Meep Keeping him isnt a debate at this point, you keep him if your Jon Robinson. The issue is do you franchise tag him or do you go ahead and unload the brinks truck?
I can respect an opinion but I broke down every snap that guy ever took in Miami and graded them even. Others know this. From pre to post snap to execution you name it. And he was ballin after week 5 of 2016 in terms of qb ask and execution. same thing I’ve been doing in Tennessee. Just not posting play by play break downs of it lol.
You pay him market rate. Only question is how many years do you do it for and that in turn dictates guaranteed money. but I still say even on a 3 year pact he’s gonna be giving you a bargain if it’s less than $90 mil guaranteed. The shorter the pact the more likely his agent demands fully guaranteed
"this kind of play"? Like 118 passer rating and 9.8 YPA play? Nobody has EVER played that way consistently. If we are talking about slightly above 100 passer rating with a YPA in the 8 range, Tanehill did that consistently when his OL was intact from 2014 to 2016.
Yeah this is when you absolutely have to use z-scores, and many other QB's have done stuff over 8-game ranges more impressive. I mean.. Matt Ryan played at this level (2.47 z-score comparable to Tannehill 2.48 currently) over an entire 16 game period. And Steve Young in 1994 was more than a full standard deviation higher at 3.57. If you ignore z-scores you're just capturing passer rating inflation (or Y/A inflation), which isn't something specific to that QB but applies to all QB's in a given era.
I still hold that passer rating would be a far better stat, and far more useful in determining the QB ability/contribution, if it only contained yards and TDs that occur directly as a result of the QB. In other words, QB stats stop as soon as the receiver catches the ball. Everything after that is only on the receiver. And yes, I understand that QBs can help generate YAC with proper ball placement, but that would seem to be frought with the same issues that we complain about PFF for, with that determination being interpretation by the person watching the film. It would be much easier to just end the tracking for the QB at the catch. You might be kind of penalizing QBs who are accurate, but that seems better than artificially inflating all QBs stats, and it doesn't penalize QBs who don't have good YAC receivers while unfairly benefiting those that do.
Even Matt Ryan in 2016 only did it in 4 game stretches. He had one at the start of the season and one at the end. Steve Young was unbelievable in 1994. Tannehill is in very rare company and he will come back to earth a bit but I still think he will continue to be very good.
Yeah I'd agree with that. It's a little hard to actually produce this type of passer rating right now (otherwise I'd do it) because even in play-by-play databases (which I have for an 8-year period) they don't tell you air yards on that specific play. In other words, air yards are computed in bulk and not reported in "play-by-play" data.
If you’ve read the book, and I have, what cbrad is doing is the opposite. Most of the ways to lie with stats involve truncating the data, hiding the uncertainty, splicing apples data onto oranges data. Cbrad has been very clear about transparency and making sure the data is not cut or sliced into portions that fit a particular argument.
Yes... there's the total contract amount, then there's the guarantee, which you get as long you meet requirements. Kirk Cousins' contract was unique in that it was a fully guaranteed three-year contract worth $84 million total. This is the first fully guaranteed NFL contract and the highest paying contract in NFL history. Andy Dalton is currently nearing the end of a six-year deal worth $96 million which included $17 million in guarantees. Tannehill is slated to make $2M next year currently. The Franchise Tag would be ~$26M + for the 2020 season. Signing him to a $100M contract for around $20M a season could be a smart move if he continues to play well.
No way he takes $20M per IMO. He won't demand top QB contract money, but $20M is only $3M over his last one. $23-25M would be fair for all sides I'd think.
I was being as conservative as possible...I 'believe' they'll come in around $27-28M per...but could see higher.
The talk about Tannehill‘s future contract brings up another possible contributor to his current performance: he’s in a contract year.
Aah, so you think he was intentionally holding back the past 7 seasons while waiting for this moment? That's pretty sound logic I guess.
This is completetly off topic, but since your a stat nerd, would you consider LeBron better than Jordan? Or are they not compareable beacause of different era's?
Titans WRs were scrubs under Marcus and he couldnt hit them in strides. They look completetly different with Tanny.
I'm not speaking for cbrad, but I suspect that what he'll say is that adjusting for era is important to the degree that the sport changes fundamentally across eras. In football, and in this discussion in particular, this pertains to the rule changes in the league that have made the passing game be featured far more than it was in other eras. As they say, "it's a passing league." But that wasn't always true. It wasn't always a passing league. Consequently the players passing the ball function differently today than they did in 1970 for example.
I know that Tannehills performance with the Titans has been the highlight of this season for me, followed by the Dolphins showing some spunk and winning a few games.
I haven't analyzed any basketball stats so I can't answer the first question. But in general you can compare across eras as long as you equate distributions across eras. So if the mean and standard deviation of the distribution of some stat (whatever you're interested in) changed from year X to year Y, then you adjust the stats in one of those years to match the other so that the distributions match (probabilistically match). The next question is which stat to use, and ideally you have something like wins above replacement (WAR) in baseball. People have tried to create a WAR in basketball but it's nowhere near as good as in baseball because basketball boxscores are pretty stats-sparse (e.g., you have little information on defense) and because basketball involves more interaction among players for any play. But in general yes you can compare once you adjust stats to a common scale.
Anyway.. I was curious about something: how different are the yards to go on 2nd and 3rd down that Tannehill has faced in Miami and Tennessee? If they were very different (e.g., if he's consistently facing short yardage situations in Tennessee on 2nd and/or 3rd downs) that could help explain the huge jump in ratings. But look at these graphs (I'm only comparing 2019 vs. 2016+2018 in Miami): There's not much difference. More 2nd and short in Miami, but more 2nd and 5-ish in Tennessee, while 3rd down isn't that different (except 3rd and 9 which apparently hasn't happened yet in Tennessee). Data are from pro-football-reference play finder: https://www.pro-football-reference.com/players/T/TannRy00/passing-plays/2019/ So whatever is going on with Tannehill in Tennessee it has relatively little to do with yards to go on 2nd and 3rd down. Oh, and both histograms are normalized so that the height is probability. EDIT: forgot to say these are passing attempts only.
Common denominator in Tannehill balling in 2016 and 2019?? The running game being the defenses focal point. If he has it all opened up for him? He’s a STUD
Im not sure how bad the Dolphins roster was with Tanny. But if my memory serves me right, didnt you guys have Landry, Parker thats actually ballin".. Stills wasnt bad either. Ajayi had couple years of being a monster.
Interested in #cbrad take on this. In 2014, in the eight regular season games 4 through 11, Tannehill posted the following passer ratings: 109.3 83.3 123.6 73.3 125.6 81.8 114.8 104.9 In 2016, in the eight regular season games 6 through 13, he posted the following passer ratings: 97.4 99.4 86.8 130.6 89.3 130.6 63.1 124 In his eight games this season he's posted the following passer ratings: 78.1 120.1 109.8 82.3 133.9 155.8 131.2 140.4 Unless I'm mistaken, neither of the top two samples of passer ratings is significantly different from the bottom one.
And? Has nothing to do with Tannehill. I am just interested in seeing what it would like. Who would be the top passers. I think it's more accurate in some ways.
That's correct they're not significantly different, but it doesn't really mean anything because that's solely due to small sample size. You have 8 games vs. 8 games. That's rarely going to be significantly different for almost any situation. Simple example: choose games 1-8 instead of games 4-11 from 2014 and you'll also get no significant difference (and it's even less so once you properly adjust ratings). That's why when you apply the t-test you have to keep in mind that at least one of the sets (and ideally both) needs to be large enough so that the result of the test isn't just due to small sample size. So when I test for significance between Tannehill's games in Miami vs. Tannehill's games in Tennessee I'm including all 88 games in Miami vs. 8 in Tennessee. That "8 in Tennessee" might still raise some eyebrows in terms of small sample size, but we're talking passer ratings here that I know tend to be normally distributed as opposed to something like the distribution of serum samples from people infected with malaria (not even close to normally distributed), and since the t-test assumes normality in the differences between samples I'm confident enough in applying the test to 88 vs. 8. That's the math answer. The science answer is that you can't just select any subset of the data without there being some independently identifiable difference in that subset, and I don't see how you can show that in this case. Can you really show which games were the games where "the running game was the defense's focal point"? If you can't show that then it's just cherry picking data.
No, they're not as good as this stretch. He's gone of 130 4 times this year and just 2 times in those 2 stretches overall.
Appreciate your response. So help me understand if you would why, if the question for someone is, "is this stretch of eight games significantly different from any other Tannehill has played," any eight game stretch he's played historically wouldn't be legitimately used in that comparison?
Yeah.. the answer to this one relies on the very precise usage of "any" and "all" in mathematics. You said "any" but I think you meant "all". You asked whether the current stretch of 8 games is significantly different from ANY other stretch of 8 games Tannehill has played. That technically means you only need to find ONE example to say "yes". But I'm pretty sure you didn't mean that. If however you really mean "all", then you have to compare to all 8 game stretches simultaneously. So what you'd want to do is to look at the (sampling) distribution of the means of all 8 game stretches and see what the z-score of the current 8 game stretch mean is. Naturally you will have to adjust ratings first or some of the differences will be due to passer rating inflation.
Landry is not an offense driving weapon. Parker missed many games, was injured all the time...this is the first season he's really shown that he's not a wasted pick. Ajayi had those 4 monster games, but was most definitely not a monster for several years.
I agree. Though, I wasn't talking about CBrad. I appreciate his knowledge in the area of statistics and analytics.
Depends on the terms, guarantees, years, etc... but Tannehill is an ultimate team guy type. I can easily see him taking less money to help stock the shelves and keep the team in playoff contention as opposed to simply wanting to "get paid" or be one of the highest paid QBs in the league.
Roster wise, we had some good pieces, but we never really had them going well at the same time. Landry has been the second most divisive player here behind Tannehill in recent memory. Some people loved his effort and leadership. Other people saw his production as underwhelming. Ajayi was a monster for about half a season. Parker has broken out this season, but before this season he has been perpetually disappointing because of injury. Lamar Miller was good, but underused. Spent far too many games in Philbin’s doghouse for no good reason. Mike Wallace, well he was commented on a few pages back. The only TE worth more than a bucket of warm spit he had here was a half decent year was Charles Clay, and then the FO let him walk to a division rival.