Patriots have the 12 ranked D 1/4 into the season - Discuss

[ivanvamp]

You can't "take away" any one game that "skews" the stats, or you'd have to do that for everyone. So what you do is calculate how the Pats' opponents have fared against their non-Patriots' opponents (on average), and then compare. So, through four games....

Houston
vs. NE: 20 points, 325 yards, 21 first downs
vs. other: 25.3 points, 443.3 yards, 24.0 first downs
DIFF: -5.3 points, -118.3 yards, -3.0 first downs

Jacksonville
vs. NE: 31 points, 480 yards, 27 first downs
vs. other: 19.0 points, 346.7 yards, 17.3 first downs
DIFF: +12.0 points, +133.3 yards, +9.7 first downs

Detroit
vs. NE: 26 points, 414 yards, 25 first downs
vs. other: 22.3 points, 382.7 yards, 21.0 first downs
DIFF: +3.7 points, +31.3 yards, +4.0 first downs

Miami
vs. NE: 7 points, 172 yards, 11 first downs
vs. other: 25.0 points, 324.0 yards, 16.3 first downs
DIFF: -18.0 points, -152.0 yards, -5.3 first downs

NET DIFF: -7.6 points, -105.7 yards, +5.4 first downs

So they're better than average in points and yards, and worse than average in first downs. That sounds pretty much like a team whose defense should be ranked somewhere between 12th and 14th.

 

[RobertWeathers]

Nice stats but I always reserve judgement til late November.
Do people eat pudding on Thanksgiving?
I guess the proof will be in the pudding.

No.

My sister in law makes a kick-ass apple and pumpkin pie.

My mom makes a lemon meringue and a key lime pie.

I'll typically have a piece after dinner and another a couple of hrs later.

 

[patsinthesnow]

In 2011 & 2017 the team was still next to dead last in yards per game, 3rd down defense, passer rating etc. despite facing mediocre competition. It’s a step forward that they’ve delivered Watson/Tannehill’s worst games in 10+ starts with Flowers on the field. Still remains to be seen what the unit can/will be.

 

[jimnance]

No.

My sister in law makes a kick-ass apple and pumpkin pie.

My mom makes a lemon meringue and a key lime pie.

I'll typically have a piece after dinner and another a couple of hrs later.

Yeah, I've never seen pudding at a Thanksgiving meal.
I was just trying to stay on topic.

 

[Ring 6]

In 2011 & 2017 the team was still next to dead last in yards per game, 3rd down defense, passer rating etc. despite facing mediocre competition. It’s a step forward that they’ve delivered Watson/Tannehill’s worst games in 10+ starts with Flowers on the field. Still remains to be seen what the unit can/will be.

No they weren’t.

Every team on average faces mediocre competition.
I keep waiting to hear about all these teams who allow more points are better because somehow they are only scheduled against top 10 teams.

 

[RobertWeathers]

Yeah, I've never seen pudding at a Thanksgiving meal.
I was just trying to stay on topic.

I'm one of the biggest offenders on this board to not staying on topic. I do that once it seems we've exhausted the discussion.

Back to the discussion....

 

[Palm Beach Pats Fan]

I guess the proof will be in the pudding.

The proof of the pudding is in the eating.

That's how that saying used to go, but it's gotten twisted around over the years.

the proof of the pudding is in the eating - Wiktionary

 

[Deus Irae]

You can't "take away" any one game that "skews" the stats, or you'd have to do that for everyone.

Noting a statistical outlier is a perfectly valid thing to do. The problem here isn't the existence of that game's stats. The problem is the existence of that game's stats in what is just a 4 game sample.


And I do like to make a homemade version of key lime pudding.

 

[Horace's Ivories]

I am pretty optimistic about the secondary continuing to improve and with depth..

baked granola and chocolate chip "bread" pudding w/ maple syrup

 

[Nunchucks]

Small N allows outliers to to carry more weight than they should. Playing offensively challenged teams will not help.

Miami was awesome, lets hope the trend continues, not give the Patriots an "atta boy" because we all know the defense looked TERRIBLE for two games. Not the "can't cover anybody" terrible, but the "the guy in front of me keeps kicking my arse all over the filed" terrible.

Lets not use statistics to cover for what we saw with our own eyes.

 

[One-If-By-Sea]

I am not cherry picking any game. I am saying the statement about the Pats having the 12th best D is useless since the difference between the 11th and 27th best is very small anyway.

It does not mean it is useless. The Pats are the 12th best D. A very small difference between 11 to 27 means the variance or deviation is small, so the ranking could swing week to week, or the different defenses are similar in ability. If the number of games played, the sample size, increases and the definition becomes less variable, that is not a bad thing either. But the Pats are 12th right now and that small sample size statistic that has a small deviation is not useless. It is what it is, minimizing or exaggerating it serves no purpose.

 

[One-If-By-Sea]

Noting a statistical outlier is a perfectly valid thing to do. The problem here isn't the existence of that game's stats. The problem is the existence of that game's stats in what is just a 4 game sample.


And I do like to make a homemade version of key lime pudding.

You can't really note an "outlier" in a 4 sample statistic.

I am traditional - the best is fresh picked apples in a homemade apple pie with vanilla bean ice cream.

 

[Kdo5]

We'll really see where they are at when they take on KC. I am very skeptical and the Miami game isn't going to change that.

 

[ivanvamp]

Noting a statistical outlier is a perfectly valid thing to do. The problem here isn't the existence of that game's stats. The problem is the existence of that game's stats in what is just a 4 game sample.

Yes you can take away the Pats' best game, so long as you do that for everyone else too. Just taking it away from the Pats, but allowing other teams their best defensive games, skews the data.

And I do like to make a homemade version of key lime pudding.

I don't care.

 

[Deus Irae]

You can't really note an "outlier" in a 4 sample statistic.

Sure you can, particularly when it's as obvious as this set of 4. But, as I noted, the problem is the small sample size, regardless of the existence of any potential outliers.

Yes you can take away the Pats' best game, so long as you do that for everyone else too.

You don't need to do it for everyone else, because you're noting the outlier for context, not to go back and redo the numbers. Saying you have to go back for every other team isn't even remotely accurate.

Just taking it away from the Pats, but allowing other teams their best defensive games, skews the data.

Context. It's all about context. It's not about trying to re-slot. Other teams are irrelevant.

 

[Kontradiction]

Small N allows outliers to to carry more weight than they should. Playing offensively challenged teams will not help.

Miami was awesome, lets hope the trend continues, not give the Patriots an "atta boy" because we all know the defense looked TERRIBLE for two games. Not the "can't cover anybody" terrible, but the "the guy in front of me keeps kicking my arse all over the filed" terrible.

Lets not use statistics to cover for what we saw with our own eyes.

Agree but N = population size, n = sample size. In the Patriots' case, n usually includes 6 games against the Dolphins, Jets, and Bills who are traditionally anywhere from terrible to mediocre... especially on offense.

 

[PATRIOTSFANINPA]

if only the team played every game against a Ryan Tannehill offense the stats might resemble the 86 Bears - too bad its not possible

 

[Kontradiction]

By the way, I've weighed in other stats in my thread to give us a better idea of what we're really looking at and have compared the to-date offensive rankings of our competition for context. Right now, we're looking at a middle of the pack defense that has faced below average offensive competition.

 

[Bruins29]

By the way, I've weighed in other stats in my thread to give us a better idea of what we're really looking at and have compared the to-date offensive rankings of our competition for context. Right now, we're looking at a middle of the pack defense that has faced below average offensive competition.

BTW I did not mean to supersede your thread.

 

[Kontradiction]

BTW I did not mean to supersede your thread.

No worries. Ian pays for the bandwidth anyway.