Ryan Tannehill

"AGuyNamedAlex wrote:

I think that is valid if you can say that all (most even) HOF QB have similar numbers. I'm not sure if that is or isnt the case, but I'd guess not.

What would matter is whether the variation among the numbers of Hall of Fame quarterbacks is significantly less than the variation between their numbers and the numbers of non-Hall of Fame quarterbacks.

What that would do is distinguish the two groups, Hall of Famers and non-Hall of Famers, from each other, statistically speaking.

So there could be variation in the numbers among Hall of Fame quarterbacks, as long as it was significantly less than the variation in the numbers between Hall of Fame quarterbacks and non-Hall of Fame quarterbacks. You wouldn’t need Hall of Fame quarterbacks to be statistically identical to each other to distinguish them statistically from non-Hall of Fame quarterbacks.

Just in case you guys didn't realise.

"AGuyNamedAlex wrote:

I think that is valid if you can say that all (most even) HOF QB have similar numbers. I'm not sure if that is or isnt the case, but I'd guess not.

They're all well above average statistically so you can just substitute "well above average" for HoF QB. Here are the career regular season only z-scores of all HoF QB's whose rookie year was 1966 or later:

Ken Stabler = 0.4051
Bob Griese = 0.7704
Dan Fouts = 0.8233
Warren Moon = 0.3908
Kurt Warner = 1.1918
Jim Kelly = 0.7422
Troy Aikman = 0.4316
Steve Young = 1.8627
Terry Bradshaw = 0.3516
Roger Staubach = 1.3849
Brett Favre = 0.6032
Dan Marino = 0.9476
John Elway = 0.3703
Joe Montana = 1.5602

The average is 0.8454 which is top 20th percentile (out of 32 QB's that's top 6-7). So really, instead of talking about HoF QB's you can just go back to argument about needing a QB that plays in that top 20% range often enough that you can expect to win a SB maybe once per decade (which is WAY better than we've had it!).

Oh.. something else. Eli Manning's career regular season z-score is -0.1381 lol. If he gets in due to those 2 playoff performances he will be a true statistical anomaly in the HoF, at least in the SB era. I hope he doesn't make it.

"The Guy wrote:

In a different thread about passer rating.

That's not at all what I said. I said that in a skills test like throwing at targets, dropping balls into buckets, those sorts of things, I'd be surprised if there's the massive variation that some of you think there would be, between the worst and the best. I'd even bet that there'd be some "bad QBs" that would do better than some "great QBs." The worst QB in the league is still elite at his position. It's why he's even on an NFL roster. These are guys who can do that shit in their sleep. I said where the variation I think would come, is in processing speed when things break down. The worst QBs respond worse under pressure of any sort than the best. Clearly some guys are better scramblers than others, but, were not talking necessarily about running here.

That's what I was saying about individual ability.

"The Guy wrote:

So again, which of the current or probable Hall of Fame quarterbacks who have won Super Bowls in the past 25 years are or will be in the Hall of Fame due to popularity and not individual ability?

You have to flesh that out if you are going to argue that Hall of Fame status is a measure of popularity and not individual ability. Otherwise you’re just talking theory and providing no support for it.

Dude. Eli Manning hasn't had, imo, a HoF career. But he'll get in for his name and the 2 rings.

"resnor wrote:

Dude. Eli Manning hasn't had, imo, a HoF career. But he'll get in for his name and the 2 rings.

I didn’t even include Eli Manning in the group of Hall of Fame quarterbacks involved in 20 of the last 25 Super Bowls. I counted him as one of the average quarterbacks who won a Super Bowl during that period.

"The Guy wrote:

I didn’t even include Eli Manning in the group of Hall of Fame quarterbacks involved in 20 of the last 25 Super Bowls. I counted him as one of the average quarterbacks who won a Super Bowl during that period.

So what? I'm saying, he's probably gonna be in the HoF, based on his name and 2 rings. It's speculation.

"resnor wrote:

So what? I'm saying, he's probably gonna be in the HoF, based on his name and 2 rings. It's speculation.

I’m not sure what point you’re trying to make.

"The Guy wrote:

What would matter is whether the variation among the numbers of Hall of Fame quarterbacks is significantly less than the variation between their numbers and the numbers of non-Hall of Fame quarterbacks.

What that would do is distinguish the two groups, Hall of Famers and non-Hall of Famers, from each other, statistically speaking.

So there could be variation in the numbers among Hall of Fame quarterbacks, as long as it was significantly less than the variation in the numbers between Hall of Fame quarterbacks and non-Hall of Fame quarterbacks. You wouldn’t need Hall of Fame quarterbacks to be statistically identical to each other to distinguish them statistically from non-Hall of Fame quarterbacks.

I didnt necessarily mean identical, just that they should be within the same range. Kind of what you're saying really.

"AGuyNamedAlex wrote:

I didnt necessarily mean identical, just that they should be within the same range. Kind of what you're saying really.

Then you intuitively came up with an analysis of variance to support your point. Good work.

https://en.m.wikipedia.org/wiki/Analysis_of_variance

"cbrad wrote:

I think it's more complicated than that.

On one hand there's evidence that "pressure on the QB" negatively affects passer rating, even if you define it like FO does. On the other hand, many mobile QB's that are good at evading pressure have very high sack% (e.g., Russell Wilson's career sack% of 8.37% is higher than Tannehill's 7.93% yet it's pretty clear RW > RT when it comes to pocket awareness). The QB's with the lowest career sack percentages are pocket passers with quick releases (Peyton and Marino are best all time at 3.13%).

This basically suggests that sacks and sack% are not good proxy measures for "pressure due to OL on the QB", and if that's true then you would expect a low (but still negative) correlation between sack% and passer rating. Nevertheless it's really hard to intuit how the correlation is essentially zero like with Mariota. That really surprised me.

So what’s being said here as I read it, and I think people should take note of this, is that sacks are more strongly related to quarterback play than they are to offensive line play.

"cbrad wrote:

Ultimately, stats and the observations that aren't captured by stats have to be consistent. So in any endeavor where you apply statistical analysis you have to ask the question "does this make sense?". And I attempt to do that.

For example, consider how I debated the issue of the degree to which passer rating becomes a measure of QB ability with large sample size. After acknowledging that passer rating is technically a team stat, I pointed out that the only constant in the formula itself is the QB and showed that the variance in win% over 5-year periods for individual teams is about half that of what you'd expect if team strength were randomly selected from game to game, which suggests that about half of the effect of surrounding cast is removed with large sample size.. but only half.

That's an example where I would argue I used "solid logic" with statistics, which really helps to properly interpret the stat. The hypothesis is consistent with the observation that passer rating doesn't account for all kinds of things it should if it were a better measure of QB ability while explaining the high correlation with how people tend to rate QB's with career (adjusted) passer rating.

Conversely, there are times where using "solid logic" I argue that the human observation component should be completely ignored and that you should only rely on stats. Best example was with the use of in-game win probabilities. You can easily do an experiment to show humans are terrible at estimating how much each play in a game contributes to the end result.

The issue with sack% is going more in the other direction, where I think there are good reasons to doubt its utility as a measure of "pressure due to OL on the QB" as I outlined in the post you quoted. However, just like with passer rating sack% has one constant in the formula itself, the QB, so it does reflect on the QB ability more and more over large sample size as evidenced by sack% being the highest correlated passing stat for a QB when he changes teams.

Either way, whenever I present stats I try to use what you call "solid logic" to properly interpret them. Challenge my interpretation if you want, but it's not like I'm ignoring what we see when I present stats nor am I inconsistent in my interpretations (it's not like I change them from argument to argument). I'm just cognizant of the weaknesses of both approaches. Humans (including experts) can't agree on what they see (and if you can't agree you don't have a "measure"), they have imperfect memories and are biased, etc... while with stats you have no such problems but you remove tons of information. Have to leverage both.

Great response.

I actually understood every word.

If I can get it, you'd think the rest of the folks here could also get it.

This should separate the interested fans from the self-interested fans!

"The Guy wrote:

I’m not sure what point you’re trying to make.

Yes you do.

"The Guy wrote:

So what’s being said here as I read it, and I think people should take note of this, is that sacks are more strongly related to quarterback play than they are to offensive line play.

And if stats tell you that poor oline play doesn't lead to sacks, then I'd question my stats.

"Irishman wrote:

Great response.

I actually understood every word.

If I can get it, you'd think the rest of the folks here could also get it.

This should separate the interested fans from the self-interested fans!

The quibble I've always had, with a lot of the statistical approaches used in here and in sports in general, is that macro stats are often used to make definitive conclusions about X player/unit/team, which ignores the reality that, very often, things happen counter to the statistical conclusion.

It's like gambling, sure over time, the house will win over all, but people still do win and if we made the same stat based arguments we do here about, say, winning the lottery, we'd be telling people that won the lottery that they didn't.

I miss the Phins league on NFL.com that I was in with you guys. That was fun.

"resnor wrote:

And if stats tell you that poor oline play doesn't lead to sacks, then I'd question my stats.

What the statistics in that post illustrated is that quarterback play is a moderating variable between pressure and sacks, which is an intuitive finding, given that quarterbacks vary in their awareness and evasion of pressure. Hopefully this isn't another piece of common knowledge that's going to be taken up as eminently and endlessly debatable.

"resnor wrote:

Yes you do.

Well if you're willing to leave it at that, it must not be that important.

"cbrad wrote:

They're all well above average statistically so you can just substitute "well above average" for HoF QB. Here are the career regular season only z-scores of all HoF QB's whose rookie year was 1966 or later:

Ken Stabler = 0.4051
Bob Griese = 0.7704
Dan Fouts = 0.8233
Warren Moon = 0.3908
Kurt Warner = 1.1918
Jim Kelly = 0.7422
Troy Aikman = 0.4316
Steve Young = 1.8627
Terry Bradshaw = 0.3516
Roger Staubach = 1.3849
Brett Favre = 0.6032
Dan Marino = 0.9476
John Elway = 0.3703
Joe Montana = 1.5602

The average is 0.8454 which is top 20th percentile (out of 32 QB's that's top 6-7). So really, instead of talking about HoF QB's you can just go back to argument about needing a QB that plays in that top 20% range often enough that you can expect to win a SB maybe once per decade (which is WAY better than we've had it!).

Oh.. something else. Eli Manning's career regular season z-score is -0.1381 lol. If he gets in due to those 2 playoff performances he will be a true statistical anomaly in the HoF, at least in the SB era. I hope he doesn't make it.

This is really what's noteworthy here in my opinion.

The very simple explanation for all this as I see it is that the league has evolved such that the modifications in the rules of the game prioritize the passing game. Teams that win are those that exploit those rules to their benefit, by having better personnel in the passing game. Obviously that increases considerably the importance of the single player who is involved in every passing play -- the quarterback.

It's no surprise that a vast majority of Super Bowls during that era have been won by the league's best quarterbacks, those who by virtue of their individual performance are far more likely to be inducted into the Hall of Fame.

The above really isn't debatable in my opinion. The game can be a team sport, and other parts of teams can be important as well, but those facts don't change the above.

This is why every team clamors for an elite QB. Why would they do that if it wasn't so important? If the fact that "it's a team sport" were more important, they wouldn't clamor for such QBs. They'd sit back and relax and assemble pieces of teams at random, rather than prioritizing certain pieces over others and paying them accordingly.

Again, none of this should be debatable in my opinion. It's common knowledge.

"The Guy wrote:

And I could swing the pendulum completely the other direction and say that Tom Brady is a terrible quarterback who’s done nothing individually during his career, has been nothing but the beneficiary of his surroundings, and that none of the Patriots’ success or Super Bowl wins have had anything to do with him at all.

And if I did that, then both of us would be making a strawman argument that no one in this thread on either side has stated.

Yes, but at least I'd agree with your statement. Brady has benefited from a really smart coach, a really smart lip reader, and a well conditioned supporting cast. I personally don't think he's anything special without those "added benefits".

"KeyFin wrote:

Yes, but at least I'd agree with your statement. Brady has benefited from a really smart coach, a really smart lip reader, and a well conditioned supporting cast. I personally don't think he's anything special without those "added benefits".

Hi Key! How are ya bud?

"KeyFin wrote:

Yes, but at least I'd agree with your statement. Brady has benefited from a really smart coach, a really smart lip reader, and a well conditioned supporting cast. I personally don't think he's anything special without those "added benefits".

Unfortunately there hasn't been any sample of play in which Brady has been divorced from those surroundings to permit such a conclusion (or its opposite) with any certainty.

Personally I would put my money on its not being an either-or, but a both, where both Brady and his surroundings are exceptionally good, which is likely the best explanation for any dynasty.

I don't think there is any situation in which a relatively poor or mediocre quarterback's surroundings can elevate him to career-long Hall of Fame-level play. You can get a very good season from that on occasion, like Andy Dalton in 2015 for example, but you're not going to get career-long Hall of Fame performance from it.

"The Guy wrote:

This is really what's noteworthy here in my opinion.

The very simple explanation for all this as I see it is that the league has evolved such that the modifications in the rules of the game prioritize the passing game. Teams that win are those that exploit those rules to their benefit, by having better personnel in the passing game. Obviously that increases considerably the importance of the single player who is involved in every passing play -- the quarterback.

It's no surprise that a vast majority of Super Bowls during that era have been won by the league's best quarterbacks, those who by virtue of their individual performance are far more likely to be inducted into the Hall of Fame.

The above really isn't debatable in my opinion. The game can be a team sport, and other parts of teams can be important as well, but those facts don't change the above.

This is why every team clamors for an elite QB. Why would they do that if it wasn't so important? If the fact that "it's a team sport" were more important, they wouldn't clamor for such QBs. They'd sit back and relax and assemble pieces of teams at random, rather than prioritizing certain pieces over others and paying them accordingly.

Again, none of this should be debatable in my opinion. It's common knowledge.

Teams want the best players possible at every position. You always try to boil things down to false dichotomies.

"resnor wrote:

Teams want the best players possible at every position. You always try to boil things down to false dichotomies.

And they assemble and pay those players at random? There is no prioritization in that regard? Guards and linebackers are chosen number-one overall and paid like quarterbacks?