The Inconsistency of the Defense in Numbers

[PF_SU]

Digging a bit deeper into the splits of the defense in the first four week it's extremely startling just how inconsistent the defense has been. Not from a game-to-game basis but from a half-to-half basis within a single game. If you take the best halves of each game the defense looks like a legit contender for best defense in the league. Take the worst halves of each game and the D looks like the worst defense out there. Now, every team will experience a certain measure of fluctuation. The best defenses will look mortal for a drive, a quarter or even a half. And the worst defenses can put together an occasional fight too. But the Jekyll-and-Hyde like inconsistency of the defense through week four has been extreme. To put it into numbers:

Week 1 vs Arizona

First half: 30 plays, 157 yards, seven points, one three-and-out, 3-7 third down conversions in five drives. 6 plays, 31.4 yards and 1.4 points per drive.

Second half: 29 plays, 183 yards, 14 points, 3-5 third down conversions in four drives. 7.25 plays, 45.8 yards and 3.5 points per drive.

Week 2 vs Miami

First half: 28 plays, 138 yards, three points, three three-and-outs, 1-5 third down conversions in seven drives. 4 plays, 19.7 yards and 0.43 points per drive.

Second half: 34 plays, 320 yards, 21 points, 4-5 third down conversions in five drives. 6.8 plays, 64 yards and 4.2 points per drive.

Week 3 vs Houston

First half: 34 plays, 136 yards, zero points, 4-8 third down conversions in five drives. 6.8 plays, 27.2 yards and 0 points per drive.

Second half: 34 plays, 127 yards, zero points, one three-and-out, 2-7 third down conversions in five drives. 6.8 plays, 25.4 yards and 0 points per drive.

Week 4 vs Buffalo

First half: 42 plays, 221 yards, 13 points, 4-7 third down conversions in four drives. 10.5 plays, 55.3 yards and 3.25 points per drive.

Second half: 35 plays, 122 yards, three points, one three-and-out, 3-8 third down conversions in five drives. 7 plays and 24.4 yards and 0.6 points per drive.

Splits

Best halves of each game: 127 plays, 544 yards, 13 points, six three-and-outs, 9-27 third down conversions (33.3%) in 22 drives. 5.8 plays, 24.7 yards and 0.6 points per drive.

Worst halves of each game: 139 plays, 860 yards, 48 points, no three-and-outs, 15-25 third down conversions (60%) in 18 drives. 7.7 plays, 47.8 yards and 2.7 points per drive.

Combined: 266 plays, 1404 yards, 61 points, six three-and-outs, 24-52 third down conversions (46.2%) in 40 drives. 6.7 plays, 35.1 yards and 1.5 points per drive.

Comparison

Minnesota Vikings: 5.8 plays, 26.8 yards and 1.11 points per drive.
Seattle Seahawks: 5.4 plays, 24.5 yards and 1.18 points per drive.
Denver Broncos: 5.9 plays, 24.3 yards and 1.39 points per drive.
.
.
.
Atlanta Falcons: 6.3 plays, 37.9 yards and 2.59 points per drive.
New Orleans Saints: 6.3 plays, 38.3 yards and 2.61 points per drive.
Washington Redskins: 6.6 plays, 40.2 yards and 2.73 points per drive.

 

[FCB02062]

Yup, the Pats defense just isn't that good right now , or at the very least it's not meeting expectations

 

[PScottman]

What is the definition of good? Is it the same as expectation? Is it potential? Star to JAG ratio? Media coverage? How about consistency? What portion of defense? Scoring? Yards? Turnovers? Sacks?

There are a lot of ways we can measure this defense and that measure will also morph as we go through the season, just like always. If the defense continues like this into week 7 and 8...let the circular firing squad commence.

 

[Gronk]

It's interesting to see laid out in figures but I'm not sure there's much to truly conclude because we're taking the other teams' overall numbers at face value but hyper-analyzing the Pats' splits. As you said yourself, every team will experience some level of fluctuation, but your experiment basically assumes that the Pats fluctuation is out of the ordinary in a way that is statistically significant. You're probably right that their best v. worst half has varied more than an average teams, but



In any case 4 games is a pretty small sample size - and of those one of them is a shutout, and the loss that has caused a lot of the hand-wringing over the defense was a game in which they only gave up 16 pts with an offense that struggled to do much. There are things they need to improve on, yes, but I don't think it's correct to say they're not good.

 

[Uptown]

The D has also been on the field A LOT longer due to the offense not being able to sustain drives (mostly with Brissett at the helm).

Having The GOAT will help the DEF simply by keeping the OFF on the field longer & allowing the DEF to rest & make adjustments. Also, having our leader Ninko back is vastly underrated!

 

[RayClay]

Football isn't a game that lends itself to statistic analysis. We went 3-1 with a backup and a rookie and they pitched a shutout in the game they needed to. I think our defense looks real good. There are some changes and i expect them to come together later in the season as guys like Long, Flowers, Cyrus Jones and Valentine get settled and improve. Hightower and Collins together will cure a lot of ills too.

I really like our defense this year, I don't know what people are expecting.

 

[Ring 6]

Let me introduce you to something called sample size. Remove the 2nd half a Miami playing with a big lead and a rookie qb who never played before and this statistic you invented if worst half and best half normalized.

 

[PF_SU]

It's interesting to see laid out in figures but I'm not sure there's much to truly conclude because we're taking the other teams' overall numbers at face value but hyper-analyzing the Pats' splits.

Coincidentally I started doing the same thing for the top-three defenses (Vikings, Seahawks, Broncos):

Minnesota Vikings

Best: 122 plays, 439 yards, 16 points, 8 three-and-outs, 6 turnovers, 24 drives. 5.08 plays, 18.29 yards, 0.67 points per drive.

Worst: 145 plays, 772 yards, 34 points, 5 three-and-outs, 4 turnovers, 21 drives. 6.9 plays, 36.76 yards, 1.62 points per drive.

Seattle Seahawks

Best: 108 plays, 478 yards, 19 points, 11 three-and-outs, 3 turnovers, 24 drives. 4.5 plays, 19.92 yards, 0.79 points per drive.

Worst: 128 plays, 668 yards, 35 points, 5 three-and-outs, 1 turnover, 19 drives. 6.74 plays, 35.16 yards, 1.84 points per drive.

Denver Broncos

Best: 125 plays, 368 yards, 12 points, 7 three-and-outs, 2 turnovers, 22 drives. 5.68 plays, 16.73 yards, 0.55 points per drive.

Worst: 142 plays, 733 yards, 52 points, 5 three-and-outs, 4 turnovers, 21 drives. 6.76 plays, 34.9 yards, 2.48 points per drive.

New England Patriots

Best: 127 plays, 544 yards, 13 points, 6 three-and-outs, 2 turnovers, 22 drives. 5.77 plays, 24.73 yards, 0.59 points per drive.

Worst: 139 plays, 860 yards, 48 points, 0 three-and-outs, 3 turnovers, 18 drives. 7.72 plays, 47.78 yards, 2.67 points per drive.

So, as can be seen, when the defense is on they are matching the performance of a top-3 defense. Yes, they allow more yards, but that shouldn't surprise anybody knowing Belichick's philosophy. In terms of points allowed in the better halves of each game only the Broncos do better and only by a single point. But when they are off, they can stop anything. If I had to point out a single stat it would be the three-and-out one. No three-and-outs for the Patriots' D in their worse halves against the Cardinals, Dolphins, Texans and Bills.

The point here is not to show how "bad" the defense is right now, but that their performance is fluctuating wildly. They have the tools to be a top-10, if not top-5, defense. But they have to get their gaffes under control. Hopefully starting today.

Let me introduce you to something called sample size. Remove the 2nd half a Miami playing with a big lead and a rookie qb who never played before and this statistic you invented if worst half and best half normalized.

I know the issue of sample size. But even if you remove the second half of the Miami game you still end up with a per-drive basis of 8.1 plays, 41.5 yards and 2.1 points per drive. They were unable to stop the Cardinals in the second half and they were dragged around the field by the Bills in the first half. That has to improve. But once the improve on consistency this defense can be very good.

 

[Ring 6]

Coincidentally I started doing the same thing for the top-three defenses (Vikings, Seahawks, Broncos):

Minnesota Vikings

Best: 122 plays, 439 yards, 16 points, 8 three-and-outs, 6 turnovers, 24 drives. 5.08 plays, 18.29 yards, 0.67 points per drive.

Worst: 145 plays, 772 yards, 34 points, 5 three-and-outs, 4 turnovers, 21 drives. 6.9 plays, 36.76 yards, 1.62 points per drive.

Seattle Seahawks

Best: 108 plays, 478 yards, 19 points, 11 three-and-outs, 3 turnovers, 24 drives. 4.5 plays, 19.92 yards, 0.79 points per drive.

Worst: 128 plays, 668 yards, 35 points, 5 three-and-outs, 1 turnover, 19 drives. 6.74 plays, 35.16 yards, 1.84 points per drive.

Denver Broncos

Best: 125 plays, 368 yards, 12 points, 7 three-and-outs, 2 turnovers, 22 drives. 5.68 plays, 16.73 yards, 0.55 points per drive.

Worst: 142 plays, 733 yards, 52 points, 5 three-and-outs, 4 turnovers, 21 drives. 6.76 plays, 34.9 yards, 2.48 points per drive.

New England Patriots

Best: 127 plays, 544 yards, 13 points, 6 three-and-outs, 2 turnovers, 22 drives. 5.77 plays, 24.73 yards, 0.59 points per drive.

Worst: 139 plays, 860 yards, 48 points, 0 three-and-outs, 3 turnovers, 18 drives. 7.72 plays, 47.78 yards, 2.67 points per drive.

So, as can be seen, when the defense is on they are matching the performance of a top-3 defense. Yes, they allow more yards, but that shouldn't surprise anybody knowing Belichick's philosophy. In terms of points allowed in the better halves of each game only the Broncos do better and only by a single point. But when they are off, they can stop anything. If I had to point out a single stat it would be the three-and-out one. No three-and-outs for the Patriots' D in their worse halves against the Cardinals, Dolphins, Texans and Bills.

The point here is not to show how "bad" the defense is right now, but that their performance is fluctuating wildly. They have the tools to be a top-10, if not top-5, defense. But they have to get their gaffes under control. Hopefully starting today.


I know the issue of sample size. But even if you remove the second half of the Miami game you still end up with a per-drive basis of 8.1 plays, 41.5 yards and 2.1 points per drive. They were unable to stop the Cardinals in the second half and they were dragged around the field by the Bills in the first half. That has to improve. But once the improve on consistency this defense can be very good.

The defense is allowing 15.25 points per game. It is already very good.

 

[A Defiant Goose]

Reminds me of the 2010 and/or 2011 seasons. The defense would play well and then once we had a sizable second half lead they'd give up a **** ton of yards and let the other team milk the clock out. The end result would be a win with a defense that somehow allowed 400+ yards but only 17 points. Maybe they just go into prevent defense mode too soon, that's what I saw in the Dolphins game at least.

 

[farn]

I still believe in this D. As Fall moves on, so will the D. If HT gets healthy - that will help.

If this D can allow 15 pts as it "struggles", I can't wait for the winter D.

 

[ausbacker]

I still believe in this D. As Fall moves on, so will the D. If HT gets healthy - that will help.

If this D can allow 15 pts as it "struggles", I can't wait for the winter D.

Through 5 games, the Patriots D is averaging 14.8 PPG, good for 4th in the NFL.

 

[mosslost]

Yup, the Pats defense just isn't that good right now , or at the very least it's not meeting expectations

It's number One in the AFC for points against though.

 

[ALP]

Cmon guys we see this every year

Other defenses are plateau from beginning of season to end, but we are not and haven't been in years, our D's have always started weaker/slower and about halfway through the season, evolved into what they truly are for that season

I think this year it's going to happen earlier because of the small amount of turnoever this past offseason, but this is not our true playoff defense, so let's hold off on the comparisons at the moment

 

[long distance]

.. and the teams we played have combined 9-16 record .. and the statistic merry-go-round can go on and on ..

The fact is yesterday was the first outing of the actual 2016 team .. and evaluating D regardless of the team situation can hardly have sense.

The Bengals will be a good test. They're still a solid team with season on the line coming to Foxboro and the key playmaker in the O Eifert will be (reportedly) returning.

The D is obviously still in the making. Hightower's injury, Ninko's absence, McClellin's injury, Valentine's injury and inconsistency at CB are slowing the process a bit. The emergence of Roberts (and partly also Hamilton) will help. What this D needs is a healthy Hightower and a solid #3 corner. BB is clearly searching and evaluating this position for several weeks now exchanging Coleman, Cyrus and JJ - but unfortunately no one really grabbed the opportunity. The biggest hope for this weak link that would solidify the whole group IMO lies in Rowe - also because of his size and agility. I can only hope next week he is ready to go, he gets the chance and grabs it.

The other thing I really hope is they bring another DT (or Valentine if healthy or A Johnson from PS) to the active roster so that we have a normal big body rotation again because the workload on Branch and especially Brown is getting ridiculous and dangerous ..

 

[Username]

The defense has been solid, but the DBs, with the exception of Chung, have been average at best thus far. And that's being kind. McCourty has been strangely nowhere to be seen on quite a few deep balls this year, leaving struggling CBs isolated. Hopefully it'll all come together soon because if the secondary improved we'd be pretty tough to move the ball on.

 

[Kontradiction]

The defense has been solid, but the DBs, with the exception of Chung, have been average at best thus far. And that's being kind. McCourty has been strangely nowhere to be seen on quite a few deep balls this year, leaving struggling CBs isolated. Hopefully it'll all come together soon because if the secondary improved we'd be pretty tough to move the ball on.

Logan Ryan has been pretty dog **** this year and Butler has been hit or miss compared to his 2015 campaign. McCourty has been fine although he did arrive just a bit late on Barnidge's TD pass. But that happens from time to time when you're responsible for as much as he is.

 

[long distance]

The defense has been solid, but the DBs, with the exception of Chung, have been average at best thus far. And that's being kind. McCourty has been strangely nowhere to be seen on quite a few deep balls this year, leaving struggling CBs isolated. Hopefully it'll all come together soon because if the secondary improved we'd be pretty tough to move the ball on.

McCourty is a beast and main reason our DB looks relatively ok.

 

[PF_SU]

I think McCourty's been playing well this season. There's a reason why we shutdown the deep passing game of the Texans and Cardinals without having a monster pass rush. Butler's been up and down, the Miami game was especially bad. Perhaps he was bothered a bit by an injury? Logan Ryan I think played fairly okay against DeAndre Hopkins, the other games, not so much.

Our secondary currently has a size issue. Hope Rowe can get on the field soon. His size can be valuable against tall WRs and some TEs.

 

[PScottman]

McCourty has been strangely nowhere to be seen on quite a few deep balls this year, leaving struggling CBs isolated.

Game planning has him doubling a certain player or one-on-one with a WR or TE vice playing center field (which also by the way is nowadays either split or left to Harmon as well) as much. So he's not supposed to be seen as it would mean he's leaving his own assignment to roam.

Another example was that he was closer to the line more (Cleveland being a run-heavy team) yesterday vice in deep cover-2 or high 3. He busted through for 1-2 TFL if I recall correctly.