I wanted to go a little higher because I am, well...you know...a ****ing homer, but I went with solid and dependable. Only because I think it will take him some time, once the games start to really count, to knock the rust off. As the season wears on, I think he will get better as he gets used to the play speed and the new guys we have. That will be good, IMO, it's better to build and peak later in the season anyway. :-)

 

I think a better supporting cast at both TE and Oline allows RT to get to the top of the 2nd tier QB's (95 rating)

 

When I’ve looked Tannehil’s production
1) He was severely handicapped by Mike Sherman’s antiquated offense in his first two seasons.
2) There is clear statistical evidence that shows opposing D’s were taking advantage of Bill Lazor’s known playcalling tendencies in Lazor’s second season as OC.

Therefore I feel very comfortable in saying Tannehill’s true floor, assuming a successful recovery, is a low to mid 90s passer rating.

I want to believe that he is a genuine 100+ passer rating QB. However until he shows it over a full season I remain skeptical.

 

I went with solid and dependable, as he's coming back from the injury. I expect solid and dependable, with some higher rating games thrown in.

 

Tannehill has been a very mediocre QB his entire NFL career and that was before he suffered a serious knee injury. Therefore I suspect he will once again be mediocre at best in 2018 and the Dolphins will end the season 6-10 or possibly 7-9.

 

I'm predicting just under 95 rating.

Tannehill was ranked 12th in passer rating in 2016 and also 12th if you use all starting QB's best 8-game stretches in 2016 (proper comparison for that 8-game stretch where Tannehill had a 100.13 rating). And 12th in passer rating is generally around 93.5-94, so I'll go with that.

 

I don't know man, with respect, after watching the Cutler **** show last year, I sure wished Tannehill was in there. I am not going to argue that Tannehill needs a good supporting cast to be effective, but when he has that he is good enough for us to win and much better than mediocre. I also believe his leadership behind the scenes is underrated, so that goes a long way too. Sometimes it's not all about stats.

 

As a long time Dolphin fan, I hope you are right about Tannehill and I end up being completely wrong. I just don’t see Tannehill as the type of QB that can get the Dolphins beyond mediocre and I believe that by the end of the 2018 season, Gase will be looking to upgrade the QB position if he is still the HC in 2019.

 

Excellent point! I am only guessing, but I think Tannehill would be able to say right now that he has a complete understanding of our playbook and with what Gase wants to do with the offense. That should help him play faster and get better as he builds on his experience this year.

 

90 qbr floor 95 qbr expectations 100 plus qbr hopes

Health’s the only way he doesn’t see another contract extension in Miami

 

Dependable starter. Any franchise qb needs to be able to periodically put the game on his shoulder and go win it. Thus far, it has eluded him. He's got the skills, but has to up his game. Ajayi changed everything his last time out. We'll see.

 

Agreed. :knucks:

 

I really think you haxe to separate bad coaching and QBing. As we’ve seen with Jeff Fisher and QBs like Goff and Foles bad coaching can make QBs look much worse than they really are.

I defer to other experts who showed the play design Mike Sherman used was well and truly past its use by date. I started some threads 2 years ago looking at Lazor’s playcalling, and there is clear statistical evidence showing that Lazor’s playcalling when behind on the scoreboard was so predictable that opposing Ds were able to shut us down, also in his second year opposing Ds were able to get excellent results from the blitz. Because these 2 areas are things Tannehill had no problem with when Sherman or Gase were his OC it is clear the problem was Lazor.

In situations where RT17 has had decent coaching (Lazor when tied or with a lead/Gase) he has consistently shown he is a 95+ passer rating QB.

In my book that makes him better than mediocre. However we haven’t been able to get a good read on if he can be a consistent good/great QB.

 

His last three years playing, he had a 92.8, 88.7 and 93.5 rating. I'm really not sure how anyone would label that as mediocre. That's under three different head coaches.

From 2014-2016, here are his totals for per game ratings:

100+: 15
95-99: 4
90-94: 5
85-89: 4
80-84: 5
70-79: 9
under 70: 5

 

Small correction: 13 100+ ratings from 2014-2016, not 15:
https://www.pro-football-reference.com/players/T/TannRy00/gamelog/

I'll tell you what's a much bigger correction, but not in your post.

There's one Tannehill stat that just makes NO sense and that's his 124.0 rating in his last game against Arizona in 2016. pro-football-reference.com, espn.com and nfl.com all agree that he had 15 completions in 20 attempts for 195 yards, 3 TD's and 1 INT in that game.

OK.. well those individual stats correspond to a 134.4 rating, NOT a 124.0 rating. I have no idea what's going on there, but someone messed up there big time.

 

It's the only discrepancy I've found in career gamelogs for like 40+ starting QB's. That's thousand of games where my program predicts the same passer rating as found in their passer rating column.. except this one.

Maybe there's another one somewhere else, but I just find that pretty interesting lol. Anyway.. I've found other errors before so it's worth keeping in mind that there are errors on rare occasions even at the official nfl.com site (we know it has to be an error due to the inconsistency..).

 

I ran the numbers out of curiosity myself here: http://www.primecomputing.com/

According to them, it amounts to a rating of 123.958

Either way, the main thing is that RT17 is very capable of putting up ratings of 95+ when things don't fall to crap around him. And since he's been over a 90 rating in two of the past three seasons, and close in the other, I feel really, really confidant in saying that if he's healthy enough to play, that's his absolute floor.

And if you take away his game vs Baltimore in 2016 (where all Dolphins crap the bed), his season rating jumps to 96.28. So again, as long as he's healthy, saying that he's in tier #2 is only saying that he'll continue to play about as well as he was before. That hardly seems like a leap!

 

OK this is seriously weird.

Let's just do this manually using the formula:
https://en.wikipedia.org/wiki/Passer_rating

a = (COMP/ATT - 0.3) x 5 = (15/20 - 0.3) x 5 = 2.25
b = (YDS/ATT - 3) x 0.25 = (195/20 - 3) x 0.25 = 1.6875
c = (TD/ATT) x 20 = 3/20 x 20 = 3
d = 2.375 - (INT/ATT x 25) = 2.375 - (1/20 x 25) = 1.1250

and..

Passer rating = (a+b+c+d)/6 * 100 = 134.375, just as I computed.

So something is seriously wrong here.


EDIT!! I figured it out lol.

I'm wrong!!! Why? Because they say if any component is greater than 2.375 you have to set it to 2.375. That occurs for c. So if you set c = 2.375 then it is 124 ha!

OK.. this was a good exercise. It is very interesting that this discrepancy hasn't shown up before in previous calculations (or the effect was too small to notice). OK I'm editing my program to include this stupid rule lol.

 

That's actually exactly why it was so hard to detect before. When you have multiple TD's with very few attempts, you push passer rating closer to that artificial ceiling = "perfect" game, and any discrepancy with the true formula won't show up!

 

There was no qb freedom in year 2 with lazor. We ran a basically chip Kelly like qb limiting offense and we did it with no tempo even.

We booted by design into wide 9 pre snap edges with clean releases hell I think that tape vs Miami alone with lazor is what got Mario Williams his swan song money in miami. Miami booted into him by design in buffalo. It was a theatre of the absurd. If they tried to account for that wide 9 edge it was bringing the tight end post snap across the formation. With a wide 9 de look pre snap no less and a clean release off the los. Clown shoes.

The only thing anyone should do with that lazor tape and the 2017 qb tape with jay cutler is burn it.

I don’t want to be a Debbie downer but miamis top candidate in 2018 for a Mario Williams like return is Robert Quinn. Let’s hope he’s playing for more than a pay check.

 

I've never understood why there is the ceiling in the first place, or why an imperfect (yet likely awesome) game can be considered perfect. I guess its to keep the numbers all within a reasonable number, and easy to compare without some crazy outliers.

 

I had missed that too.

It probably makes zero difference when calculating season ratings, but it may create holes in game to game stats, or other low volume stats.

 

Now that he has 3 receivers who can take it to the house, a big tight end who can make jump balls in the end zone, 2 running backs who can take a simple screen pass to the house or a wheel route to the house, he will have over 100+ ratings! For the first time, Tannehill has an offense full of playmakers. Oh yea, and Devante Parker! Oh yea, and an interior line that that is better than blocking then a macaroni strainer is at holding water.

 

Parker can take it to the house too :up:

I bet that with all of the other talent around them, RT will hit Parker a few times this season where he has a lot of room to run and he'll leave the defense behind. We've seen Landry and Rishard Matthews do it, so there's no reason Parker can't.

 

At the end of the year people will talk up qbr as gospel but what they won’t do is actually try to or account for even what is driving it. Ie is it a actual qb problem or other things around the qb.

Since Ryan Tannehill has been there it has never been a qb problem. That much I can say with absolute conviction. He gets blamed for a lot cause he’s the qb and well maybe it comes with the territory but why tape evaluators and Adam Gase even never blame him is cause the tape never says they should.

I don’t care about qbr although I do agree with using it as a projection for qb play. What I do and have done for years is grade out the qb overall relative to the rest of the team and if the qb isn’t grading out top 5 on a consistent clip regardless of who else shows up that week the only way I’m gonna be competing for anything is if my d and or run game are all time like worthy. And in the case of the run game that hides qb deficiencies at an alarming clip cause the luxury of playing the position mostly off pa or in plan a even on offense cause you are ahead of the sticks or never chasing the scoreboard can not be accounted for. However what it can and does do is inflate overall qb play into thinking or the perception that I have an “elite qb” when really all I have is a place setter or one carried by things around him. Dak Prescott in 2017 a perfect example.

That dudes been carried by the Dallas o line since he showed up. His warts show up in a big way when he has to either chase the scoreboard or play behind the sticks. Playing qb in 2nd and 6 or 3rd and 4 is a luxury that cannot be understated with qb qbr and overall grade. Play in 3rd and 8 plus like tanny did an entire year and that’s separate from a top 5 oline overall and then show me what your qbr looks like.

That kid saw more rush 3 drop 8 or rush 4 drop 7 and contact on the qb before his back foot could hit than you could ever count.

 

Quinn is absolutely playing for an extension. If he sucks this season, Miami can and will release him with zero dead money. He has every reason to try and get one more big contract.

 

I forgot about Parker when I first started writing. So that is 4 potentially big-time receivers.

 

My personal opinion with Quinn is he’s gonna take him 10 mil drop 6 sacks in pin ear opportunities and be cut next offseason. It’s a risk I’d take no doubt but barring us getting teams behind the scoreboard I don’t see him making the impact.

I’d bet on Charles Harris in year 2 with more awareness and a rediculous work ethic over him. But I would like to see Harris come in 260 to camp with that same 10 yard box pass rusher game

 

The problem is people rarely when looking back or judging a qbs play account for the outliers. As in what is he dealing with what are the weaknesses what are the strengths what is driving it. Some love to throw out robber coverage as tannys kryptonite well those are distance to sticks driven more than anything else. What that means is in the progression the opposition is giving you pre snap the sticks there they take it away post snap with a bait look pre snap into a post snap run under into the hook zone or pre snap sticks concept.

That wouldn’t change regardless of qb as long as you allow them to get you behind the sticks. So is that really a qb problem or something qbr or even pff don’t account for situational play? I say it’s the latter. Provided the qb doesn’t lack head and eye discipline in their drops

And I have asked that question more times than I can count from fellow tape evaluators that have carried themselves as authorities on the subject and never once gotten an answer. Just a token “he needs to be better vs robber coverages” never accounting or even acknowledging situational play or explaining what different the qb in a pre snap progression read based offense should do.

 

Did the input some of Moore's stats that game in error?

Never mind, you guys figured it out before I read more...

 

It's because passer rating is what mathematicians call a "linear function" and is linearly related to win% only over a certain range. For example, from 2002-2017, the best-fitting line between 2017-adjusted passer ratings and wins in a 16 game season is: W = 0.1785*PR - 7.51. Solve that for 16 wins (win% = 100%) and you get 131.71. So really the ceiling should be lower than 158.3 from that point of view because otherwise you'd predict a win% way beyond 100%.

Mathematically what that really says is that passer rating shouldn't be a "linear" function but rather sigmoidal in nature.. something like in this link (one could fit what's called a cumulative normal distribution to get that):
http://www.fairlynerdy.com/wp-content/uploads/2017/09/2_Normal_Curve_Cumulative_Distribution-1.png


Here's some interesting trivia though. If we don't use the "correct" formula and go back to the one I was using (which doesn't naturally give you that perfect passer rating ceiling), then the highest unadjusted rating for a single game in NFL history (SB era from 1967) is from Fran Tarkenton in 1970 against St. Louis. He had 15 completions in 18 attempts for 280 yards, 5 TD's and no INT's giving him an unadjusted rating of 228.935 LOL.

2nd on the list is Ben Roethlisberger in 2007 against Baltimore. 13 completions in 16 attempts for 209 yards, 5 TD's and no INT's for an unadjusted rating of 228.385. When you see numbers like this you start to realize why they put in an artificial ceiling.. really ad hoc though. As I said they should be fitting cumulative normals.


btw.. using that equation: W = 0.1785*PR - 7.51, you see how unlikely it would be for Tannehill to put up a 95 rating and Vegas be correct with us winning only 6 games. 95 rating in that formula gives you an expected 9.44 wins!

 

I’m not sure why the multiplier is what it is. However, my best guess is that using 1960’s/early 1970s as your baseline (i.e. when the pr formula was created) 100 pr would equal 100% expected win ratio.

 

I think they realized they needed a ceiling for the rating to make sense, but they didn't really know where to put that ceiling. Got to remember that in 1973 when the formula was adopted they didn't have personal computers like you do today. Serious data analysis might have been done on something like an IBM RAMAC that literally weighed a ton and had a whopping 5MB of storage lol. So I wouldn't be surprised if they couldn't calculate expected win% using every game played (or at least they didn't want to go to the enormous trouble to do so).

What I do know is that they wanted "average" to be at 66.7 rating. For each of those 4 components the average was defined to be 1 (that part is easy to calculate so I know they could scale things there). Then as you see they add 4 of these "1's", divide by 6 and multiply by 100, giving you 66.7 as the target average. If that's the targeted average, I'm guessing they probably thought any ceiling should be at least twice that, but why it's 2.375 times that and not 2 or so I don't know.

 

I think RT will be a 95+ passer this next season. As Unlucky pointed out, and even though he's coming off of ACL surgery, he was pushing that 95 mark his last 3 full seasons. And that was with some major crap on the o-line, no real running game in 2 of those seasons, and a defense that couldn't stop an 80 yr old woman from crossing the street.

If RT finishes this season under 95 I'll be surprised.

 

It can be hard to stop an 80 year old woman from crossing the street. Back when I was in Texas, my great aunt and family went out to Corpus Christi for Easter dinner. She was 99 at the time (she would die shortly after making it to 100), and we had reservations. Well.. we went there and they told us we didn't have reservations lol.

So.. first my cousin (who was an army Captain back then.. he died later in a car accident near Houston) gets into an argument.. no avail. Then my uncle, then my dad. Nothing. It was left up to me to come up with this obviously brilliant idea: "let Grace (my great aunt) do the talking.. see if they have the guts to tell a 99 year old woman she's wrong!"

We got a table immediately lol.

 

I think you should read *The 100 year old man who climbed out of a window and disappeared*. If you don’t mind subtitles there’s a pretty good movie version of it too. I think your great aunt would have approved.