Slightly OT: Why did the Ravens abandon the air coryell system?

[InigoMontoya]

That is literally a difference of one single game.


No division has ever had a set of pass defenses as terrible as the 2014 NFC South. Not across the WHOLE division.

You haven't thought this thru. Your argument here, implies that we can never pay any attention to a QB's year-by-year stats: they can never tell us anything. That's way too strong a statement.


"Randomly", sure.


"Prove" is a strong word to use with statistical arguments. I don't need a p-value below a certain threshold. "Evidence" is enough for me.
Here's some:

Joe vs NFC South, 2014
4 games, 77 of 108 (71%) for 1134 yds (10.5ypa) with 11 TDa and 2 INTs, for a passer rating of 131.5.

For comparison, here's Joe's postseason run in '12:

4 games, 73 of 126 (58%) for 1140 yds (9.0ypa) with 11 TDs and 0 INTs, for a passer rating of 117.2.

The NFL's passer rating takes a lot of criticism from modern analytical approaches -- for decent reasons, it doesn't include sacks or fumbles, and it can't adjust for quality of opposing Ds. Fair enough. But it does provide a nice quick-&-dirty number for looking at per-attempt productivity. Passer rating uses four primary stats: completion %, yards-per-attempt, TD% and INT%. Some pretty important dimensions for a QB. Here's Joe vs the 2014 NFC South compared with Joe's 2012 postseason on those 4 metrics:

stat	12post	14NFCS
comp%	58%	71%
Y/A	9.0	10.5
TD%	8.7	10.1
INT%	0%	1.8%

Those zero INTs in the 2012 postseason jump out at you -- talk about clutch! -- but Joe's 1.8% vs the NFC South is excellent too. For comparison, Mahomes this year was at 2.1%. Drew Brees led the league in passer rating this year with 115.7 (lower than Joe's postseason run): Brees' INT% was at 1.0%.

On every other metric, Joe is distinctly better across the board in 2014. Better than the 2012 postseason! Joe played at an MVP level in the '12 postseason.
Here's the last 10 years of NFL passer rating leaders, with their efficiency stats, along with Joe's 2012 postseason:

Year	QB	Comp%	YPA	TD%	INT%	Rating
2018	Brees	74.4%	8.2	6.5	1.0	115.7
2017	ASmith	68%	8.0	5.1	1.0	104.7
2016	Ryan	70%	9.3	7.1	1.3	117.1
2015	Wilson	68%	8.3	7.0	1.7	110.1
2014	Romo	70%	8.5	7.8	2.1	113.2
2013	Foles	64%	9.1	8.5	0.6	119.2
2012	Rodgers	67%	7.8	7.1	1.4	108.0
2011	Rodgers	68%	9.2	9.0	1.2	122.5
2010	Brady	66%	7.9	7.3	0.8%	111.0
2009	Brees	71%	8.5	6.6	2.1	109.6
12post	Joe	58%	9.0	8.7%	0%	117.2
12NFCS	Joe	71$	10.5	10.1	1.8%	131.5

Joe's 2012 postseason fits right in with these other numbers. It's toward the higher side, but not at all outlandish: very similar to Matt Ryan '16. But! "One of these things is not like the others. One of these things just doesn't belong!"

Joe played out of his mind in the 2012 postseason. Historic. His stats vs the 2014 NFC South were distinctly better, across the board. Not "a little" better, not "slightly" better: a whole 'nother level.

And yet you all want to tell me that these numbers did NOT "distort" Joe's 2014 overall stats?

That's preposterous. It's retarded. It's not even worth entertaining.

I think some more compelling evidence would be if Big Ben, Andy Dalton, and whoever the Clowns had at QB in 2014 had similar differences in their stats against the NFC South. If those QBs lit it up against the NFC South in 2014, but were demonstratively worse vs. the rest of the league, then your argument holds more water to me.

[JimZipCode]

Which is why Kyle Shannahan would have been a perfect hire that made sense. If I am not mistaken he was available after Kubiak left and we went with Trestman. If I'm misremembering that I apologize.

You're right, he was available then – he was in the building! Byrne reported that he was one of the guys interviewed.

I have never understood why Harbs didn't pull the trigger on K Shanny. He would have been a great hire in '14 instead of Kubes; this is one of the things I have complained about before, that Harbs has tended to go with the "old" coordinator, even when a better option exists.

Shanahan was obviously the ideal hire in '15 after Kubes left. But I think the window was very narrow – not much time between Kubes deciding on Denver and Shanny hired in ATL.

[GreatWhiteNorthRaven]

That's preposterous. It's retarded. It's not even worth entertaining.

Ironic.

[Jazz1988]

Which is why Kyle Shannahan would have been a perfect hire that made sense. If I am not mistaken he was available after Kubiak left and we went with Trestman. If I'm misremembering that I apologize. But offensive coordinator hiring outside of kubiak has left a lot to be desired which is a problem.

Kyle Shanahan was actually not availalble when Kubiak left. I could be wrong but I think he was hire by The Falcons before even made their official announcement of their current head coach Dan Quinn.

[JimZipCode]

Kyle Shanahan was actually not availalble when Kubiak left. I could be wrong but I think he was hire by The Falcons before even made their official announcement of their current head coach Dan Quinn.

I think that's true too. He was available the season before, when our search dragged on and eventually landed on Kubiak.

Byrne wrote something on the Ravens site about Shanahan being in the building a couple times. I forget what the circumstances were.

[HotInHere]

That is literally a difference of one single game.


No division has ever had a set of pass defenses as terrible as the 2014 NFC South. Not across the WHOLE division.

You haven't thought this thru. Your argument here, implies that we can never pay any attention to a QB's year-by-year stats: they can never tell us anything. That's way too strong a statement.

...

That's preposterous. It's retarded. It's not even worth entertaining.

Okay, I'll try one more time, then I'm out.

Why the fuck does it matter that those four teams were in the same division? Let's say that in 2014 Joe played against the four worst defenses in the league, and they were the Panthers, Saints, Bucs, and Falcons. Let's say in 2015 they also played against the four worst defenses in the league, and they were the Raiders, Dolphins, and Browns twice.

Your 'analytical' method would exclude the Panthers Saints, Bucs, and Falcons from your 2014 numbers, and the Rams, 9ers, Seahawks, and Cardinals from your 2015 numbers. Do you see how that makes absolutely zero sense?

Honestly, I applaud you trying to use actual data to back up your case. And I believe your case might actually be right. And I don't expect something as in-depth as board stats gurus like Filmstudy and organizedchaos. I do, however, expect a basic understanding of statistical analysis if you're going to use it.

Grouping the NFC South together because all four bad defenses happened to be in the same division that year is just pointless. If you want to make your point, then make your point.

And btw, bringing a comparison to his 2012 postseason actually undercuts your point. You're trying to show how Joe got padded stats against weak competition in 2014, so we shouldn't count that as a good performance by Joe (and Kubes.) And then you point out how he was able to rack up amazing stats while playing against some of the very top defenses in the NFL in the 2012 playoffs. So which is it? Is it level of competition, or does is it good QB play and playcalling?

Look, I started out saying I think you may have a valid point. I'd actually love to see an analysis of the overall caliber of defense we faced in 2014 vis-a-vis the seasons surrounding it if you ever felt like doing it. Reach out to Filmstudy -- he would probably be happy to help.

But unfortunately, your current methodology is, to use your words, "preposterous and re..."; nevermind, I won't use your words.

[Cleo&Dilly]

Shanahan represented TOO BIG A THREAT to Harbaugh. He would NEVER VOLUNTARILY hire a guy who could develop into a viable head coaching alternative.

[HotInHere]

Shanahan represented TOO BIG A THREAT to Harbaugh. He would NEVER VOLUNTARILY hire a guy who could develop into a viable head coaching alternative.

Kubiak. Pagano. Caldwell. Pretty sure those guys became head coaches.

Go Broncos!


Sent from my iPhone using Tapatalk

[JimZipCode]

Grouping the NFC South together because all four bad defenses happened to be in the same division that year is just pointless. If you want to make your point, then make your point.

That is the point. Those bad defenses happened to be clustered there, and we got the advantage of playing them. Which is fine, all by itself; random stuff happens every year.

But the problem is, it coincided with a year that we switched offensive coordinators. People look at Joe's overall stat line that year, and conclude that Kubiak's WCO is/was a great fit for him. Maybe not so much. Maybe that conclusion was always based on a faulty premise.

You see? It wouldn't matter when Joe's "outlier" reg season games were – they're bound to be somewhere – except that we have been using those to drive an evaluation of which OC was best for the Ravens. Without even being aware of the distortion! "Flacco was at his best with Kubiak – if only we could get an OC in here like Kubiak – etc." Well, maybe that was all wrong.

Taking it further, it's hard not to wonder if that same statistical blindness influenced the next three years of coordinator choices for us. An unfounded theory of "great fit", built on a failure to notice that those four bad defenses happened to be in the same division.

And btw, bringing a comparison to his 2012 postseason actually undercuts your point. You're trying to show how Joe got padded stats against weak competition in 2014, so we shouldn't count that as a good performance by Joe (and Kubes.) And then you point out how he was able to rack up amazing stats while playing against some of the very top defenses in the NFL in the 2012 playoffs. So which is it?

I think pretty much all of us accept that Joe's performance in the 2012 postseason was Absolute Peak Joe. The four game streak that defines his career

In the four games against the NFC S in '14, Joe's stats were distinctly better than Absolute Peak Joe. Not by a smidge, either: distinctly better. Joe put up better numbers in those games than he did in the the best performance of his life – notably better. Probly just coincidence?

Look, I started out saying I think you may have a valid point. I'd actually love to see an analysis of the overall caliber of defense we faced in 2014 vis-a-vis the seasons surrounding it if you ever felt like doing it. Reach out to Filmstudy -- he would probably be happy to help.

The most galling part of this thread is people thinking they have to gently, condescendingly explain statistical analysis issues to me. Data mining, curve fitting, crystal ball effects, small sample sizes, serial correlation: yeah yeah yeah.

Keep it simple. Glance back at the chart in the first post of my thread. Look up Joe's 2 games with the highest passer rating of his career. Or his 10 games with the highest "Adj Yds / Att" (PFR's pet stat). Think about it a little.

[JimZipCode]

Kubiak. Pagano. Caldwell. Pretty sure those guys became head coaches.

And Rexy (though he was already here).

Quite a few ex-HCs too, like Leslie Frazier & Spagnuolo.

Roman got buzz for HC in previous cycles. Wink might too.

[HbgPARavenfan]

Which ironically is not what the philosophy of the WCO was.

....

they switched to limit downfield throws because our staff is turnover adverse