Great point. So many more variables when looking at Tannehill before he started in Tennessee... Everything was different. When Henry got Tannehill as his starting Qb only one variable changed.
Well and I think the glaring contextual clue revolves around the usage of both players and how Tannehill performed when the usage centered around him (i.e., high-volume passing games). Tannehill had the second-fewest pass attempts per game in the league, behind only Lamar Jackson, and when was involved in high-volume passing games, his performance plummeted. So the evidence indicates that when the offense revolved around the use of Henry, Tannehill's performance was stellar, whereas when it revolved around Tannehill, it was far worse. Now, one obvious explanation for that could be that Tannehill's passing got the team out to signficant leads during games, which then allowed the team to ride Henry the rest of the game and run out the clock more quickly, thus lowering the passing volume for Tannehill after he'd already put together a good performance earlier in the game. However, that explanation isn't supported by the data: the Titans were 0.14, 0.44, and 0.05 standard deviations below the league norm in point differential after the first, second, and third quarters of games in 2019, respectively. That suggests the Titans were playing neck-and-neck with teams on average in 2019, yet they were choosing to feature Henry as the centerpiece of their offense regardless, again to the tune of Tannehill's having the second fewest pass attempts per game in the league. So you have this extremely strong correlation between these two players' performances, far stronger than that for any of their counterparts in the league, and one of the players was the centerpiece of the offense while the other was "along for the ride" so to speak. That makes it pretty open-and-shut in my opinion. And then add to that the significant decrement in Tannehill's performance against non-stacked boxes, when opposing defenses were presumably focused on stopping him and not Henry.
Yeah that's a good point. There's no comparable data point like that with Tannehill, and you actually do get statistical significance with Y/A using 2019 numbers alone: games 1-5 (can't include game 6 because Mariota and Tannehill split attempts) vs. the rest. The problem is that you don't see statistical significance once you include 2018 (e.g., with all games where Henry had 10+ rushing attempts) where Vrabel was still coach. And that can't be discounted. Anyway, I'm only defending the position that there's evidence the causal relationship goes both ways. Personally, I do think Tannehill benefitted from Henry more (the centerpiece of that offense) but that's opinion, not something I'm suggesting can be inferred from stats.
Not sure how you would even quantify it, but I think the passing game and running game were co-centerpieces. IMO, they fed off of one another. I also think it is clear that the Titans put a bigger emphasis on signing Tannehill over Henry. Finally, the offensive turnaround was huge, immediate, and undeniable when Tannehill took over. I couple that with individual stats (accuracy, depth of throws, aggressiveness, etc) that are indicative of extremely high performance by the QB.
Every QB benefits from an effective running game... Tannehill is no exception. The fact that Tannehill played at level of precision few other QB's in the NFL are even capable of... Combine that with Henry gaining almost twice as many yards per carry with Tannehill as the starter... and You cannot in any way shape or form realistically diminish Tannehills importance to that offense.
No, Henry's workload changed significantly in terms of volume. He went from averaging 18.8 carries per game to averaging 22.8 carries per game thereafter (including the playoffs). That may not seem like much, but it's actually more than a standard deviation more carries per game, in the context of the 32 NFL running backs with the most carries on the season in 2019. So, it wasn't only Tannehill that changed. There was apparently a decision made by the coaches to feature Henry more, and that again fits with the notion that Tannehill benefited from Henry more than Henry benefited from Tannehill. And as I pointed out before, that appears not to have been a function of getting early leads in games and then riding Henry.
And no, they don't. The 31 correlations between passer rating and yards per rush in post #6348 -- some of them somewhat strongly negative -- are a testament to that. The degree to which Tannehill's performance was associated with Henry's was uniquely strong in the NFL in 2019, far greater than that for any of their counterparts.
In the regular season Henry averaged exactly 19 carries per game in the second half of the season. ( Games 9 through 16 ) He averaged 18.9 carries from games 1 through 8 So no difference in the regular season for workload for Henry between Tannehill and Mariotta But nice try to make it seem like there was by including the playoff games. In the regular season the only difference was in terms of yards per carry... Henry in the first half of the season 3.8 yards per carry Henry in the second half of the season 6.3 yards per carry... I wonder what happened half way through the season.... Maybe a change in the starting QB? Hmmm What we know for sure is that the Titans became the #2 scoring offense in Football after Tannehill took over for Mariotta... Second only to Baltimore... Even better than teams like San Fransisco and the Chiefs. Again points strongly to Tannehill
So Looking at Henry getting the exact same number of carries a game for Mariotta and Tannehill... Here is the difference in play between the 2 QB's Tannehill completes 11% more passes. Tannehill throws his passes more than 2 yards further per attempt Tannehill has a QB rating 25 points higher And remember that Henry got the same number of carries per game under Mariotta as he did Tannehill. Henry increases his yards per carry by more than 50% with Tannehill with the same workload. All pointing to the importance of Tannehill to this offense.
Those figures for the regular season were 18.8 (Mariota) and 21.1 (Tannehill). In the playoffs Henry averaged 27.7 carries per game, while Tannehill averaged 16.7 pass attempts per game. The fact that the team centered the offense around Henry even more when the stakes were highest -- in the playoffs -- speaks volumes. I have no idea why you'd exclude those games as a measure of the volume of the offense attributed to Henry versus that attributed to Tannehill in 2019.
Did you go game by game and attempt in each and go back through the film to see who was the QB... because there were several times Tannehill came in to relieve Mariotta before being named starter week 7. The fact of the matter is that games 1 through 8 he averaged 18.9 carries per game... and in games 9-16 he averaged 19 carries a game If you have access to a site that shows in the games where both QB's played how many carries Henry had... Please show this information... I will gladly do the additional calculations and come up with numbers.
If Tannehill was named the starter in week 7, why is your cutoff point game 9 for determining Henry's carries per game with Tannehill as the starter? https://www.pro-football-reference.com/players/H/HenrDe00/gamelog/
Guy without specific breakdowns for each game with Tannehill and Mariotta in the games they both appeared I looked at Henry's usage early and late in the season... Went to ESPN... Season splits... carries per game 1-8 Carries per game 9-16 It showed that Henry was not being worked any harder in the second half of the season than he was in the first half.
And your link does not break down carries per Tannehill as opposed to carries per Mariotta. So unless you broke down every minute of game tape in the games where they both appeared you are pulling your numbers out of your arse. please tell me how you calculated the carries per game exactly.
Sure, but it's easy enough to simply determine the carries per game he got with either quarterback as the starter: https://www.pro-football-reference.com/players/H/HenrDe00/gamelog/2019/ The irony here is that in post #6366 you implied that I was being misleading.
There is a long established precedent that you do not mix playoff data with regular season data. This goes double for small sample sizes. Comparing a blend of data set A (regular season) and data set B (playoffs) against data set A is totally inappropriate. Drawing a conclusion that relies on blending regular season and playoff data together when you know that the regular season data alone does not support your hypothesis is beyond dubious. The evidence that you knew the regular season data did not fit your hypothesis is the fact it is kept separate and uncontaminated from playoff data, so you had to do extra work to blend them together to find a result that fitted your hypothesis.
He ran slightly more often in the games Tannehill was QB (21.1 vs 18.8). It is purely a function of playing with the lead more often when Tannehill was the QB. Not hard to understand. On the season, he averaged 16.8 carries per game in losses and 22.4 carries per game in wins. How big a difference depends on how you treat the Denver game. Henry had 12 carries with Mariota at QB and only 3 with Tannehill at QB. If you prorate the carries in that game, the carries per game is very close.
Finfan That was my point all along. The numbers are identical in terms of carries per game for the regular season. Furthermore the Titans became the #2 scoring offense in football for the games that Tannehill started as QB and that happened without Henry getting a heavier workload. The fact that Guy would not disclose how he was breaking down games like the Denver game told me all I needed to know. Trying to lump the playoffs into it also spoke volumes... He was trying to manufacture numbers to rally behind
That is his MO. Nevermind that the three playoff games were played on the road in NE, Baltimore, and KC. He also fails to recognize that the first half of all three playoff games featured a balanced run/pass ratio: vs NE: 14 Henry runs, 9 Tannehill passes vs Balt: 11 Henry runs, 10 Tannehill passes vs KC: 16 Henry runs, 17 Tannehill passes The first two games got skewed in the 2nd half because of game situation (close game at NE and a lead at Balt). The Patriots defense was guarding the pass and daring the Titans to run the ball. Against Baltimore, the Titans got off to a big lead (throwing the ball) and Baltimore just could not stop the running game after that. No need to keep throwing the ball. Grinding the clock is how you win in that situation.
BTW, Henry had 132 carries in the first half of games and 177 in the second half of games. That doesn't look like a team that is running early to setup the pass. That looks like a team that is running the ball to close out wins.
He also misrepresented a chart the showed passing EPA against stacked and non-stacked boxes as solely about Tannehill, when it was for the Titans passing offense. Mariota started just over 6 1/2 games.
I mentioned closing out games with a running back previously... Like the Patriots did when they had blount... Get up by more than a score in the second half and start using Blount.
Its funny you say that... Mariota has been criticized by not throwing the ball to WRs hand where it needs to be, wich made our WRs look awful than they actually were. Ryan is doing the opposite, his throws are giving WRs chances after catch.
I have it on good authority that Tannehill gets to throw to wide open receivers because, you know, Derrick Henry..... stacked boxes..... and stuff....
I have to ask... What are you basing it on then? Preconceived notions? Lol. Honestly, though, it's hard for me to understand you, of all people, going outside of stats on this.
I believe he was comparing the improvement of Tannehill in Miami to Tenn vs Henry before and after Tannehill. Looking at that comparison, you can make the case that Tannehill improved more and therefor got the bigger benefit. But, IMO, the comparison is invalid since so much was different between Miami and Tenn. Coaching, supporting cast of offense, defense, scheme, healthy of the QB, etc. Literally everything was better. IMO, the better comparison is the 16 or so games that Tannehill had between 2014 and 2016 with his OL in tact. He was healthy, had a decent OL, a running game that was working, and he put up top 10 numbers in terms of efficiency (measured by passer rating).
Right, that's my point. If you thought Tannehill was at best an average QB, then you're going to attribute his success in Tennessee to anybody but him, and cbrad absolutely thought that. But you can't compare Tannehill in Miami to Tannehill in Tennessee. You can compare Tennessee last year without Tannehill to Tennessee last year with Tannehill.
This passes are amazing. A breath of fresh air to see a QB with good accuracy and arm strength. Now go look at Mariotas highlights and look at the way the ball floats in the air...
Part of the issue here is that several people involved in this debate are considering Tannehill's performance in 2019 to indicate validation of their belief that his surroundings in Miami were terribly inadequate. Consider however that that could be true, but it isn't necessarily the case. It's possible that his surroundings in Miami were no different from average, and that his surroundings in Tennessee in 2019 consisted of a perfect storm of ingredients that vaulted his performance to what it was. In other words, his surroundings in Tennessee were possibly not just adequate, but exceptionally good, in terms of what Tannehill in particular needs to excel. If that's the case, then the question in terms of Tannehill's performance becomes, how likely are those surroundings to be replicated, because Tannehill's performance hinges on them. Again we've seen this with other QBs in the past -- Andy Dalton and Nick Foles for example, who enjoyed one great season apiece and who have never again performed at that level. Clearly they must've had exceptionally good surroundings for them in those single seasons, and those surroundings, whatever they were, have been impossible to replicate. This is why it'll likely take an analysis of next season, and how it differs from 2019, to adjudicate this issue definitively. And this is also why any judgment of a QB's individual ability can't involve his performance in just one season -- it has to involve his performance over several seasons. Obviously if a QB is dependent on a set of surroundings that can exist in only one season of a career and can't be replicated (i.e., Dalton and Foles), he has far less individual ability than a QB who can perform well across various levels of quality of his surroundings. Consider the following scenario: someone holds a gun to your head and demands that you bet everything you own on whether 1) Ryan Tannhill's performance in Miami represents the true reflection of his individual ability, or 2) his performance in Tennessee in 2019 represents the true reflection of his individual ability. The person holding the gun to your head then gives you the option of deciding now and betting everything you own, or waiting until the end of the 2020 season and using that information in deciding on your bet. Do you bet now, or wait until the end of 2020? And if you decide to wait until the end of 2020, why would you do that? Why isn't this open-and-shut for you already?
The Dolphins already proved it is the case. Just look at the turnover on the roster and coaching staff. If you want to name all the players and coaches from the 2014 through 2016 Dolphins that have excelled elsewhere (other than Tannehill, of course) go ahead. Take a look at the starting OL in 2015: Jason Fox Billy Turner Mike Pouncey Dallas Thomas Brandon Albert Three would be out of the league in 2 years or less, one is/was injury prone, and the other would take another few years to become a journeyman starter. The NFL decided that this was a bad line. Surely 2014 was better: Ja'Wuan James Mike Pouncey Samson Satelle Daryn Coledge Brandon Albert Two would never play another down in the NFL, one was a rookie, one was out of position and missed part of the year, and the last missed 7 games. Jesus man, give it up.
Note the following: https://theathletic.com/1552465/202...ading-way-kc-boasts-new-improved-run-defense/
Then why, from 2012 to 2016, was Tannehill ranked 14th of the 18 QBs for whom there are data, regarding how he performed under no pressure? Why didn't his performance vault up to the level of his supposed individual ability when the conditions were especially favorable (i.e., no pressure) during that time period? https://www.footballoutsiders.com/stat-analysis/2017/quarterbacks-and-pressure-2016