Tua is not the Problem

Pass block win rate is a measure of the offensive line. Period.

 

Numbers need context.

 

"The number one thing that creates more wins is passer rating."

I fundamentally and completely disagree. Passer rating is a result. It isn't decided until the final whistle. Passer rating doesn't do ****. It doesn't throw the ball, it doesn't score touchdowns. Nothing. Rating is a result that attempts to numerically represent everything the QB does.

Game prep has nothing to do with statistics, other than as a numerical representation of what the other team is good or bad at. It is a numerical quantification of what you can see with your eyes when watching the games.

 

Partly, but it’s not all of it. If a QB holds the ball longer than expected, rolls out a different direction or is late, calls out the wrong protections…all affect line play. There many other examples.

 

That is not how pass block win rate is calculated.

 

Have you been to elementary school lately? I can confirm that they don't do the same math we did in elementary school.

 

How isn't 30/30 for 500 yards better than the 10/10 for 100 yards?

If your model is treating these as the same impact on win %, then something is wrong with your model.

 

Dude.. near zero correlation is the end of the argument.

Still, 30/30 for 500 yards has a higher passer rating (118.75) than 10/10 for 100 yards (108.3). If you're interested in why look up passer rating. Anyway, I think we're done with this debate now.

 

But that isn't true. You are assuming both were in wins. Again, if 5 of those are 60 yard bombs for TDs, then yes, that QB most definitely massively impacted win%.

 

Pass block win rate measures how long a lineman can sustain his block. I "think" (going off of memory) they've created this stat with about 2.5 to 3 seconds as the 'win". It's not a very good stat. What if the play in question requires the QB to get rid of the ball faster than 3 seconds? Is that a fail because the olineman didn't block for that amount of time? What if a play requires the oline to block for 3.5 seconds? What if the QB is sacked at 3.2 seconds? That's a fail, but according to this stat, it's a win. Different passing plays need different amounts of time. I appreciate the attempt to measure oline play, but this doesn't pass the muster.

Bottom line, QB play affects oline play. There's no arguing this fact. RT has shown that he can inconsistently play at a high level...most NFL QB's can do that. That's why they're in the NFL. What RT cannot do is play at a high level most of the time. What's really funny is that you've mentioned the number "18" about 100 times now. In 2019, RT's best season, his two best games, he threw for 18 and 19 times. In fact, in 5 out of his 12 games played that season he threw the ball less than 22 times. lmao...We should disregard those games according to you. If we did, instead of finishing the season with a passer rating of 117, he would have had around a 95.

 

No dude... The more efficient a QB is the more likely his team is to win. That efficiency is measured as passer rating. Of course it isn't perfect, but it's proven time and time again to be true and the best we have.

If you've been watching football you'll consistently hear about "analytics". What they are talking about are the PREDICTIVE analytics.

"Predictive analytics is the use of data, statistical algorithms and machine learning techniques to identify the likelihood of future outcomes based on historical data. The goal is to go beyond knowing what has happened to providing a best assessment of what will happen in the future."

The NFL hires people to do math for all type of situations... "Should I punt on 4th and 1 or go for it?" There's a metric for that. There's a model that shows the odds of what happened in the past to PREDICT the future.

 

I think there is a disconnect here. Basically, if your eyes are telling you that Tua isn't playing great, but he is leading the league in PR, then you are wrong. Now there are caveats, like PR needing volume to be accurate. I believe it was originally developed to grade a seasons worth of play. So a game or two with a high PR isn't that great of an indicator of overall play. If I remember correctly, Cbrad uses 150 attempts as a general minimum to have a meaningful score. The reason that he uses PR amongst all of the various stats is its high correlation to winning and it's transparency. ESPN has QBR, and claims it is a more nuanced measurement that also is highly correlated with winning, but who knows how it is calculated? They admit that they update the formula periodically but not what it is. If I am wrong about any of that I hope cbrad corrects me. This is my understanding after reading a bunch of posts.

If you want to argue that you see something that means his play isn't sustainable, it doesn't change what has happened so far, and so far Tua has been great. I personally want a full season to call him great. I am super stoked he has been playing like he has.

 

While we're on the subject of efficiency stats, there are two other simple ones that are essentially as good as passer rating, just not as well known. They're very similar to passer rating except that neither includes completion percentage and one includes sack percentage.

AYA (adjusted yards per attempt) = (Passing yards + 20*Passing TD - 45*Interceptions)/(Passing attempts)
ANYA (adjusted net yards per attempt) = same as AYA except you subtract sack yards in the numerator and add # sacks in the denominator.

ANYA actually beats out passer rating by a razor thin margin (nothing statistically significant) in terms of correlation to win% in the last 20 years: 0.68 to 0.675, with AYA coming in right behind them at 0.66. This shows that including sacks is marginally more informative than including completion percentage.

Guess who leads right now in AYA and ANYA? Yup.. Tua.

 

That kind of makes sense. I can see how yards would trump comp%. Is a 50% comp% with 400 yrds passing more efficient than 60% comp% with 300 yards?

 

I think this righty vs lefty perception bias theory is a lot of BS. People don't say Tua has a weak arm because he is a lefty, they say it because he has a weak ****ing arm compared to other NFL QBs. Michael Vich was a lefty and no one ever called him weak armed, because he actually had a strong arm, Tua does not. Has nothing to do with lefty perception bias.

If a ball is underthrown it's underthrown, doesn't matter if you are a righty or lefty you will be labeled weak armed if you are regularly underthrowing balls or throwing with poor velocity.

People come.up with all sorts of crap for why people **** on Tua. It's not because he is a lefty. It's because he has a relatively weak arm, and is injury prone. There in no great mystery or perception bias going on here.

 

No stat should be used in isolation. Stats should not be used without film study either. All the stats and film study from 2021 confirm the evaluation that the Titans OL was terrible at pass blocking.

You've responded to the wrong guy with this one. You were debating with cbrad about pass attempts. But, I'll respond.

Over a 26 regular season game period (2019 - 2020), Tannehill had the highest passer rating in the league. Let that sink in. Not just a good passer rating, the highest. Nearly two full seasons. He did not enjoy the best supporting cast or the easiest plays to make. These assertions can be supported with independent stats.

In 2021 and 2022 the supporting cast declined significantly. This can be confirmed by looking at the injuries, player losses, various independent evaluations of the players in the other positions, and by watching the film. It is no coincidence that the drop off in Tannehill's play (both volume and efficiency) coincides with the drop off in talent and the loss of their OC.

My opinion of Tannehill has not changed. He is a very good QB that is capable of elite level efficiency. What separates him from the elite QBs (IMO) is that the situations in which he can perform at that level are narrower than the elite QBs. Primarily, he needs a bit better pass protection than guys like Mahomes, Rodgers, and Josh Allen. He doesn't need great pass protection (he has never had great pass protection), just average. What separates him from average QBs is that his ceiling is higher in those situations. A passer rating of 117 with a YPA average of 9.6 is very very rare. I still think Tannehill is a top 8 - 12 QB in talent overall.

BTW, my arguments are equally applicable to Tua. That is the beauty of my thought process, I don't need to twist it depending on if the QB plays for my team or not. Tua's first two seasons were "Tannehill in a bad situation" like numbers. Not a surprise since he was in a similarly bad situation. Improve the pass blocking, play calling and receivers and, just like magic, his play improves..... wow.

Tua's efficiency this season is also in the very very rare category. Passer rating of 112 and YPA of 9.6. Look at the next gen stats and compare passer rating and CAY. There are only 3 QBs in the top 10 of CAY with a passer rating over 100 (Tua, Allen, Geno Smith). Their passer ratings are lower and CAY are significantly lower. Only one other QB (Winston) has a CAY over 8 and his passer rating is 79. It is very very difficult to maintain a high level of efficiency while pushing the ball down the field.

It is not hard to understand.

 

I highly, highly encourage you (no, I dare you) to go here - https://www.titansreport.com/forum/4-titans-and-nfl-talk/ and claim that the surrounding cast on offense in 2021 and 2022 is/was about as great as it could/can be. LOL. I guarantee you that you will get roasted. Please start that thread and link it here.

 

I'm pretty sure brad said the same thing a few days ago, and I think most of us now agree with that point. Tannehill has certainly had elite games and elite stretches in his career- both in Miami and in Tennessee.

The only thing I'd add is not to put too much weight in "elite" or any term. When we present something as an absolute, we're usually lying to ourselves and letting our opinions jump to a conclusion. For instance, I feel like I'm pretty elite as a grill master, a writer/marketer, and several other things. But then I see trained chef make a steak on television or YouTube and I'm like, "LOL, I suck at grilling." It's all perspective and mostly based on personal opinion, even when we have the stats or tape to back it up.

I certainly have my opinions on RT and Tua. Tannehill is elite with the right protection; he's average to below average without it. Tua is elite in the RPO offense with quick timing passes, but he doesn't have anywhere the negatives of RT when forced to improvise. Both are good quarterbacks though.

Again, that's based on the "facts" I choose to look at and my own personal opinion...but my gut instinct has always carried more weight than the numbers.

 

I loved the thread "Henry is done", LOL. Thirty-one other teams would sign him in a heartbeat.

 

For me elite means top few in the world at something at any given time. I can distinguish between elite total results and an estimation of the ability of the QB. Consider the professional grill master starting with a crappy steak, crappy seasonings, and a crappy grill that doesn't reach the right temp. I'd guess that you wouldn't put yourself in the top grill master category when compared to professionals, but if you have the best cut of meat, the best seasonings, and a perfect cooking surface, you'd likely out do the pro who has crap. So, maybe elite backyard grill master?

You and I agree, Tannehill's weakness is the off schedule stuff. My big pet peeve is when people translate that to mean, he needs perfect surroundings. That is BS because we have literally seen him thrive in less than perfect surroundings. In order to make their case, they need to both exaggerate the good surroundings and down play the bad surroundings..... That has happened again in this thread.

Tua is showing to be elite in much more than the RPO stuff. IMO, his best traits are his quick feet, quick release, anticipation, ability to look off defenders, and accuracy. People want to talk endlessly about arm strength. The fact is that if you can read defenses, throw with anticipation and hit your spots, you can minimize the lack of elite arm strength. If you need to see the guy open before you are willing to throw it and need to rely on arm strength to get it in, then lack of arm strength will kill you.

Tua's average arm strength shows up when he is late on throws or cannot step into the throw at all. He cannot rely on throws where he has no base and he cannot just throw it harder if he is late. But, those should be the exceptions. His ability to throw with anticipation means he can release it sooner and that helps his OL and doesn't allow defenders to jump the route. I have seen tons of throws where it looks like the ball is taking too long to get there but when it arrives, there is no defender close.....

 

To me, arm strength is great and I'd love to have a quarterback with an arm like Drew Bledsoe that could toss the ball 85 yards. But how often does a QB get the time and the timing to throw an 85 yard TD strike through the air? Or even a 60 yard pass, for that matter? It's certainly one of those "great to have" traits, but it is certainly not a deal breaker.

Specifically for Tyreek and Waddle, Tua will learn to release the ball earlier...likely before either receiver passes the corner or a safety in their routes. He's learned from year one to year two what "NFL open" means, and eventually he'll figure out how to time our blistering speed seconds before our boys are open. It has to be pretty darn hard since you have to check the safety, the receiver, and the corner before releasing, but I do believe that we'll see it more and more in the years to come.

The good news is that Tua's already connecting with Waddle on those 15-20 yard strikes in stride for the easy TD. Tyreek is a different beast though, his explosiveness is off the charts and it feels like he can cover 10 yards in a half a second. Corners intentionally play him way off the ball with safety help for that reason, so it's going to be rare to see him streaking down the sidelines without additional coverage over the top. The kid is a nightmare, LOL, and it's a very good problem for us to have that we can't hit him in stride very often 40-50 yards away. He's still coming down with the ball regardless of coverage.

 

I'm. Not. Arguing. Against. Tua.
My only concern right now about Tua is whether or not his head will hold up long term.

You guys have assumed things about my stance that aren't true, and you now interpret my posts through this lense that I'm anti-Tua.

I said FROM THE BEGINNING, that I believed Tua would have a relatively long career, barring significant injury. I just didn't think he would play at elite levels. He's showed me that he can play at an elite level.

 

His efficiency is still a result.

Those odds aren't predicting the future. They're just giving you the probability that you will be successful. You can still be unsuccessful for any number of reasons.

Again, I really think you guys put way too much focus into stats.

 

Yes. This is the kind of real discussion this board was started for.

 

lol.. that's still in your head?? Can't get it out eh? Anyway, what I said is fact: Titans fans were near unanimous in agreeing that the only reason they couldn't progress further in the playoffs is because of Tannehill. I went to those boards and read through threads for a few days after that debacle to see the reaction because to everyone else (except you apparently) it was patently clear Tannehill LOST the game for them. And that's what they said too, that this year (last year) there were no excuses (i.e., no excuses from the surrounding cast). Tannehill simply couldn't carry a team and overcome adversity when it counts.

btw.. Vrabel understands that too. That's why there was such QB controversy after he drafted Willis. Anyway, as I said in that long thread about Tannehill, it's true that what Tannehill did in 2019 and 2020 is statistically significant compared to what he did in Miami. It's also true that he's back to his performance level in Miami since then: statistically average. That is not how a "very good" QB performs over time. 2019 and 2020 are going down as flukes in Tannehill's overall career if his play since then is any indication. Guy is an average to slightly above average QB who had 1-2 years of great play.

 

let me try and explain it a different way.

Let’s say we have an alligator infested river. Over the past 10 years we recorded the number of attacks on people. 425 people have been attacked over 10 years. We start to divide up the people by things like color of clothing, swimming vs not swimming, day vs night visits to the river, etc. We find, through our analytics, that 90% of those attacked were swimming at night. 5% were swimming during the day, and the remaining 5% were just laying near the shore when attacked. Sounds like we shouldn’t swim at night, right? That’s a prediction. Why? Well, through our data collection we found that 10,000 people had been in the water at night over the last 10 years and yet only around 382 had been attacked. there’s a good chance you won’t be attacked even if swimming at night, but the GREATEST chance is if you do.

Same with going for it on 4th and 1 with 3:30 minutes to go in the 4th quarter down by 2 points. We can predict the greatest chance of winning or not winning based upon past results.

 

"Efficiency" is a trait of the QB. The stat is only a measure of that trait. Doesn't mean the measure is perfect, but what is being measured is a cause, not an effect. So Y/A, passer rating, etc. are all measuring causal factors. The stat is updated after each play (a causal factor).

The probability you'll be successful IS the prediction. Most scientific modeling predicts probabilities of events, not "precisely X will occur with no uncertainty". It's actually interesting historically because before the 19th century you'd have been more correct than incorrect. Newton's theory of motion for example is purely deterministic, no room for uncertainty. But as statistics rose to prominence in the 19th century models started to incorporate uncertainty so you see prediction of "expected value" accompanied with confidence intervals, etc.

 

This is awesome!

 

I'd love to hear from some of them now. They didn't have Chris Canty on there, but he was trash, too. Of course, after the Baltimore game, he switched from "Tua sucks because he's not a good QB" to "Tua still sucks because I question his durability". Like a lot of *ahem* Dolfans, he has switched from basing his smears on proven performance to predicting eventual failure.

 

Again, I don't necessarily agree that "efficiency" is a trait. "Efficiency" is a quantified representation of perhaps unquantifiable innate traits. Traits like perception, brain processing, etc etc. The better your innate traits, the more "efficient" you will be.

 

The definition of a trait is any distinguishable (i.e., through measurement) quality or characteristic, usually of a person. By definition efficiency is a trait. So if one person solves a particular type of math problem faster than another person does, that's a trait of that person by definition. Or if someone processes football relevant information faster than another person, that's a trait. Y/A and passer rating are measuring some amalgamation of such individual traits, and by definition they themselves are measures of traits.

 

So then take my challenge.....

Wrong. Still..... I would agree with you IF the Titans roster in 2021 and 2022 had not fallen off so dramatically. That is why you needed to make the BS claim that I am challenging you to test out. You won't because you know you are wrong.

 

What fans think now isn't the point. What they thought in the few days after that debacle when memory of the game is fresh is precisely what I said. Either way, over long periods of time a "very good" QB will demonstrate well above average career rating after adjusting for era. 2019 and 2020 certainly helps Tannehill, but if he keeps up what he's doing now there's no way around it: he'll only be statistically slightly above average over a long career, and that means he's not a "very good" QB. We'll see of course. His career isn't over.

 

LOL. What disappointed fans think in the days after a terrible point is how the supporting cast and Tannehill should be judged? LOL. Man, you're better than that.

Reach back out when Tannehill under performs with an average or better pass blocking OL. It will be the first time in his career.

You just cannot defend this statement: "Tannehill had possibly the best surrounding cast you could ask for in the playoffs last year and completely bombed. How can you argue it's surrounding cast when you see that?"

How can the 31st rated pass blocking OL, no decent TEs, and a depleted WR corps be considered the best supporting cast you could ask for? It can't and you know it.

Now, I will admit that Tannehill bombed. That is a fact.

 

How did we get back to Tannehill?

 

At least they remember what happened in the game. FAR better than asking them a year later like you're suggesting we do.

danmarino was alluding to this earlier, but the QB has a massive influence on how good pass blocking is. Statistically, the two most stable stats (among commonly used ones) when QBs change teams are sack% and completion%. In other words, pass blocking is in many ways more indicative of QB ability than OL per se.

I said that just a few weeks ago in a thread where people were saying we have one of the worst OL in the league, and that Tua, once he comes back healthy, would get killed behind it and likely not last long. Well, I predicted that this "worst OL" we saw with our backup QBs would suddenly start looking like a decent OL pass blocking-wise precisely because pass blocking is in many ways a QB stat.

That's exactly what's happened. Tua with his quick release covers up for tons of OL issues and the OL doesn't look so bad anymore. So Tannehill often playing with supposedly a "bad OL" is in many ways indicative of him not being good at evading pressure. He's had that problem since playing here in Miami.

In other words, OL stats like pass block win rate really aren't good measures of OL ability. Not all stats are equal, and while I give credit to ESPN to putting together PBWR — it's the best attempt I've seen so far at an OL stat — in absolute terms it's not a good "OL" stat because it's so confounded by QB ability.

btw.. Titans were 24th in PBWR in 2021 not 31st:
https://www.espn.com/nfl/story/_/id...opping-blocking-leaderboard-win-rate-rankings

 

FinFaninBuffalo's posts are almost exclusively about defending Tannehill. Don't worry, after tomorrow's game this will be a Tua thread again.

 

Right. But "efficiency" is a stat that incorporates all traits. When someone is "efficient" that is a result of many individual traits working together. Again, efficiency is a result, looking at many individual things. Interesting that passer rating is also a result, that many treat as a causal action.