Tannehill: where he ranks stats wise in different categories

[Nawledge]

How does Phillip Rivers numbers correlate to winning? That **** is nonsense shouright.

 

[Shouright]

can you explain the stat correlation to me? like really...its nonsense. If a QB has great numbers but his defense sucks and the team goes 8-8 how do those stats correlate to winning?

Correlations are based on more than just a sample of one. When you take hundreds upon hundreds of QBs and teams over years and years and you find that as a QB's YPA, for example, increases, so does his team's winning percentage the vast majority of the time, you have a meaningful correlation.

Now, that doesn't mean there aren't exceptions to the rule, such as when the 2010 San Diego Chargers had an adjusted net YPA of 7.8 and didn't make the playoffs, which is what you're highlighting with your example above. This is why I said no correlation is perfect.

Here's a good article on it if you're interested:

http://www.footballperspective.com/correlating-passing-stats-with-wins/

 

[NBP81]

Correlations are based on more than just a sample of one. When you take hundreds upon hundreds of QBs and teams over years and years and you find that as a QB's YPA, for example, increases, so does his team's winning percentage the vast majority of the time, you have a meaningful correlation.

Now, that doesn't mean there aren't exceptions to the rule, such as when the 2010 San Diego Chargers had an adjusted net YPA of 7.8 and didn't make the playoffs, which is what you're highlighting with your example above. This is why I said no correlation is perfect.

But still, the stats you use are just a manipulation of the raw stats in the op... So if doing well in most of them, there`s a good chance you`ll do well in the ones you use... But like you said... there are exceptions you know...

 

[Shouright]

How does Phillip Rivers numbers correlate to winning? That **** is nonsense shouright.

Well you're right in that there are other parts of the team involved, as well. The Chargers on defense are surrendering some of the worst quarterback play in the league, in terms of the QB variables most strongly correlated with winning. So what I'm saying about the correlation between QB variables and winning still holds true in a roundabout way, because not only do you need good QB play to win, but you need to defend QB play, as well.

See here:

http://www.nfl.com/stats/categoryst...TEAM_PASSING&d-447263-s=PASSING_PASSER_RATING

If you notice, the Chargers are surrendering an opposing QB rating of 101.6 and an opposing YPA of 8.5, on average. Those numbers are horrible, and they're too poor for Philip Rivers to surmount despite his very good play, hence their mediocre record.

Now, if the Chargers didn't have Philip Rivers, they'd look a lot more like the Jacksonville Jaguars out there, which again illustrates my point about QB play and its relationship with winning. Rivers alone is able to keep the Chargers afloat despite their horrendous pass defense.

 

[Shouright]

But still, the stats you use are just a manipulation of the raw stats in the op... So if doing well in most of them, there`s a good chance you`ll do well in the ones you use... But like you said... there are exceptions you know...

It's the "manipulation" of the stats (as you put it) that gets you something more meaningful in terms of a correlation winning. It isn't just yards or attempts, which are measures of volume, but yards per attempt, which is a measure of efficiency. In other words, passing efficiency is more valuable in terms of winning than passing volume.

 

[Nawledge]

Well you're right in that there are other parts of the team involved, as well. The Chargers on defense are surrendering some of the worst quarterback play in the league, in terms of the QB variables most strongly correlated with winning. So what I'm saying about the correlation between QB variables and winning still holds true in a roundabout way, because not only do you need good QB play to win, but you need to defend QB play, as well.

See here:

http://www.nfl.com/stats/categoryst...TEAM_PASSING&d-447263-s=PASSING_PASSER_RATING

If you notice, the Chargers are surrendering an opposing QB rating of 101.6 and an opposing YPA of 8.5, on average. Those numbers are horrible, and they're too poor for Philip Rivers to surmount despite his very good play, hence their mediocre record.

Now, if the Chargers didn't have Philip Rivers, they'd look a lot more like the Jacksonville Jaguars out there, which again illustrates my point about QB play and its relationship with winning. Rivers alone is able to keep the Chargers afloat despite their horrendous pass defense.

so what happens when a qb sucks and the defense and run game take a team to 2 AFC championship games...your correlation to winning stats are meaningless.

 

[FinfanInBuffalo]

It's the "manipulation" of the stats (as you put it) that gets you something more meaningful in terms of a correlation winning. It isn't just yards or attempts, which are measures of volume, but yards per attempt, which is a measure of efficiency. In other words, passing efficiency is more valuable in terms of winning than passing volume.

And passing efficiency (for or against) is a team stat.

 

[CSONKA1966]

Except for signing Wallace ,we have not done Tanny any favors this year no running game ,the guy was sacked more than anybody the first half of the season ,then you throw the whole Martin mess on his plate. Yet the Kid keeps progressing and we are in the mix of things. I have been a fan since 1971 and I think we might have something great here

 

[Shouright]

so what happens when a qb sucks and the defense and run game take a team to 2 AFC championship games...your correlation to winning stats are meaningless.

Like I said, no correlation is perfect. There are exceptions to the rule. However, you want going for you what gives you the best chance to win, and the individual stats that are most strongly correlated with winning are just that, despite the inevitable exceptions.

 

[Nawledge]

Like I said, no correlation is perfect. There are exceptions to the rule. However, you want going for you what gives you the best chance to win, and the individual stats that are most strongly correlated with winning are just that, despite the inevitable exceptions.

so you just admitted that the stats u speak of are meaningless.

 

[NBP81]

It's the "manipulation" of the stats (as you put it) that gets you something more meaningful in terms of a correlation winning. It isn't just yards or attempts, which are measures of volume, but yards per attempt, which is a measure of efficiency. In other words, passing efficiency is more valuable in terms of winning than passing volume.

I know, all IM saying is your stats still come from his stats so if Tannehill is doing relatively well in in his stats, there`s a good chance he`ll do well in yours...get with the program bro!

 

[Shouright]

And passing efficiency (for or against) is a team stat.

It appears to be correlated fairly strongly with the consensus perceptions of QBs' individual ability, however, despite the inevitable exceptions to the rule:

http://www.pro-football-reference.c...pos_is_db=Y&draft_pos_is_k=Y&draft_pos_is_p=Y

---------- Post added at 08:40 PM ---------- Previous post was at 08:39 PM ----------

so you just admitted that the stats u speak of are meaningless.

I think we're done now. :)

---------- Post added at 08:40 PM ---------- Previous post was at 08:40 PM ----------

I know, all IM saying is your stats still come from his stats so if Tannehill is doing relatively well in in his stats, there`s a good chance he`ll do well in yours...get with the program bro!

Not necessarily, bro. :)

Later folks. Have a great evening. :up:

 

[FinfanInBuffalo]

It appears to be correlated fairly strongly with the consensus perceptions of QBs' individual ability, however, despite the inevitable exceptions to the rule:

It is even more strongly correlated to the QB's ability combined with a strong supporting cast just as the variability year by year shows. It is a team sport after all. Something you seem to have a hard time grasping.

 

[NBP81]

It appears to be correlated fairly strongly with the consensus perceptions of QBs' individual ability, however, despite the inevitable exceptions to the rule:

http://www.pro-football-reference.c...pos_is_db=Y&draft_pos_is_k=Y&draft_pos_is_p=Y

---------- Post added at 08:40 PM ---------- Previous post was at 08:39 PM ----------

I think we're done now. :)

---------- Post added at 08:40 PM ---------- Previous post was at 08:40 PM ----------

Not necessarily, bro. :)

And why would that be? For example, your favorite y/pa stat... It takes into account, attempts, completions, yards(all of those are in the op) but adds sacks and yards loss(which are certainly debatable as to how much of a responsibility falls on the QB)... So yeah it becomes more of a team stat than anything, even if you wont admit it...

 

[Nawledge]

And why would that be? For example, your favorite y/pa stat... It takes into account, attempts, completions, yards(all of those are in the op) but adds sacks and yards loss(which are certainly debatable as to how much of a responsibility falls on the QB)... So yeah it becomes more of a team stat than anything, even if you wont admit it...

his stats correlating to wins is pointless, I provided 2 examples of why they don't mean **** and his response was "no correlation is perfect"...well then why should anyone give a ****?