Who is watching the games this year?

[Dave Gray]

And yet another problem with the statistic is, what is it measuring?

Is is measuring offensive line play?

Yes.

[Dolfanalyst]

Yes.

You don't know that with any certainty, and even if you did, the fact still remains that it's weakly correlated with passing statistics throughout the NFL.

So, for the sake of argument, imagine that it is in fact a measurement of offensive line play -- how do you explain the fact that there are QBs who have experienced just as much or more pressure, who have performed so much better than Tannehill?

[Spider-Dan]

I have to question your memory. One score is an early, fairly large lead?

Don't you remember?  A game can be in "garbage time" on your second offensive possession.

 

[Spider-Dan]

That may have been true for that particular game, in which the jobs of backup linemen were made more difficult by the team's surrendering an early, fairly large lead, and the Titans' pass-rushers' pinning their ears back.

So since we now agree that MIA did not "surrender an early, fairly large lead," what is the explanation for why the OLinemen were getting bowled backward?

Or are you now saying that the bowling pin linemen are not relevant?

[Dolfanalyst]

So since we now agree that MIA did not "surrender an early, fairly large lead," what is the explanation for why the OLinemen were getting bowled backward?

Or are you now saying that the bowling pin linemen are not relevant?

I'll concede that the backup linemen played poorly Sunday.

[Dolfanalyst]

Don't you remember?  A game can be in "garbage time" on your second offensive possession.

 

Actually it can't, because a score by the opposing team at that point would change the win probability a great deal.  Or did you miss that part in that discussion?

[Spider-Dan]

I'll concede that the backup linemen played poorly Sunday.

Sure, but (in your more recent posts) you appear to be implying that the pass-blocking effectiveness of the offensive line is of marginal relevance to the performance of the QB and/or the passing game.

So whereas originally the poor performance of the offensive line was significant... but caused by Tannehill's incompetence at producing points, now that we know the game was still close at that time, suddenly the performance of the offensive line really doesn't have much to do with the QB's effectiveness and we really shouldn't weigh it so heavily.  This is dynamic real-time cherry picking.

[Dolfanalyst]

Sure, but (in your more recent posts) you appear to be implying that the pass-blocking effectiveness of the offensive line is of marginal relevance to the performance of the QB and/or the passing game.

So whereas originally the poor performance of the offensive line was significant... but caused by Tannehill's incompetence at producing points, now that we know the game was still close at that time, suddenly the performance of the offensive line really doesn't have much to do with the QB's effectiveness and we really shouldn't weigh it so heavily.  This is dynamic real-time cherry picking.

We can certainly give the performance of the offensive line greater weight in an anomalous situation, without giving it the same weight in the far more typical situation.

Or are you saying the offensive line performed Sunday the way it had previously this season?

[Spider-Dan]

Pretty much, yes.

Through 5 weeks, MIA's OL is ranked #32 in pass protection by Football Outsiders.

[Dolfanalyst]

Pretty much, yes.

Through 5 weeks, MIA's OL is ranked #32 in pass protection by Football Outsiders.

From the page linked above:

Teams are ranked according to Adjusted Sack Rate, which gives sacks (plus intentional grounding penalties) per pass attempt adjusted for down, distance, and opponent. Pass blocking stats are explained further here. Our sack totals may differ slightly from official NFL totals depending on the league's retroactive statistical adjustments.

So of course that means the statistic is based largely on sacks.  With that said, how do we distinguish between the sacks attributable to the offensive line, and those attributable to Tannehill?

In other words, how do we know that statistic is measuring offensive line play exclusively?

[Spider-Dan]

Here you go.

Although once again, I find it funny that when evaluating offensive lines, we must employ a highly-nuanced consideration of all the extra factors that may play a part... but when it comes to evaluating QBs, straight YPA is plenty good enough to give a thumbs up or thumbs down!  We don't need to account for receivers, defenses, effectiveness of the run game, or any other mitigating factors; we simply look at league rank in, say, passer rating and then we can issue a verdict.

And once again, you are repeating the same cherry-picking of advanced stats: Football Outsiders' complex DVOA formula is good enough to be regularly cited by you, yet their rating of offensive lines is suspect and should be treated with a skeptical eye.

[Dolfanalyst]

Here you go.

Although once again, I find it funny that when evaluating offensive lines, we must employ a highly-nuanced consideration of all the extra factors that may play a part... but when it comes to evaluating QBs, straight YPA is plenty good enough to give a thumbs up or thumbs down!  We don't need to account for receivers, defenses, effectiveness of the run game, or any other mitigating factors; we simply look at league rank in, say, passer rating and then we can issue a verdict.

If you were to ask how we know YPA is a measure of quarterback play, I could answer that.

Now, if we were to consider for the sake of argument that Football Outsiders' "adjusted sack rate" is a measure of offensive line play exclusively (which I think is still very debatable, despite the information on the page linked above), there would still be another issue to contend with, and that is that it correlates with QBs' DVOA league-wide at only 0.28.

So, what that means is that there is roughly 92% of the variation in QBs' DVOA that isn't associated with adjusted sack rate, or, that there are QBs who do very well in terms of DVOA but are on teams with poor adjusted sack rates, as well as QBs who do very poorly in terms of DVOA but are on teams with very good adjusted sack rates.

So, ironically, if you're of the mind that Ryan Tannehill is hampered by his offensive line, you'd be better off believing that adjusted sack rate isn't a valid measure of offensive line play.  What you'd look for instead is some measure of offensive line play that actually correlated strongly with QB play, not weakly.

[Downunder Dolphan]

Only watching what is being shown on free to air here in Australia, and whatever highlights come up on the net - not worth paying for a subscription when we are this crappy.

So far they have shown just the one game (Patriots) - Dolphins just don't seem to get the coverage every year that other teams like the Broncos, Patriots, Cowboys, Packers & Giants do.