were against division foes. Maybe our scouting isn't good against non-division rivals, or maybe Haz's game planning is slow to catch on to the opponent's tendencies. What do you think?

That's good food for thought.  You see a team more, you learn them more, you game plan against them better.

 

However, the D let Dallas complete back into the first Dallas game. For much of the second half, Tony Romo sits to pee looked like Joe Montana.  And going back through the last few years, there have been multiple facepalm moments, such as long, crushing bomb-type TDs at the ends of games, against the division teams.

 

I think last year, it had more to do with playing the Eagles twice, and they were just bad. The other three division games, Dallas twice and Giants once, were all REALLY close games.  It's not like the D turned into the '85 bears.  They were better, but not great. 

 

I'm going to go with injuries and losing half of your back end on defense from the 2nd to the 3rd year and have little to no money to replace it.

This is definitely the biggest component.  They expected Tanard Jackson and Merriweather to be the starting safeties.  Both were gone. Then they lost their best pass rushing DE and LB.  So, the coverage went away, and the pressure went away.  

 

Don't forget that they lost Jenkins 2 years ago, who they drafted and looked like he would provide some push off the edge, and that never materialized.  And he wasn't quite the same last year as he recovered.

 

In my head, it comes down a few different things:

1. Injuries - two guys clearly better than their backups (Carriker and Tanard), and two guys probably better (Meriweather and Orakpo)

2. Depth - (or lack thereof) roster overhaul, draft limitations, cap penalty

3. Lack of in-game adjustment - I've actually been pretty impressed with Haslett's game-planning, but the 2rd and 4th quarters were often abysmal.

4. Star power - we have had some solid players on D, but few (if any) elite players. The frequently poor acquisitions by the previous regimes, along with limited resources probably account for this.

Regarding the last point, I think this issue is why I'm excited about these rooks. Big time potential for the defensive picks makes the risks higher, but the reward potentially higher still.

I'm not sure if the OP's stat changed later in the season, but the Defense sure looked a lot better. Hopefully this is a sign of things to come.

I agree with all of this.

 

A couple additional points:

1. They went through a scheme change.  This generally takes a few years to do, gathering both starters and depth to fit the scheme.  And right in the middle of it, they got the cap penalty, which slowed down FA acquisitions.

2. They had to rebuild the offense at the same time.  It's not like the O was great, good, or even passable. So, with their limited resources, they had to build both sides of the ball.  In fact, of the 4 drafts, they've only spent 1 1st round draft pick on a defensive player: Kerrigan. Trent was 2010, Kerrigan 2011, Robert 2012, and the 2013 pick went for Robert. 

3. Has is not blameless.  He's getting better, and he's been dealt a crap hand with players, but he's been out-coached at times.  The first play of the Bengals game was inexcusable. 

 

 

During Shanahan's tenure here, his defenses have had:

 

1) No real NT

2) Absolutely no safeties

 

Get both of those and see how much the defense improves.

I sortof disagree about the NT.  I think of all the problems, Cofield isn't one of them.  He's not the best in the league, but he clogs up the middle.  Last year, I think he was fighting up hill all year. No real threat from the DE position. 

 

The safeties have been abominable since November 2007. RIP. 

[pjfootballer]

The answer to your question will be found in his coaching record as a DC. He has never had a top 10 defense since he left Pittsburg and that was in 1998 when his defense was ranked 7th over all. He had an 11th rank defense in 1997 and a 12th rank defense in 1999. All of it in Pittsburg. Every other place he has been in the 20's. From 2006 until now he hasn't had a defense ranked under 20. His defenses in St.Louis were 28th, 31st and 31st. His defensive ranks here are 21st, 21st, and 22nd last year. So why haven't they improved? It is due to a guy that is an average to below average success as a DC.

And that Pittsburg Defense was perenial top 5 before Has took over. Even with top flight talent, he makes it average. I've been against him since day one, and he has done nothing to make me change my mind.

I agree with that. When he was hired I thought to myself "Why?" They had a guy in Gregg Blatche that had the defense playing good and it was a top 10 defense. They made the change because of turnovers. Not enough of them and we still don't get enough of them. So again I ask "Why?"

Blatche was ready to retire. I never cared for his "bend but don't break" defense anyway. Plus, we were amongst the worst teams getting turnovers with him. We've been one if the better teams getting turnovers the last few years despite the injuries and problems.

What I've wanted is a defense like Pittsburgh's that can just totally shut a team down and frustrate an offense. I'd like to see us be more aggressive in the early downs and not the late downs.

I've never liked Haslett's style of coaching since he got here. I was excited when Mike hired him, but over the years he's worn on me like Danny Smith did.

[Tulane Skins Fan]

The problem is your assumption that points per drive is a divine stat.

 

My problem must go deeper than that. I wasn't even aware there was such a thing as a divine stat.

 

I think you over-estimate the "problem."  You can find stats where there has been great improvement.  For example, the defense was second in the NFC in takeaways last year, and that obviously contributed to us scoring more points and winning more games.

 

The Haslett debate is both perpetual and annoying.  It is always laced with the comparison to Gregg Williams' defense.  Defenses that never turned the ball over and never got a 4th quarter stop.  I would take last year's defense over Gregg Williams' defense every day of the week and twice on Sundays.

[RandyHolt]

Our defense improved by leaps and bounds by finally, finally, deciding to get after QB's.  

 

It started vs Philly and Foiles mid season and was the driving force behind turning our season around, and us making the playoffs. 

 

Our record against rookie QB's over the past decade was pathetic. It has been mind boggling how we flat out refused to try to pressure QB's, from Tom Brady to rookies, we treated them all the same. We waited for them to make a mistake, and facing no pressure, they rarely did. 

 

I still want to know whose idea was it to finally abandon our lame duck defensive playbook from the past decade. 

[Tulane Skins Fan]

Goldbean got this one right.  The defense "problem" is not what people continue to push it as.  The defense was very good where it counted last year.  

[DC9]

TSF,

 

"Devine stat" is my new favorite quote.  Well played.

 

#Lawyered

[skinny21]

I don't think Mike Shanahan or the people he has consistently placed around him is a bad judge of defensive talent, and I think he's added good young players through the draft in his time here for the most part.

Jenkins stepped into a starting role only because of injury. I think Kerrigan is the only other drafted starter in three years. That's not good.

It's not good, but it's not like we've whiffed on a bunch of defensive picks.

2010 - Riley was the only defensive player drafted

2011 - Kerrigan, Jenkins, Gomes, Thompson, White and Nield

2012 - Robinson, Crawford and Bernstine.

What's nuts is that we only drafted 5 guys before the 7th and that 4/10 of these defensive players drafted have been on IR (and Robinson twice!). Hard to develop injured guys.

Oh, and nice post Goldbean. This year, I'd like to see us improve on sacks and third downs...

[Oldfan]

 

The problem is your assumption that points per drive is a divine stat.

 

My problem must go deeper than that. I wasn't even aware there was such a thing as a divine stat.

 

I think you over-estimate the "problem."  You can find stats where there has been great improvement.  For example, the defense was second in the NFC in takeaways last year, and that obviously contributed to us scoring more points and winning more games.

 

The Haslett debate is both perpetual and annoying.  It is always laced with the comparison to Gregg Williams' defense.  Defenses that never turned the ball over and never got a 4th quarter stop.  I would take last year's defense over Gregg Williams' defense every day of the week and twice on Sundays.

 

Maybe you missed the two posts, but I have already criticized my stat on the takeaway factor. The defense's stat in effect gives a takeaway half-credit. It stops the opponent's drive early. But the advantage to the offense would shift over to make the offense's drive stat too high. So, that number six rank for the offense might be lower. The team "Net" stat should be more accurate. 

[pjfootballer]

POSTED BY BEDLAM

 

"I really don't understand the Cofield hate - There is the argument that your NT has to be the Sam Adams road grader or Cody (ravens) man mountain but there has be a trend towards more the pocket penetrating NT (like Ratclif in Dallas) and the roles of the the DL is always evolving . In 2011 Cofield played better than the 2012 Cofield but I think the DL were disappointing throughout last season ... but that might also be more a function of the lack of creativity

 

 

In the 2011 draft with the D we were still picking in the top 10 when we took Kerrigan and could have taken the best 3-4 DE in years in JJ Watt , but got good trade down value ...insted of going for the home run we went dependable blue collar types ... value over potential - and got a very good OLB in Kerrigan - and a potentially very good DE in Jenkins and some role players in Gomes and Nield but these are always going to be those JAG in many peoples minds when they talk of the skins and similar guys are on every single team in the NFL".

 

I see your point with Cofield, but I want to add another valid point you bring up with Kerrigan. I really hate to hear our fans whine and cry about not getting Watt.  I would rather have the multiple picks to strengthen the depth of the team then to single out one guy who may or may not have been able to elevate our defense.  Kerrigan doesn't get enough credit IMO and too many look at Watt and "wonder" what he would be like on this team.  But Watt went to a Texans defense that was already stacked.  I get tired of the "we could have had JJ Watt" BS, when we got a great player in Kerrigan and more good players with the picks that we acquired. We also addressed DE with Bowen and Carriker like you said.

[DC9]

People first need to realize the goal of the Haz scheme. 

 

It's not a "smashmouth" defense or a "bend but don't break" defense or even a "pick your poison defense".  It's a "trap" defense.

 

The overall goal of the defense is to show consistent looks in certain packages... and get the offense thinking one thing, and then trapping/baiting them into a turnover once they've reached a comfort level by tweaking the scheme a bit (which has been planned out before the game). 

 

This isn't the ONLY goal, but it's the overall goal.  You sacrifice a lot of yards in between the 20's using this style of play, but it pays dividens in the turnover department.

 

Now, the biggest problem with Haz, is he only plans one or two deep in his traps.  And frankly, he sucks at in-game adjustments. 

[Jumbo]

People first need to realize the goal of the Haz scheme. 

<edit>

 

Now, the biggest problem with Haz, is he only plans one or two deep in his traps.  And frankly, he sucks at in-game adjustments. 

 

Relatively minor shortcomings. <insert old halo smiley and old grin smiley>

[DC9]

People first need to realize the goal of the Haz scheme. 

<edit>

 

Now, the biggest problem with Haz, is he only plans one or two deep in his traps.  And frankly, he sucks at in-game adjustments. 

 

Relatively minor shortcomings. <insert old halo smiley and old grin smiley>

 

 

Haha, no doubt!  And I think Rah (or my longshot dream... Spanos) would change the scheme and mentality up to put the players in more of a position to be successful based off of our talents. 

 

But the point I was making, is that in terms of the overall goal of his defensive scheme, he should have high marks for last season. 

 

Now, a lot of that goes to Robert taking care of the ball (turnover differential, etc) and the offense running over 50% of the time.  Chicken and the egg.

 

Do we have the same "defensive success" in terms of turnovers if Rex is the quarterback?  Doubtful.

 

I think Oldfan did a post on that a while back in the offseason.  I can't remember if it was him or something else, but the topic was close.

[TSO]

People first need to realize the goal of the Haz scheme. 

 

It's not a "smashmouth" defense or a "bend but don't break" defense or even a "pick your poison defense".  It's a "trap" defense.

 

The overall goal of the defense is to show consistent looks in certain packages... and get the offense thinking one thing, and then trapping/baiting them into a turnover once they've reached a comfort level by tweaking the scheme a bit (which has been planned out before the game). 

 

This isn't the ONLY goal, but it's the overall goal.  You sacrifice a lot of yards in between the 20's using this style of play, but it pays dividens in the turnover department.

 

Now, the biggest problem with Haz, is he only plans one or two deep in his traps.  And frankly, he sucks at in-game adjustments. 

 

 

This is perfectly said and I've always understood the defense in this manner. It's why I never minded when the opponent seemed to rack up the yards on a drive because I was expecting for it to end in a turnover. The problem last season was that Haslett went into safe mode after the injuries and decided to play a cover 3 shell with one ILB blitzing pretty much every passing down.

 

As for your last statement, Haz has changed in my mind. 

 

Like I said before, after the bye, all I know is that Has, Slowik and Rah really started to gel and we started to see those turnovers pop up again. We confused offense's and our defense was effective more often that not, finishing the season VERY strong by holding a Dallas team that was absolutely ON FIRE offensively to under 20 points and then holding a Seattle team in the playoffs that was also ON FIRE offensively to 13 points for almost the entire game while our offense went three and out over and over again.

 

So, yeah. Stats can go to hell on this. I'm excited about our defense and saw big improvement. I just can't believe we'll all of a sudden revert back to Haz' old habits. I think the coaches know how to utilize their personnel right now and have found the secret recipe.     

[DC9]

This is perfectly said and I've always understood the defense in this manner. It's why I never minded when the opponent seemed to rack up the yards on a drive because I was expecting for it to end in a turnover. The problem last season was that Haslett went into safe mode after the injuries and decided to play a cover 3 shell with one ILB blitzing pretty much every passing down.

 

I think it depended on the opponent and the weapons which they possessed (ie: You play Romo sits to pee and Dez Bryant differently then you play Weeden and Josh Gordon, etc) but I agree with you 95% on this.

 

As for your last statement, Haz has changed in my mind. 

 

Like I said before, after the bye, all I know is that Has, Slowik and Rah really started to gel and we started to see those turnovers pop up again. We confused offense's and our defense was effective more often that not, finishing the season VERY strong by holding a Dallas team that was absolutely ON FIRE offensively to under 20 points and then holding a Seattle team in the playoffs that was also ON FIRE offensively to 13 points for almost the entire game while our offense went three and out over and over again.

 

Good points and I'm in on Haz this year as well.  But I'd like to see some consistency.  Especially with his new toys and the perceived depth we have.  Personally, I don't care if it was a trust thing or if it was a team effort thing when it came to gameplanning, keep it up.

 

So, yeah. Stats can go to hell on this. I'm excited about our defense and saw big improvement. I just can't believe we'll all of a sudden revert back to Haz' old habits. I think the coaches know how to utilize their personnel right now and have found the secret recipe.     

 

Agreed.  You have to look at overall schemes and goals of a defense before you cherry pick stats.  There are players that even say "Yeah, we let them have that completion so that we could take advantage of it two formations down the road in the game."

[Puma Kane]

I think the defense is bad, due to a combo of injuries and too much risk taking in the zone blitz scheme on short downs.

Teams seem to pick up short yardage on short down plays lol. Then the defense seems to tire and Jim wants to blitz, when our guys are dragging the ground.

There are major differences between Greg Williams and the Haz. One can call plays, where the other struggles.

I would love to see Haz drop 9 or 10 on short downs.

[Oldfan]

Goldbean: So the more appropriate question might be whether the defense is improving in light of stats that, when adjusted for strength of schedule, scenario, and baseline league-wide performance, suggest an upward trend, even though unadjusted defensive PPD is not improving.

 

Forgive the late reply. I only saw your post after TSF said you were right.

 

And... he's right. You're right.

 

The DVOA can't be much better than the PPD's Net stat (.92 correlation to wins) despite all the effort they make to include more variables. Those variables are mostly accounted for in PPD.

 

However, although I mentioned that the unit stats were less reliable, it never occurred to me to switch to DVOA for this OP.

[HogHeaven1983]

Probably a HUGE Stretch ...

 

Redksins get big leads and give up big yards in prevent defenses?

I am thinking of the Dallas game in Dallas

And didnt that happen in the Vikings game too?

 

Seriously I think all those injuries to our secondary (and the T.Jackson suspension) killed us last year ... HUGE passing plays against us

[DC9]

Probably a HUGE Stretch ...

 

Redksins get big leads and give up big yards in prevent defenses?

I am thinking of the Dallas game in Dallas

And didnt that happen in the Vikings game too?

 

Seriously I think all those injuries to our secondary (and the T.Jackson suspension) killed us last year ... HUGE passing plays against us

 

I'm going to be honest here, and it's not going to be a popular opinion.  Everyone wants to go for the jugular all the time, etc.

 

But I don't really think we were all that worried about Dallas on Thanksgiving.  As much as I was sweating and as uncomfortable the ending was, if you are smoking a team like that... I think you can afford to change the game plan up and allow a team to work their way down the field, but make sure that they "earn" it.  Like, 5-8 minute drives.  Then get the ball back and milk 2-4 minutes off, that's a win in my opinion.

 

The big play in the Dallas game came from Dez Bryant and it was more poor tackling in the secondary and the fact that Romo sits to pee was back there for 20 minutes before he threw the ball.

 

Bottom line with those types of games, you'll trade scores for time off the clock, but big plays are killers. OR - as long as you don't turn it over, you should win those games.

 

The Vikings game, if memory serves, they jumped on us early and then we came back and scored in bunches on them.  The defense had a "sandwich game" where the 1st and 4th we were shaky and the 2nd and 3rd we were pretty good.

 

Not having Tanard hurt, but their were other things afoot during that part of the season.  Namely the lack of a 3rd down ROLB.  Haz was toying with Chris Wilson and Lorenzo... he finally gave way to RJax the following week against the Giants.