Eagles vs Patriots SBLII GDT

[BAConrad]

Also DVOA doesn't "isolate units" when rankings total defense. Don't know where the hell you got that from. @lancerman

[lancerman]

8 minutes ago, BAConrad said:

There's never been a team in history IIRC to win a SB with a Defensive DVOA as bad as the Pats' this season (ranked 31st in the league). And DVOA puts more weight into games played later in the season...when this Pat's defense was supposedly "dominant", as you or someone said. 

The worst in this century I believe was that Ravens defense in 2012 where they finished TWELFTH in the league that season. And the Pats defense is 2nd last.

Not saying DVOA is the end all be all but it's by far the best measurement of defense and offense ATM. Far better than PPGA 

 

Oh I know they have a bad DVOA. Last year had very poor DVOA as well. You just don't play the game by isolating offenses and defenses which is what DVOA does. DVOA is historically not a great indicator of how likely the defense is going to perform. PPG is more predicative because there are 3 phases of the game and each one compliments the other. Like a big part of the Patriots defensive success is a really strong special teams. However whoever plays the Patriots has to deal with that special teams and the field positioning advantage that comes with it. Another is the ability of a strong offense to chew up clock and keep the defense off the field. You still have to play the New England defense. 

I agree that DVOA is a better metric at purely isolating a unit and giving it an evaluation. It's however horrible at being predicative because football is a complimentary game and other units influence performance. Like it's all well in good to say the Patriots have the 31st DVOA. However that doesn't change that they are 13-3 and have the 5th best ppg in the league. You still have to account for that the Patriots have the best offense in DVOA and the 3rd best special teams in DVOA. It also could be a function. 

Also your numbers are off. The Patriots were 16th in DVOA last year and won the Super Bowl. They were also 12th in DVOA in 2014 when they won. They were 30th in 2011 and were also far worse in ppg and still were a hairlength from winning that one. They were 11th in 2007 and nearly went undefeated (and the team that beat them only scored 17 points). The Colts in 2006 were 25th in DVOA and they won the Super Bowl.

So while I agree that DVOA is good at isolating a defense, it's not very predicative.

[lancerman]

Just now, BAConrad said:

Also DVOA doesn't "isolate units" when rankings total defense. Don't know where the hell you got that from. @lancerman

Do you even know what I meant by that or are you just trying to start an argument.

[BAConrad]

Just now, lancerman said:

Do you even know what I meant by that or are you just trying to start an argument.

Reading your newest post I think I misunderstood. But either way , your point makes no sense. (I thought you meant they isolate positional units on defense)

Anyway, Why the hell does it matter? FO uses DVOA to measure teams total efficiency, as well as their offense and defense. 

By that logic, PPGA is also moot because it doesn't take offense or special teams into account. It's just an arbitrary stat telling how many points your TEAM gave up on a weekly basis. 

DVOA is much more complex and thorough.

It's not perfect. But you saying it isn't a good indicator of performance is laughable since the SB winners have always been one of the best teams in DVOA on a yearly basis. 

 

[BAConrad]

Honest question: if the Pats defense is as good as you say, why are they so low in DVOA? 31st is BAD. Why is it so low despite the fact that they were 5th (iirc) in PPGA? 

Tbh I don't remember ever seeing such a large discrepancy in these two statistics/formulas. 

DVOA is the best indicator there is ATM of how good a team is in general, or on offense, defense, etc. So how could an elite defense be ranked that low? 

[BAConrad]

Also, I admit I was mistaken about the Ravens being the lowest rated unit. I didn't check.

But, those defenses like Indy that season, Baltimore in 2012, etc, all were ones that weren't elite most season but then got extremely hot for the playoff runs. Like that Ravens team got back multiple star players from injury right before the postseason.

Check the weighted DVOA, which puts more value into games played later in the season. I'm certain that there will be a large difference.

Also, just one example of DVOA being a good indicator of future performance (since you claim it isn't) is the Eagles.

We were ranked 5th overall in total team DVOA last year despite winning 7 games. Our defense finished the year top 5. And then this year same thing except we win 13 games and make the SB. There's just one example off the top of my head of DVOA accurately predicting future success.

[lancerman]

7 minutes ago, BAConrad said:

Reading your newest post I think I misunderstood. But either way , your point makes no sense. (I thought you meant they isolate positional units on defense)

Anyway, Why the hell does it matter? FO uses DVOA to measure teams total efficiency, as well as their offense and defense. 

By that logic, PPGA is also moot because it doesn't take offense or special teams into account. It's just an arbitrary stat telling how many points your TEAM gave up on a weekly basis. 

DVOA is much more complex and thorough.

It's not perfect. But you saying it isn't a good indicator of performance is laughable since the SB winners have always been one of the best teams in DVOA on a yearly basis. 

 

That's what I mean though. 

DVOA looks strictly at the unit it is measuring to give you an indicator of it's strength. PPG doesn't. It's strictly based off points so it doesn't tell you if they are doing it because they consistently have good field position, or are usually playing with a lead, or have the benefit of time of possession. 

The reason why it matters is because you don't play a defense in isolation. The Broncos have the 10th best DVOA ranking. The Broncos also have the 22nd ppg ranking. The Broncos had a 5-11 record. So what was more indicative of their teams success the DVOA or the PPG? PPG is far more predicative. 

2016 the 16th best DVOA won the Super Bowl (#1ppg)

2014 the 12th best DVOA won the Super Bowl

2012 the 19th best DVOA won the Super Bowl

2011 the 19th best DVOA won the Super Bowl

2009 the 17th best DVOA won the Super Bowl

2007 the 13th best DVOA won the Super Bowl

2005 the 25th best DVOA won the Super Bowl

2001 the 13th best DVOA won the Super Bowl.

All taken from Football Outsiders.

[lancerman]

14 minutes ago, BAConrad said:

Also, I admit I was mistaken about the Ravens being the lowest rated unit. I didn't check.

But, those defenses like Indy that season, Baltimore in 2012, etc, all were ones that weren't elite most season but then got extremely hot for the playoff runs. Like that Ravens team got back multiple star players from injury right before the postseason.

Check the weighted DVOA, which puts more value into games played later in the season. I'm certain that there will be a large difference.

Also, just one example of DVOA being a good indicator of future performance (since you claim it isn't) is the Eagles.

We were ranked 5th overall in total team DVOA last year despite winning 7 games. Our defense finished the year top 5. And then this year same thing except we win 13 games and make the SB. There's just one example off the top of my head of DVOA accurately predicting future success.

Did the Patriots defense not get alot better as the season went on.

[iothar]

17 minutes ago, BAConrad said:

Also, I admit I was mistaken about the Ravens being the lowest rated unit. I didn't check.

But, those defenses like Indy that season, Baltimore in 2012, etc, all were ones that weren't elite most season but then got extremely hot for the playoff runs. Like that Ravens team got back multiple star players from injury right before the postseason.

Check the weighted DVOA, which puts more value into games played later in the season. I'm certain that there will be a large difference.

Also, just one example of DVOA being a good indicator of future performance (since you claim it isn't) is the Eagles.

We were ranked 5th overall in total team DVOA last year despite winning 7 games. Our defense finished the year top 5. And then this year same thing except we win 13 games and make the SB. There's just one example off the top of my head of DVOA accurately predicting future success.

What was the eagles dvoa every year for the last 15 years? I feel like every year they were near the top and failed to perform the following year like many predicted.

[lancerman]

21 minutes ago, BAConrad said:

Honest question: if the Pats defense is as good as you say, why are they so low in DVOA? 31st is BAD. Why is it so low despite the fact that they were 5th (iirc) in PPGA? 

Tbh I don't remember ever seeing such a large discrepancy in these two statistics/formulas. 

DVOA is the best indicator there is ATM of how good a team is in general, or on offense, defense, etc. So how could an elite defense be ranked that low? 

Okay if you want to stick with DVOA: Why does the team that supposedly is the second worst defense in the league according to DVOA, giving up the 5th least points, went 13-3 tying them for best record in the league, has only given up 20 or more points 3 times since the start of October, and is in the Super Bowl?

If DVOA is the best indicator, why was it so wrong here? Why have 5 of the last 12 Super Bowl winners been below the half way point in DVOA rankings? Why do the teams that normally win Super Bowls have a relatively high ppg ranking though? 

It's almost like DVOA is very imperfect just like any other stat. There's a lot of context in football. For instance the Patriots usually have a solid lead in most games. As a result they end up playing prevent while teams start going pass happy on them. Because of that they give up a lot of yards, but not points because they tighten up in the red zone. Etc. 

I'm not saying they are world class defense. I've never even claimed that. I said they are not as bad as they are being made out to be and to not sleep on them. In the stat that matters they've been top 5 and have playing at a high level for awhile. A lot of that probably has to do with their offense and special teams. Guess what? You have to play them too.  

PPG tends to be a much better indicator of how a defense will actually perform when the game comes. 

[lancerman]

6 minutes ago, iothar said:

What was the eagles dvoa every year for the last 15 years? I feel like every year they were near the top and failed to perform the following year like many predicted.

Here's the last 10 years defensive DVOA and the record the team finished with. 

2017: 5 (13-3) 

2016: 4 (7-9) 

2015: 17 (7-9)

2014: 10 (10-6)

2013: 23 (10-6)

2012: 26 (4-12)

2011: 11 (8-8) 

2010: 11 (10-6)

2009: 3 (11-5)

2008: 3 (9-6-1)

I see very little correlation between how the team finished and defensive DVOA. They were 4th last year and barely over .500 and the year before they were 17th and had the exact same record. They had the same record in 2014 and 2013 but they had a top 10 DVOA one year and 17th the other. However if you look at ppg it lines up a lot better with the record. 

[WeaponX]

7 hours ago, lancerman said:

It's a top 5 defense in ppg. That's just a fact.

The Patriots defense in 29th in yards allowed per game. And they are 31st in defensive DVOA. Those are just facts.

7 hours ago, lancerman said:

Well your basing it on very little if not nothing

See above.

7 hours ago, lancerman said:

I said they were better in ppg. They have been. 

If you choose to believe the Patriots defense will be as tough of a test for the Eagles as the Vikings were last week, sure, go ahead. In reality, the Vikings defense was 2nd defensive DVOA, 1st in yards allowed per game, and 1st in points allowed per game.

In contrast, the Patriots defense was 31st in defensive DVOA, 29th in yards allowed per game, and 5th in points allowed per game.

If you want to cherry pick one aspect of one stat (best PPG since week 5), let me present to you 2 other stats that present another part of the picture.

7 hours ago, lancerman said:

They probably aren't. If you isolated units to test them like DVOA does, they wouldn't. You just aren't playing an isolated Pats defense though. Your playing one of the best special teams and offenses in the league along with a defense that tends to come up clutch in the redzone.

Then like I said before, again, not sure what the point of this discussion is. The entire basis of this discussion is me saying that the Patriots defense will be the weakest unit on the field for this Super Bowl. Other units matter when you consider the team but I'm just talking about the defenses.

7 hours ago, lancerman said:

They have. Again that's a fact.

You said this in response to me saying PPG after Week 5 is just one metric. And that's true. The one metric is a fact. See above for DVOA and yards allowed per game. 

7 hours ago, lancerman said:

Okay one of their divisional teams ended up in the playoffs (non of the Eagles divisional teams did) and the best offense they played all season was in that 11 week stretch.

And that divisional team that made the playoffs was the Bills. And guess what their offensive PPG ranking was? 22nd. What about their offensive DVOA? 26th. What about their offensive yards per game? 29th.

Here's how the rest of the AFC East did (made up 6 out of the final 11 games):

Bills: 22nd PPG, 26th offensive DVOA, 29th yards per game

Jets: 24th PPG, 24th offensive DVOA, 28th yards per game

Dolphins: 28th PPG, 27th offensive DVOA, 25th yards per game

So it's nice that the Bills made the playoffs. But it doesn't change the fact that 6/11 games in that stretch you keep referencing for best PPG since Week 5 came against 3 of the worst offenses in football. 

7 hours ago, lancerman said:

No. I said the argument you made was used last year to discount their ppg ranking. That same argument didn't mean much at all when it came to having a successful playoff run and winning the SB. So far the Pats have already made the Super Bowl.

Again, what they did last year has absolutely no bearing on this game. And it shouldn't even in analyzing this game. There's been so much turnover for them. The Patriots defense looked good against the Titans. Not so much against the Jaguars.

7 hours ago, lancerman said:

I never disputed this Super Bowl is an entirely different game. I'm just saying that your arguments have no data to back them up. I was the only person who brought any metrics into this to prove my argument. You just said "I feel", "I say" etc. I had facts to back up my points, you just said you didn't believe them. You also said you thought "this feels different" because you thought this was the worst defense the Patriots brought to the Super Bowl. It's not even remotely true. By virtually every metric the 2011 defense was many magnitudes worse and they almost one.

I can't argue with someone who is just going to say "nah nah nah I don't think anything you say is true" and then not have anything to back it up. The facts are the facts. The Patriots defense is still a top 5 defense in ppg and ppg are historically one of the best indicators of how a defense will perform.

I actually have backed up my arguments with data. See above. Unlike you, I haven't cherry picked one aspect of one stat. You want metrics, you have them above. Read them. 

Here's the comparison in metrics between the 2011 Patriots and the 2017 Patriots:

Defensive DVOA: 2011 (30th), 2017 (31st) - Actually, being lower in ranking doesn't indicate "many magnitudes worse".

Yards Allowed Per Game: 2011 (31st), 2017 (29th) - Both defenses pretty bad. 

PPG: 2011 (15th), 2017 (5th) - This is where there's a difference that shows the 2011 defense was worse.

Again, an exaggeration to say 2011 was magnitudes worse. They're similar. (But did you look at 2011 metrics after week 5 though?)

 

I've given you facts. The basis of your argument has been an extremely cherry picked statistic: best PPG after week 5. I've tried to explain that all 6 division games came after Week 5 and were against the worst statistical offenses in football this year. And now I've given you DVOA and yards per game. 

[WeaponX]

@lancerman, you use that Broncos example of discrepancy between DVOA and PPG. 

Why do the Broncos have a good DVOA ranking but poor defensive PPG ranking?

Perhaps it has something to do with how the Broncos having the 2nd most turnovers in the NFL. 

The Broncos defense is certainly talented. We all know that. Their poor defensive PPG ranking is certainly affected by how many times they turned over the ball.

The fact that it is the inverse for the Patriots should raise even more questions about the Patriots defensive unit. 

In fact, the Patriots are 31st in defensive DVOA. But they are 5th in defensive PPG. 

And maybe that has something to do with the fact that the Patriots had the 2nd least turnovers in the NFL this season.

[mcmurtry86]

The Pats defense isn't good but it isn't terrible. 

They're good at preventing points, which is their primary function. Their inability to get off the field at times means that they put too much pressure on the Pats offense (long fields, point deficits) and in a one game sample size (like the Super Bowl) it's easy to project higher points for their opponent since it's really only a matter of one or two extra good plays to turn a long 3 point drive into long 7 point drive.

Furthermore - football is all about matchups and game planning. The Pats can't stop the run and struggle on short passes. If the Eagles exploit that and Foles is half decent, they should pile up points at a rate higher than what the Pats gave up in the regular season.

[BLick12]

Well, it's a combination of things, right?  I'm a pretty strong endorser of DVOA, and while it isn't an absolute for predicting games, it does tell you a lot about a team's defense.  For a team like the Pats that has an elite offense that hardly turns the ball over and a good STs unit that doesn't often negatively impact field position, they can still win with a below average DVOA. It's not going to be the ultimate predictor of outcome, because it is only accounting for one side of the ball.  But in my opinion, it is a robust and more accurate statistic for analyzing historical performance than anything else out there or any of the basic stats like PPG.

Someone mentioned the Broncos and DVOA not being a good predictor for them. But you're viewing it as a fortune telling solution, it is not that.  What it tells you is that their defense still performed fairly well and there are likely additional factors that led to them struggling... like for instance, a highly detrimental offense.  @RollEagles basically nailed it.