Tannehill rated the 10th best quarterback in the league

[Dolfanalyst]

...Or quite simply stats are merely just stats and do not adequately indicate QB play nor do they adequately predict success or failure.

If that's true, then what does it mean that Peyton Manning has the most passing yardage and touchdown passes in the history of the league?  What does it mean that Aaron Rodgers has the highest career QB rating of all time in the NFL?  What does it mean that Dan Marino threw for over 5,000 yards and 48 TDs at a time when that was unheard of?  Nothing?  Are these statistics meaningless, and not indicative of these players' talent?

Don't throw the baby out with the bathwater.  Just because one statistic is garbage doesn't mean they all are.

[fyo]

The partial correlation between the "Overall" QB measure in the OP and win percentage, controlling for team points scored per game, is 0.06.

That's hardly surprising, since the quarterback doesn't play defense.

If you'd stop skipping the descriptions of what these statistical tests do, you might recall that the what "controlling for team points scored per game" means is that what is calculated is the correlation between winning and QB performance while keeping points scored constant. In other words, if it doesn't result in more points scored, the performance of the quarterback doesn't matter much in winning games. I think that's a reasonable statement that most would agree with and does not preclude a causal relationship between quarterback performance and winning.

[Dolfanalyst]

That's hardly surprising, since the quarterback doesn't play defense.

If you'd stop skipping the descriptions of what these statistical tests do, you might recall that the what "controlling for team points scored per game" means is that what is calculated is the correlation between winning and QB performance while keeping points scored constant. In other words, if it doesn't result in more points scored, the performance of the quarterback doesn't matter much in winning games. I think that's a reasonable statement that most would agree with and does not preclude a causal relationship between quarterback performance and winning.

That was exactly my point, however, that despite the diminished relationship between scoring and winning this year that you pointed out in the post of yours I quoted, the relationship between QB play and winning, while controlling for team points scored, is what would be expected.

In other words, QBs are still doing what they do with regard to winning, despite that points scored and winning are less related than usual.

The overall point is that we should still expect a measure of QB play to be correlated with winning, despite that points scored are less correlated with winning than usual.  Thus, the QB measure in the OP doesn't "get a pass" this year in terms of determining its validity.  There should still be a relationship between QB play and winning expected, as in other years.

[fyo]

The overall point is that we should still expect a measure of QB play to be correlated with winning, despite that points scored are less correlated with winning than usual.  Thus, the QB measure in the OP doesn't "get a pass" this year in terms of determining its validity.  There should still be a relationship between QB play and winning expected, as in other years.

I disagree. If the result of QB play is, primarily, to score points, then the correlation is entirely subsumed by the correlation between scoring and winning. I.e. if A -> B -> C (A = QB play, B scoring, C winning), then any reduction in correlation strength between B and C will result in a reduction between A and C as well.

The overall point, as I see it, is that we're simply dealing with small sample sizes. Scoring points hasn't suddenly become less important than in previous years.

[Pappy13]

If that's true, then what does it mean that Peyton Manning has the most passing yardage and touchdown passes in the history of the league?  What does it mean that Aaron Rodgers has the highest career QB rating of all time in the NFL?  What does it mean that Dan Marino threw for over 5,000 yards and 48 TDs at a time when that was unheard of?  Nothing?  Are these statistics meaningless, and not indicative of these players' talent?

Don't throw the baby out with the bathwater.  Just because one statistic is garbage doesn't mean they all are.

Anybody that WATCHED Marino play knew that he was special. Same goes for Peyton Manning and Aaron Rodgers. You don't need stats to tell you this and what the stats DON'T tell you is WHY they are special. Marino wasn't special because he threw a lot of TD passes, the reverse is true he threw a lot of TD passes because he was special and the reason he was special was because of the fire in his gut, his ability to elevate his play, his desire to be the best and of course his god given talents as well. Ask Marino how to play QB and he'll tell you "pick a guy a let it fly". That's not a joke, that's the way he played from the seat of his pants. He wasn't some master tactician, the game just made sense to him, he just saw things happening before they even happened. I think there's a lot of truth to that in Peyton Manning as well, it's well documented his ability to break down tape and study opponents, but in the heat of the moment, you don't think, you just act. It's instinct. These are things you can't teach, some people are just born to be QB's, but most are not. For those that aren't they can achieve a level of play through a lot of hard work and repetition, but they will NEVER be like Marino and Manning because they can't. It's just not in them and stats won't tell you that, watching them play and realizing that the stats don't define these players, that they transcend the stats will.

[Dolfanalyst]

Anybody that WATCHED Marino play knew that he was special. Same goes for Peyton Manning and Aaron Rodgers. You don't need stats to tell you this and what the stats DON'T tell you is WHY they are special. Marino wasn't special because he threw a lot of TD passes, the reverse is true he threw a lot of TD passes because he was special and the reason he was special was because of the fire in his gut, his ability to elevate his play, his desire to be the best and of course his god given talents as well. Ask Marino how to play QB and he'll tell you "pick a guy a let it fly". That's not a joke, that's the way he played from the seat of his pants. He wasn't some master tactician, the game just made sense to him, he just saw things happening before they even happened. I think there's a lot of truth to that in Peyton Manning as well, it's well documented his ability to break down tape and study opponents, but in the heat of the moment, you don't think, you just act. It's instinct. These are things you can't teach, some people are just born to be QB's, but most are not. For those that aren't they can achieve a level of play through a lot of hard work and repetition, but they will NEVER be like Marino and Manning because they can't. It's just not in them and stats won't tell you that, watching them play and realizing that the stats don't define these players, that they transcend the stats will.

Do you think that because it's possible to see how well (or not well) a QB can play, the statistics that he compiles are meaningless?

Surely you don't think it has to be one or the other, that either eyesight or statistics are valid, and not both?

[Dolfanalyst]

I disagree. If the result of QB play is, primarily, to score points, then the correlation is entirely subsumed by the correlation between scoring and winning. I.e. if A -> B -> C (A = QB play, B scoring, C winning), then any reduction in correlation strength between B and C will result in a reduction between A and C as well.

The overall point, as I see it, is that we're simply dealing with small sample sizes. Scoring points hasn't suddenly become less important than in previous years.

The issue there (the bolded portion) is that you're proposing a very simple causal path that discounts things such as YPA, interception rate, third down conversion rate, time of possession generated by sustained drives, etc., that have less than isomorphic relationships with scoring, but nonetheless have, especially when aggregated, significant relationships with winning, in that they prevent opposing teams from scoring.  Consequently you can have causal paths between QB play and winning that don't involve scoring.

The overall issue really comes down to whether you believe the interchangeability between Tannehill and Matt Ryan, Ben Roethlisberger, Drew Brees, Derek Carr, Philip Rivers, and Tom Brady implied by the QB measure in the OP is due to 1) a decrease in the relationship between scoring and winning in 2016, or 2) the simple invalidity of the QB measure, because the QB measure involves the same sample size (eight or nine games) as does the diminished relationship between scoring and winning in 2016.

[fyo]

The issue there (the bolded portion) is that you're proposing a very simple causal path that discounts things such as YPA, interception rate, third down conversion rate, time of possession generated by sustained drives, etc., that have less than isomorphic relationships with scoring, but nonetheless have, especially when aggregated, significant relationships with winning, in that they prevent opposing teams from scoring.  Consequently you can have causal paths between QB play and winning that don't involve scoring.

Are there elements that the offense (and the quarterback) can provide that help winning without scoring points? Sure, but there's nothing that even remotely compares in magnitude to SCORING. It is far, far away the most important aspect of an offense. Everything else is nice, but the effect is hard to pick out from the swamping signal of points scored.

[MyGodWearsAHoodie]

That is some serious GIGO going on.  You use Brady's qbr but the teams win % including games he wasn't even at the stadium.  Any conclusion drawn from that is pure and utter trash.

[Dolfanalyst]

That is some serious GIGO going on.  You use Brady's qbr but the teams win % including games he wasn't even at the stadium.  Any conclusion drawn from that is pure and utter trash.

Well, feel free to do the ultra-time consuming work of paring down win percentage by QB starts, for all the teams in the league, free of charge, and we'll make sure and sit here in the bleachers and critique your work.

[MyGodWearsAHoodie]

Well, feel free to do the ultra-time consuming work of paring down win percentage by QB starts, for all the teams in the league, free of charge, and we'll make sure and sit here in the bleachers and critique your work.

Not wasting my time. just because you did a bogus analysis doesn't require me to correct it, and even so running the data on half a season would be statistically meaningless even if done correctly. 

[Dolfanalyst]

Not wasting my time. just because you did a bogus analysis doesn't require me to correct it, and even so running the data on half a season would be statistically meaningless even if done correctly. 

No, it doesn't require you to do anything, but you might want to show a modicum of appreciation for the folks putting the work in and doing the best they can with it, rather than sitting up on your perch and providing only blowhard critique.

[MyGodWearsAHoodie]

No, it doesn't require you to do anything, but you might want to show a modicum of appreciation for the folks putting the work in and doing the best they can with it, rather than sitting up on your perch and providing only blowhard critique.

But you didn't do anything worth appreciating, using bogus data to hate on Tannihill when the REAL data says he is not the problem, is not worthy of appreciation.

[Dolfanalyst]

But you didn't do anything worth appreciating, using bogus data to hate on Tannihill when the REAL data says he is not the problem, is not worthy of appreciation.

Again, I think the differences in the ways we think about these things are going to make any further discussion unproductive.

[Dolfanalyst]

Anybody that WATCHED Marino play knew that he was special. Same goes for Peyton Manning and Aaron Rodgers. You don't need stats to tell you this and what the stats DON'T tell you is WHY they are special. Marino wasn't special because he threw a lot of TD passes, the reverse is true he threw a lot of TD passes because he was special and the reason he was special was because of the fire in his gut, his ability to elevate his play, his desire to be the best and of course his god given talents as well. Ask Marino how to play QB and he'll tell you "pick a guy a let it fly". That's not a joke, that's the way he played from the seat of his pants. He wasn't some master tactician, the game just made sense to him, he just saw things happening before they even happened. I think there's a lot of truth to that in Peyton Manning as well, it's well documented his ability to break down tape and study opponents, but in the heat of the moment, you don't think, you just act. It's instinct. These are things you can't teach, some people are just born to be QB's, but most are not. For those that aren't they can achieve a level of play through a lot of hard work and repetition, but they will NEVER be like Marino and Manning because they can't. It's just not in them and stats won't tell you that, watching them play and realizing that the stats don't define these players, that they transcend the stats will.

Perfect example here today of how what you see and the statistics that represent it can be one and the same.

Tannehill's having one of the best games of his career, and his YPA is a whopping 10.2, and his QB rating 146.1 at the end of the third quarter.