Data Diving into the Defense Carrying the Offense debate

[KiwiHawk]

I was asking in relation to your NFL average points per game - whether you factored the defense and ST points out of that average in order to compare apples-to-apples. It sounded like you deducted them from our offensive points, but not the NFL average.

 

[Natethegreat]

I was going to point out the same thing. If your going to deduct points not scored by the offense you would need to reduce the overall average score as well or simply count those points since they are counted in the average.

 

[Anthony!]

One more of note, On defense we have 7 of the 11 positions paid amongst the highest5 at those positions, while on offense only 3.

For instance, if we look at the top 5 players on the Hawks by cap hit they are

Wilson 18mil
Sherman 13.6 mil
Bennett 11.2 mil
Et 10.4 mil
JG at 10 mil

So we have 28 mil offense and 35mil defense

IF we go further and look at top 10


Wilson 18mil
Sherman 13.6 mil
Bennett 11.2 mil
Et 10.4 mil
JG at 10 mil
DB at 9.6 mil
KC at 8 mil
Wagner at 7,6 mil
JOeckel at 7 mil
Wright at 6.8 mil

So offense is 44.6 mil
Defense at 57.6 mil

So basically the defense should be better than the offense, in fact over all we spend 91.6 mil on defense and 76 on offense enough said

The top 6 position groups with regards to cap spent are all defense so they should be better and frankly Sherman should realize that. How good would Sherman look if they skimped on the dline and ends like they have on Oline? Answer not very

http://www.spotrac.com/nfl/seattle-seahawks/positional/

 

[Hyak]

On adjusting the average, I did not adjust it. For 2016, it looks like the differential is close to 1 point (0.96). A quick glance in terms of impact:

2 games go from green to red for the defense (2013 win vs Arizona 34-22 and 2014 NFCC win versus GB 28-22)
0 games impacted for the offense from 2012-2016.

 

[Seymour]

I'll have to agree with Anthony. The only way to determine who is not pulling their weight, is to figure in money spent. When you are robbing Peter to pay Paul and Peter is always playing somewhere on offense, then that has to be accounted for.

 

[Hyak]

Diving into the Big Play element of things, which saw a big regression in explosive run plays on offense that was the largest impact in the Big Play differential. I really think the injuries to Wilson and Rawls were huge factors in this as their explosive plays dropped significantly from 2015 for TW and career wise for RW. On a positive note for 2017, the playoff numbers were more in line with 2015 when both were relatively healthy.

source: https://www.sportingcharts.com/nfl/stat ... stics/2011

/

 

[MontanaHawk05]

I too would like to see the impact of Special Teams analyzed. I think you'll find it's quite significant.

We seemed to be constantly making big plays on ST in 2013-2014, and that went away after 2015. Pretty simplistic analysis, I know. But another reason the defense was restocked in the draft this year is that those draftees will be reinforcing ST and look like the type to do it well.

 

[Sterlinghawk]

Nice post. Thanks for your efforts.

 

[Popeyejones]

Great job on all of this JTB!

TWO VERY VERY nitpicky notes on the OP:

1) You're relying overall average points scored/allowed to set your binary cut-off point, but for your Hawks cut-off points you're eliminating defensive and special teams scores to just isolate out the offense.

I think it makes a lot of sense to do that given your question, but because the 22.75 points scored/allowed doesn't eliminate defensive and special teams scores your comparison of the Hawks' offensive output is ultimately comparing apples to oranges (i.e. the Hawks' offense needs to score 23+ points to be above average but nobody elses does, because other teams' defense and special teams contributes to their point totals in getting to 22.75).

To not do that you'd have to do OFFENSIVE points scored/allowed (which I'm assuming isn't available), or just accept that defensive and special teams scores are going to create some noise in the data and attribute those to the offense (luckily defensive and special teams scores tend to be pretty random so over six years they should even out across the league and not pose any problems for you).

2) If you want to use round numbers 22.75 should always be 23, because teams can't score fractional points. Mathematically you'd want Offense and Defense to both get a plus 1 for greater than 23 and a minus 1 for less than 23. Any score of 23 is average and results in neither a plus or minus for either side.

To be clear I am ABSOLUTELY NITPICKING and NOT SUGGESTING you go back and do either of these things (I don't think they'd change much), but if you end up updating the stats or you're a statistical nitpicker like me ( :lol: ) I figured these things could be worth noting, though.

Again though, I don't think they'd really change anything and this all VERY insightful. Thanks! :2thumbs:

 

[Popeyejones]

I'll have to agree with Anthony. The only way to determine who is not pulling their weight, is to figure in money spent. When you are robbing Peter to pay Paul and Peter is always playing somewhere on offense, then that has to be accounted for.

Yep, and that's something that can be done! To do so you need to look ar proporational offensive and defensive spending per year. You can do that by attributing percentage of defensive spending from offensive and defensive total spending.

For each year, this is the percentage of offensive and defensive spending spent on the DEFENSE:

2013: 42.7
2014: 54.2
2015: 57.5
2016 58.6

This all comes from OTC, only going back to 2013 because they don't have spending for 2012.

From there you can get to OVER/UNDER performance by spending from JTB's table. The math is pretty simple. For 2013 you'd expect the defense to over-perform on expecations more than 42.7% of the time, and the offense to over-perform on expectations more than 57.3% of the time. We can just use JTB's +1/-1 System to do that too.

FROM THAT:

2013: Offense -1 (7 percentage points below expecations), Defense + 1 (37.5 percentage points above expectations
2014: Offense +1 (3.5 percentage points above expectations. Defense + 1 (10.3 percentage points above expectations)
2015: Offense -1 (2 percentage points below expectations), Defense +1 (11 percentage points above expectations)
2015: Offense +1 (6.8 percentage points above expectations), Defense -1 (21.4 percentage points below expectations)


From that using our +/- system we can say that the offense performs average to proportional spending expectations (2/4 years), and the defense peforms above average to proporational spending expectations (3/4 years).

That washes out the size of the differentials though, so let's just add up those differentials for each year and then divide by four to get a typical year:

In total, from 2013 to 2016 the Seahawks offense outperforms their proportional spending expectations by 1/3 of 1 percentage point per year, and the Seahawks defense outperforms their proportional spending expectations by 7.7 percentage points per year.

In conclusion, from 2013-2016 the Hawks not only leaned on their defense more than their offense, but they also got BETTER proportional monetary value from their defense than their offense:

:2thumbs:

 

[Hyak]

Cool stuff here. Thanks for the feedback. This was a first brush and I may look at it so more to reward really good performances or penalize really bad ones.

The irony is that when the spending was heavy on offense, the ROI from the players making the big money was poor with money sunk into the OL, Rice, Harvin, and Miller.

 

[Popeyejones]

Cool stuff here. Thanks for the feedback. This was a first brush and I may look at it so more to reward really good performances or penalize really bad ones.

The irony is that when the spending was heavy on offense, the ROI from the players making the big money was poor with money sunk into the OL, Rice, Harvin, and Miller.

For sure man. :2thumbs:

And great point that even back then the ROI on offense wasn't actually an ROI on where the offensive spending actually was -- Wilson, Baldwin, and Lynch collectively cost practically nothing, and that's where a ton of the offensive ROI was actually coming from.

You can do it the other way too, noting that the ROI on defense is going down from year-to-year largely because over time the Seahawks had to pay the guys who make their defense what it is at the actual market rate of second contracts. For ROI the defense is getting "worse" because the same guys cost more, not because the defense is actually getting worse.

And of course, just to say it, a major nitpick of my own post above is that it bakes in the assumption that a typical team spends equally on its offense and defense, which I'd *guess* probably isn't true. To really do THAT ONE right you'd want to do proportional offensive and defensive spending over league average offensive and defensive spending. My *guess* is that teams typically spend more on offense than defense, meaning that I might be overstating the over-performance of the Hawks' defense and under-stating the performance of the Hawks' offense when thinking about proportional spending.

The takeaway, however, still holds I think: The Hawks defense contributes more than their offense when it comes to their winning ways, but it's not like they're CARRYING the offense or anything.

Fun stuff. :lol:

 

[Hyak]

I think most teams want to have some level of balance on the roster but whenever a team goes into second contract mode for their stars, it's going to impact the roster makeup. To a certain extent, the investment in the core defensive guys at the level the Seahawks did is unusual. That said, they clearly shifted the investment from the OL to pay for it and it's hard to argue given the ROI experience.

I do think we'll see some shift as these defenders approach 3rd deals and 30 years old.

 

[Popeyejones]

^^^ Yeah, absolutely agreed.

I think I've said it here before, but the Hawks are a really fascinating case-study/experiment to me because they've really taken an NBA-Style approach to roster construction, which other teams haven't.

The NBA uses a superstar model in which you dump all your resources into two or three of your players, and then just surround them with a bunch guys on the cheap. You steal from Joe and Schmoe in order to pay your group of superstars.

NFL teams are larger but it's the same deal with the Hawks, who have 10 guys getting paid in the top 5 at their positions, and then just a bunch of guys not making anything at all.

It works in the NBA because it's such an individuated game (e.g. you amass three superstars and then it's basically everyone else's job to stay out of the way and not be too terrible).

But nobody has really tried it out in football before where the game is much less individuated (e.g. a QB is only as good as his protection; a WR needs a good QB; a good secondary needs a good pass rush; good linebackers need good d-lineman to occupy blocks).

The Hawks have been in the playoffs year in and year out so I think it's been working, but I do think there's a very big story to be written about the Hawks' cap strategy and how unique it is that has yet to be written.

 

[Hyak]

To me, the regression in 2015/2016 as a team (when compared to expectations) is tied to the poor drafts in 2013/2014 along with the departure of solid but 2nd tier players like Maxwell. As a result, the depth behind the star power starters was not anywhere close to what it was in 2012 - 2014.

One thing for sure - the proven method for a consistent winner is good drafting whether you pay all of the stars (SEA) or you let some of them walk (NE).

 

[mrt144]

Does this mean popeye gets a reprieve before we run him out on rail?

 

[Jville]

To me, the regression in 2015/2016 as a team (when compared to expectations) is tied to the poor drafts in 2013/2014 along with the departure of solid but 2nd tier players like Maxwell. As a result, the depth behind the star power starters was not anywhere close to what it was in 2012 - 2014.

One thing for sure - the proven method for a consistent winner is good drafting whether you pay all of the stars (SEA) or you let some of them walk (NE).

All good points. Although, I'll offer an amended characterization for the words ..... "poor drafts".

When Carroll and Schneider and Company reviewed the 2012 season I suspect they had a very clear read on the make up of the core players they succeeded in collecting. They also had a clear understanding that they were not going to be able to keep everyone. As a result, they chose to change their approach to the draft with regards to goals and objectives during the 2013, 2014 and (I would add) 2015 drafts to minimize core player losses. They economized by trading away 1st round picks, picking unique traits that complemented starters and filled in holes and also experimenting more heavily with conversion and developmental players as a tactic to avoid competitive bidding .... i.e. conserve draft and salary capital.

After reviewing the results from those 3 drafts, they once again altered their draft and off season approach. They began to reload as cap casualties made themselves known. In 2016, the offense saw the bulk of reloading. In 2017, the defense benefited from much needed additions. Circling back around, they're once again looked for starters and leaders. Adding bigger and faster play makers has been the the priority of the last two drafts.

In the spirit of linking opinions to data, take a look at the included link as an aid in reviewing Schneider/Carroll drafts >>> [urltargetblank]http://www.pro-football-reference.com/teams/sea/draft.htm[/urltargetblank].

I suppose I should also acknowledge that Schneider has addressed the ever changing goals and objectives for each draft since his arrival in Seattle. My views reflect what I recall from those draft debriefings. Presumably, those press conferences are archived at seahawks.com. I apologize for not gathering together all the links to post draft conferences for everyone's convenience. But it is a beautiful day, the sun is out and yard work is calling.

This is one of the better threads in sometime. :2thumbs: Your contribution is appreciated.

Go Hawks!

 

[Seahawkfan80]

^^^ Yeah, absolutely agreed.

I think I've said it here before, but the Hawks are a really fascinating case-study/experiment to me because they've really taken an NBA-Style approach to roster construction, which other teams haven't.

The NBA uses a superstar model in which you dump all your resources into two or three of your players, and then just surround them with a bunch guys on the cheap. You steal from Joe and Schmoe in order to pay your group of superstars.

NFL teams are larger but it's the same deal with the Hawks, who have 10 guys getting paid in the top 5 at their positions, and then just a bunch of guys not making anything at all.

It works in the NBA because it's such an individuated game (e.g. you amass three superstars and then it's basically everyone else's job to stay out of the way and not be too terrible).

But nobody has really tried it out in football before where the game is much less individuated (e.g. a QB is only as good as his protection; a WR needs a good QB; a good secondary needs a good pass rush; good linebackers need good d-lineman to occupy blocks).

The Hawks have been in the playoffs year in and year out so I think it's been working, but I do think there's a very big story to be written about the Hawks' cap strategy and how unique it is that has yet to be written.

This is the exact thoughts I have been having for the past 2 years. We have rebalanced the ball back to center and I hope we dont go all over the place with that ball. Offense wins games, defense wins championships.

One of the drafts and seasons, we ended up trading a unique cog in the defense and the one wheel was not as stable and only full of hot air. We lost a lot of that season. But recovered and are on track again this year with the new fill ups that we can hope will make the team work positively.

Great post Popeye......I am gonna eat some spinach tomorrow from my garden in honor of this post.

 

[brimsalabim]

Our head coach puts his emphasis on the defense. Money is spent on the defense. Our offensive is called conservatively so as not to put the defense in bad positions. Our QB takes numerous blind side shots so the defense can have a multitude of situational tackles on the roster while the offense lacks even one starting caliber tackle. Pete has scripted these very specific rolls and I get the feeling he isn't going to change anything.

 

[Sgt. Largent]

We went prevent on that last drive in 2012. Quinn would not have and we would have won.

Prevent just doesn't work. It's so conservative that it's almost like bending over.

It wasn't prevent, it was their usual cover 3 zone bracketing the receivers on the sideline............we even rushed 4 and sent a 5th blitzer on both the final two pass plays. I blame Trufant, he played WAY too shallow on the first sideline catch, NO REASON WHATSOEVER TO WORRY ABOUT THE RB AND LEAVE THE MID RECEIVER WITH THAT MUCH SPACE.

Prevent is rushing 2-3 and dropping everyone into zone.