Sounds like an invite to waste far too many neurons. We have Tannehill. What difference does it make? He's our QB and he's improved each year. As opposed to going the Moneybag route, whose been available to us since Tannehill that you would have rather had? So far based upon health, I'd take him over Luck but we weren't in a position to take Luck that year. Which of these names was the baller you wanted instead in the draft? Brandon Weeden? Brock Osweiler? EJ Manuel? Gene Smith? Blake Bortles? Johnny Manziel? Teddy Bridgewater whose leg fell off dropping back? I would have liked Derek Carr but no-one saw him playing this well. Garoppalo there isn't enough body of work to determine yet.

Weren't drafting high enough to get Winston, Mariotta, Goff or Wentz.

So, what was the decision we muffed on that would cause us to discuss Tanehill more? Who besides Carr has been available that would've done better?

 

I think there are too many unknowns to agree on which "moneyball" route is the correct one. I'm sure there are ways of looking at the same data that would lead you to the conclusion it was good to keep Tannehill as well as that it would have been better to pay a decent backup while more actively searching for a better QB.

Seen in (relative) isolation, Tannehill's contract isn't bad though because it gives (and gave) the team multiple outs with relatively little cap hit if he didn't improve to become a franchise QB (defined as good enough so that finding a new starting QB isn't a high priority). If Tannehill doesn't do well in 2017, we lose little cap-wise if we cut him after that.

As far as who we could've taken had we gone on a perennial QB search, other than Derek Carr (who Philbin wanted.. might have been the ONE good judgment that guy made) you're obviously looking at Dak Prescott.

 

A lot of people on this forum were banging the drum for Russell Wilson. Some weee saying Luke Kuechly and Wilson.

That said, it's under the bridge. We got him. He's here. He's not top 10 but his ceiling is there. And those are hard to find. I would keep finding projects later in the draft (Cousins was drafted in the 4th) and keep developing Tanny.

 

For 2017, Tannehill has the 7th highest cap number among QBs in the NFL. However, not long ago he was several spots lower, before others were let go. The #7 spot is just one thing to look at though, as there are eight other QBs between 18m and his 20.3m number. His number is pretty average for an established NFL starting QB.

For 2018, he drops to 13th. If it weren't for all three of the QBs picked in the first round in 2014 not panning out, those three would all be jumping ahead of him in the rankings that year also. As it is, Derek Carr will, and possibly Garapolo too if someone wants to throw money at him to be their starter. Also, Matt Stafford will be a UFA and will get a new deal topping RT, and Sam Bradford is a UFA too. In the end, RT will be somewhere around the middle of the league as far as cap number. I don't know how anyone could argue that that's a bad deal.

Then in 2019, you'll have Matt Ryan, Winston and Marriota all getting new deals, and potentially Alex Smith could end up somewhere as a starter too. Though it doesn't seem likely right now, Blake Bortles could rebound and get a contract from someone too. That could push Tannehill all the way down to around the 20th highest cap number amoung starting QBs in the next to last year of his deal.

I see no logic in the idea that he's a bad value.

 

I guess i dont understand what your position is then. Either way, my feeling is that long term stability at the QB position is vital, and that switching up every couple of years is futile. You need to ride out the waves in order to get the best out of your guy.

 

This is true, but you have to have the right guy, this year should tell us if we do.

 

Hopefully, people won't foolishly use playoff success to determine that.

 

One way to test whether the increasing amounts of money spent on the QB position is justified is to see whether the expected number of extra wins per increase in passer rating point has decreased relative to previous years.

That is.. take all passer ratings of all starting QB's in a given year, calculate the expected number of wins for that QB had he started all 16 games, then find the best-fitting line through that data (so.. x-axis would be passer rating and y-axis is expected wins if all 16 games were started.. then fit a line), and do that for every year.

The slope of the best-fitting line for any year tells you how many extra wins you should expect for an increase in passer rating by one point. If that slope starts to consistently decrease, then you have evidence people are valuing the QB too much because you aren't getting as many extra wins for an above average QB whatever the current average (measured by passer rating) is.

Hope that makes sense.. and here's the data:
https://ibb.co/jsvvha

The overall trend is actually going slightly up, not down, meaning that if you just go by the overall trend the QB may actually still be slightly undervalued if that's even possible lol.

But as you see the last two years the slope has been less than in years before. That is, in 2014 you could expect 0.2 extra wins if you increased passer rating by 1 point, but in 2015 it went down to 0.11 wins and in 2016 it was 0.14 wins. We'll have to see if this is just due to random variation or whether this signifies a new trend, but if that lower slope remains for a couple more years, then I think you may have evidence for overvaluation of the QB position.

 

that 5th best passer rating corresponds with some respected football people that say Ryan tannehill is excellent at the pre and post snap read game.

 

I was determined to stay far away from yet another Tannehill conversation...but I took the bait on your comment anyway.

How many years have we said that exact same thing? We all said it in years 2, 3 and 4. Some were ready to give up and Gase helped Tannehill get back to what he does best....so now we're saying it again in year six. Year six and we're still playing the "wait and see game." Do you even realize how insane that sounds? The average quarterback gets one bad season and he's tossed to the curb.

Personally, I'm all the way sold on Tannehill and his long-term value...even though I don't think he's 100% where he needs to be. He has so many intangibles though that reminds me of the great ones.

 

I won't judge him like that, if he plays well.

 

The great ones however, don't take until year 6 to break out, while RT hasn't had a bad year since his rookie year, he hasn't had a breakout year either.

Last year, the year that many say he's taken big steps, he threw 19 TDs and 12 INTs, the worst TD/INT ratio since his rookie year, those are bad numbers, other numbers don't matter when you have a TD/INT ratio like that.

Richmond Webb on the Perfectville podcast said "this is the year because he has all the things in place to be properly judged"(paraphrasing), and I happen to agree with him.

 

Years 3-5 yes we've said hey wait another year. But this year we finally have Gase. Gase looks like the real deal. People thought Lazor was the real deal, then he went into IFDGAF mode.

So this year, is the first let's wait and see year, with a good coach. Nothing Gase has done has me question him as a coach. Sparano I was never hot on. Philbin I disliked from the beginning. Gase hits the right buttons. I mean, there is no "bad coaching" excuse next year.

 

Minor point but..
In 2012 (rookie year) Tannehill had 12 TDs and 13 INTs with a ratio of 12/13 = 0.923
In 2013 (year 2) Tannehill had 24 TDs and 17 INTs with a ratio of 24/17 = 1.412.
In 2016 Tannehill's ratio is 19/12 = 1.583, so his 2016 ratio was better than both his rookie year and 2013.

In any case, you're right those are bad numbers. The average TD/INT ratio among starting QB's in 2016 was 2.543, but that stat is totally skewed by Brady's unreal numbers (he had 28 TDs and 2 INTs!!! for a ratio of 14) so taking the median is better. Even taking the median you get 1.996.

Either way, Tannehill was 10th worst out of 30 QB's in 2016 that started at least 9 games.

 

there are a fouple of things I want to point out a out pay.
1) Most importantly what someone is paid is based on the expectation of future production. So the market for QB salaries is set based on projections.
2) It is not a complete free market. Salaries are held down most notably by the rookies scale, but also the threat of being tagged. Very few genuine starting quality QBs actually hit the free market.
3) Capflation. Deals that seem overpriced today might become very reasonable in 2 or 3 years down the road.

Basically only the unrestricted FA QBs who signed their deal this off season are being paid genuine market rates, and even the their numbers have been inflated because of artificial restriction of supply (for example Kirk Cousins being tagged) but almost every other QB is being relatively underpaid.

When you look at 'moneyball' approach, you're trying to find assets that are undervalued. This is a little difficult in the NFL because of the greater difficulty in transferring top tier player assets in the NFL vis a vis the NBA or MLB. So the 'moneyball' thing to do is to develop your own QB and sign him to a long term deal while you still have leverage. Ideally if you make it a 3rd or 4th rounder like Wilson or Prescott you will get the maximum value. Be warned if you hesitate you could end up like the Redskins with Cousins. Also the failure rate in the later rounds is much higher too, so I'm pretty sure the Wilson/Prescott approach is like walking into the casino and putting all your money on one number at the roulette wheel.

So from a moneyball perspective the phins have chosen the right path to take with Tannehill. Whether he is worth his contract is a different question, but the approach is good.

 

As for the effect of PR on W-L record I crunched my data a little differently to cbrad.
I looked at seasons 2002 to 2017. I checked the correlation of win% to team PR made and allowed. I did this both season by season and for the 16 year period as a whole. I also adjusted the data to a 2016 base to compareolder years and recent years equally, since average PR is raising by 0.75 points/year in the period I looked at.

Overall the correlation between win% and PR is .67, with the trend lend going from 68 and for every +2.5 in passer rating you gain 1 expected win.
Using round figures a PR of 90 should get you 8-8 and a PR of 100 should get you 12-4.

But if you slice the data season by season then there is much more variance. I don't have the computer I did the number crunching here with me now, but iirc the correlation was as low as .45 in some years and as high as .75 in others.
So the expected effect of a QB's PR isn't as stronly predictive for any one given season.

The correlation between PR difference and win% is 0.80. But the seasonal variation (iirc it went from .75 to .85) was much lower than offensuve PR. So essentially some years the defense is strong, other years the offense is strong, but the difference is the king.

So from a moneyball perspective the difference between Tannehill's pay and a top level QB is roughly $5 million a season. If the top level QB can maintain more than a +5 differential in PR then I really can't see another area of the team you can add $5M in value (eg upgrading a mediocre safety to a consistent probowl safety) that reliably gains you the extra 2 wins a season.

The problem with looking at Matt Moore is his history of concussion. Trent Green and scrambled eggs comes to mind. The reason we have him as the backup is no one wants to rely on him as a starter.

 

based on the lack of talent and coaching and scouting this team had before Gase, I'm not sure if anyone could've succeeded at qb. So financially, Tannehill has been a waste up till last season. Now with Gase, the on-the-job-training is over, the depreciation is dusted off, and in year two, we hope to have the qb in place, without having to get a rookie to come in like a few teams did this spring. I think this team failed in going ahead and getting Tannehill in 2011, and not properly building the team and making huge mistakes surrounding him. Mike Wallace, trading up for Jordan, losing an entire draft class, etc....

There is a the debate on who the other qb should and could have been and when that should have happened. I'm happy what's in-house today. He has chemistry with the WRs and is expected to flourish this season.

 

I always thought we signed RT too early. However, it was a gamble that had to be done, as if he had blown up, re-signing him would have been ridiculous. They preferred to sign him than to give him the fifth year option/ franchise/lose him eventually. I don't know if we are +- financially from that now two seasons later.

Looking at year two, Gase and this team believe in their qb. whether he was drafted too high, or has been inconsistent, Miami is happy on who he is now. Miami has made many more mistakes than the inconsistent play from Tannehill

 

Are you even thinking about life beyond 2017 though? Is Matt Moore the answer in 2019? 2022? You might catch lightening in a bottle for one season with a guy like that, but he's not going to be your answer at the position.

Ryan Tannehill is good enough to play well and be the quarterback of a team that makes the playoffs. Once in the playoffs, anything can happen. If we weren't in the same division as New England, I'd say that we have as good a shot at making the conference championship game as most other teams.

 

good points.
I will say this, no matter what, Miami has to address the qb position, even as a backup, after this season. Miami is building a good team, and would hate to have to depend on Doughty or Fales or something like that

 

Here's a graph of correlations between PR and expected win% for all starting QB's through NFL history:
https://ibb.co/dtBz8F

Correlation stays the same whether you adjust PR or not to any year so that won't change anything. Despite recent correlations being high, the average correlation from 2002-2016 is 0.599 and it's 0.592 if you combine all the data so we have some methodological disagreement here. 0.67 is really high in comparison so we're doing something different. Can you be specific about how you calculated that?

Just so it's clear, I'm taking PFR's stats for each year (e.g. for 2016: http://www.pro-football-reference.com/years/2016/passing.htm), allowing PFR to filter out all stats for "non-qualifying" QB's, then calculating correlation between PR and win percentage where win percentage is over games started.

As far as the best-fitting line to PR in any given year, there I think there's some serious disagreement. As pointed out before I'm getting this:
https://ibb.co/jsvvha

You're saying that for every 2.5 increase in passer rating points you get one expected win? My graph shows it should be more like double that (as in more like 5 PR points for one extra win). Let's just look at this for a single year so we can compare. This is the graph of expected wins vs. PR for all starting QB's (according to PFR) in 2016:
https://ibb.co/jNwfha

You're saying that line should go from 68 (I'm guessing it's at zero at 68?) and see an increase of 1 win per 2.5 per passer rating points?? Such a line doesn't fit the data. Besides, that would mean you get to 16 wins if you increase passer rating by 2.5*16 = 40 points, so at 108??

I think you're doing something wrong here. In any case, since these stats come up a lot, we need to make sure the differences are purely based on which passer ratings are being considered so can you show explicitly how you're calculating this stuff?

 

It's a simple fact, that after the ENTIRE offense and offensive staff got over the learning curve, Thill was a Top 10 QB in the league. That's inarguable. It actually happened. The only thing that stopped him was an uncharacteristic injury.

What we pay versus how he plays is a steal for us. Steal.

 

If Moore is just as good as RT then Moore could command the same salary and/or draft picks.

 

The greats didn't have new coaches and new offenses each of their 1st 6 seasons either.

 

wow, shocking statement. I think you are equating the team being mediocre and taking that out on tannehill. now that this team is building up. I know on a small sample size, Moore did ok. However, he played like crap in the last two games against playoff teams. I can't have that over a long stretch. Tannehill's salary is irrelevant. This team was never cap stricken. They actually could have spent more. If Gase (the ultimate qb whisperer), didn't make a move on a qb, then why the pressing need to get rid of him?

 

look you are beating a dead horse. If RT continues this, he will be out. I got bashed for saying that. THE EXPECTATION is that he will improve significantly. The past is the past