This is a kind of follow on from this post- http://www.thephins.com/forums/show...ehill-should-be-a-better-QB-now-Lazor-is-gone In that thread there was a lot of discussion about how coaching (specifically Lazor's playcalling when behind) hurt RT. In that thread the assertion was made that RT ha been adversely affected by (1) poor OL play. (2) Poor coaching (3) lack of commitment to the run game and (4) inability to audible. Now for (1) and (4) I can't think of a good way to analyze statistically. But I have decided to see what we can find from the run game stats. Some Housekeeping to start with. - Every QB should have a higher average passer rating the more the team runs the ball. Teams that are ahead on the scoreboard will rush more to use up time. Teams that are behind will pass more. Your QB play is a big part of whether you are in front or behind. - Because each team has different offensive philosophies and skillsets you can't compare to NFL averages. It has to be a team based analysis. that's too much work for this little black duck to do. If someone wants to do they should be looking at teams where a non-rookie starter completed all 16 games, because back-up QBs and rookie QBs have a higher ran-pass ratio than starting veteran QBs. - Run pass ratio is more important than rushing attempts. Rushing attempts is linked to how many plays your offense runs not the teams commitment to the running game. - I didn't look at rushing yards, as I am more concerned about commitment to the rushing game rushing. I looked only at the last 2 years under Lazor because the Sherman offense was structured differently. RT17's passer rating over the last 2 years is 90.2. The top 8 games of run/pass ratio: 164 completions from 234 attempts for 1750 yards. 18TD/4Int. 110.2 passer rating 268 rushing attempts from 502 plays. 53.4% rush/pass ratio. The middle upper quartile 196 completions from 286 attempts for 2157 yards. 14TD/6Int. 98.2 passer rating 196 rushing attempts from 482 plays. 40.7% rush/pass ratio. The middle lower quartile 184 completions from 292 attempts for 1910 yards. 7TD/5Int. 82.7 passer rating 153 rushing attempts from 445 plays. 34.4% rush pass ratio The bottom 8 games 211 completions from 364 attempts for 2178 yards. 12TD/9Int. 76.0 passer rating 127 rushing attempts from 491 plays. 25.9% rush/pass ratio Under Lazor there were 16 games were the rush/pass ratio was higher than 36%. Of the 16 games with a rush/pass ratio of 36%+RT had a 100+ passer rating 9 times. Of the 16 games below 36% he had one 100+ passer rating. While I expected RT17 to have a lower pass rating with lower rush/pass ratios three things surprised me. 1) The extent of the drop off. 2) I was expecting a more linear trend, but it seems more like a huge drop at the halfway point. 2) The drop off happens at lower than 36% run pass ratio, which is IMO a really low number.
Nice analytics, good writeup and summary. Now, if you can sum up the difference in qbr rating when he is or isn't PISSING BLOOD...
Actually it doesn't need to be strong. Average should be OK. But it does show that abandoning the run and passing 3 times out of 4 is a poor strategy when RT17 is your QB. It's like putting all your money on one number at the roulette table.
Nice Post Pauly.. He does play well when play action is in the repertoire and when he's in manageable down and distances.. hopefully an improved scheme, playcalling and more talent for the run game, it will improve his rating 5 to 7 points.
hmmm. Just thinking that scrimping to save cents on Lamar Miller might end up costing us dollars in RT's performance. Also it makes RB a target in rounds 1-3, and probably means no trade up unless it is to get Ezekiel.
Gase's offenses: Broncos ran the ball (vs. Pass) 40.6% of the time in 2013, 42.2% in 2014. The Bears 47.2% in 2015. Dolphins - 2013 37.0%, 2014 40.1%, 2015 36.9%. Even in Manning's ridiculous 2013 season with all the passing records and whatnot, they still ran the ball more (percentage-wise) than any Dolphins season the past three years.
Good post Pauly. Some seem to expect any decent QB to be a miracle worker who can put up top numbers in any situation, but that's just not realistic. RT17 has proven, as you've shown, that even with the other handicaps that he was under, as long as the team tries to run the ball, he can be a good QB. Why this is ignored, or even looked down upon, by some is beyond me.
Been saying this for a few years now: every franchise QB - with the exception of Aaron Rodgers, who sat - had a strong running game or feature back behind them for their first couple seasons when they came into the League. Brady (run-heavy offense), Ben (Bus), Peyton (Faulk), Eli (Tiki), Rivers (LT), Brees (LT), Russ (Marshawn), Flacco (prime Rice), Matt Ryan (prime Turner going for 1800), etc. It's a thing. And here we've saddled Tannehill with the mighty Daniel Thomas and a slow-developing under-utilized Lamar Miller.
Just to recap a little bit to look at analysis from Fin-d's angles and their effects on Tannehill. 1) Poor coaching. (offensive playcalling) Under Lazor RT had a passer of 101.2 when tied or ahead and a passer rating of 84.6 when behind. This differential is (a) much bigger than the NFL average and (b) did not exist under Sherman. So the drop in performance IMO is more attributable to Lazor than RT (discussion:http://www.thephins.com/forums/show...ehill-should-be-a-better-QB-now-Lazor-is-gone) 2) Poor OL. In games where Pouncey, Albert and James started and finished RT had a passer rating of 107.0. A quick estimate is that his passer rating in other games comes to about 85.2 http://www.sun-sentinel.com/sports/...roof-ryan-tannehill-perfo-20160405-story.html 3) Lack of commitment to the run game. In 16 games with Lazor where the Phins ran more than 36% of the time (i.e. more than average) RT had a passer rating of 103.6. In the games they rushed less than average his rating was 79.0 4) Inability to audible. I can't think of a way to measure but it is definitely reflected in (1) and (3) Obviously there is a lot of overlay in the different analyses, but each of 1, 2 and 3 is looking at different slices of data but the story is consistent. Firstly we should dismiss the notion that Tannehill is an 'average' QB. He has either performed at probowl level or below replacement level,with very little in between. Different people may take away different conclusions from that. However I believe that the probowl level QB performance is from such large slabs of data that they represent his true ability, and that his poor performances are a reflection of a cumulative effect of several poor circumstances. Other people may conclude he is a below replacement QB a la Geno Smith who has had a few good games here and there. My response to that is the data fits the hypotheses. If RT17 was truly inconsistent you shouldn't get such strong patterns (also the hypotheses were proposed before the data was analysed). Secondly for the team to get the best out of him they need to commit to the run game. Some QBs can be effective if you give up the run and go pass wacky. RT17 is not one of them. I'm not saying RT17 can't be effective passing 50 times a game, but that your gameplan shouldn't rely on him passing that much every week. Thirdly, my respect for our previous coaching staff diminishes every time the data gets sliced. I'm just a nobody on the interwebs typing on a fanblog and I can prove pretty easily that common fan complaints about their personnel management, game calling and use of RT17, were valid. Seriously these guys should have been way better than that,
Thank you, thank you, thank you. I honestly believe its all tied together too. I think the overwhelming problem is oline. I think there are 4 on-field ways to deal with a bad pass pro oline: - Run the ball to keep the defense honest. - Audible out of the play based on the looks the defense is giving. - Scheme around the oline's deficiencies in other ways, like extra protection, for example. - Have elite elusiveness. Of those 4 tools, Thill doesn't have the last one and the other 3 were taken from him against his will. That means to me, he doesn't need a run game per se, he just needs some tools to deal with a crap oline. Get him a decent oline that stays healthy, he'll be fine without audibles and a run game. Has a crap oline....then let him audible proper and/or don't abandon the run. He's been told to build a house in 4 months and then they took his hammers, saws and screwdrivers.
Actually some of it was me trying to disprove Fin-D's theory. My theory was it was more on the playcalling. Either way what really sucks is that the problems were evident in the data in 2014. Then we changed nothing of substance and exactly the same things happened again in 2015, only worse because the rest of the league knew it was coming. Not just not fixing the O-line, but not going out to get another Knowshown Moreno type back to spell Miller, gambling on rookie/unproven guards when we could have gotten average NFL starter quality guards without breaking the bank. Basically in the 2015 off season the FO and coaching staff went out and bought new rims and a new paint job for their car, but didn't repair the leaking headgasket or busted fanbelt.
Just when I think you can't be any more ignorant on things, I go and click "View post" on one of your posts, and I realize immediately why I blocked you to begin with. Hahaha
I really didn't want to do this, since I thought it was self evident, but. Same exercise as the first post, but looking at rushing yards per attempt instead of rushing percentage. The top 8 games of rush yards/attempt: 153 completions from 263 attempts for 1723 yards. 13TD/5Int. 86.4 passer rating 268 rushing attempts for 1220 yards.6.4 yards/attempt. The middle upper quartile 178 completions from 272 attempts for 1960 yards. 15TD/9Int. 91.2 passer rating 196 rushing attempts for 962 yards. 4.8 yards/attempt. The middle lower quartile 206 completions from 300 attempts for 2067 yards. 14TD/3Int. 99.4 passer rating 153 rushing attempts for 769 yards. 3.9 yards/attempt. The bottom 8 games 218 completions from 341attempts for 2455yards. 9TD/7Int. 85.6 passer rating 127 rushing attempts for 417 yards. 2.6 yards per carry. I also didn't want to do this either since I wanted to look at \rt17s passing, not team performance but your lack of knowledge drives this. Win Percentage of each quartile Rush/pass ratio 1st quartile: 100% 2nd quartile: 50% 3rd quartile: 25% 4th quartile 0% Rush yards/attempt 1st quartile: 50% 2nd quartile: 37.5% 3rd quartile: 75% 4th quartile 12.5% (edited because I lost track of 1 win) So "stuffing the run" has a negative correlation with RT17s performance. i.e he generally performs better in games with a poor yards/attempt than he does in games with a good yards/attempt Further "stuffing the run" is much worse predictor of beating the dolphins than rush/pass ratio.
I love what you're doing, Pauly, I'm just afraid the usual suspects will ignore it, and trot out the same defeated arguments.
Another interesting thing for me was sorting the various offensive categories and stacking them against win percentage. The best predictor by far of win/loss is rush/pass ratio. I don't think at all that it means teams should be rushing more. I think it is much more a reflection of teams that are ahead on the scoreboard rush more than usual to eat up the clock. Which brings me to a criticism of advanced stats. I don't care whether PFF, DVOA or whatever site or method. These look at plays in terms of their efficiency purely as a product of yards gained, or significant events like TDs, fumbles etc. However, NFL teams have two resources (a) points and (b) time. The next level for advanced stats is to combine yardage efficiency with time efficiency.
Well there are limits to what this type of analysis can do. It is in no way a replacement for film study, I don't try to do film evaluations because I'm not enough of an expert to do it, For example finster and I had a discussion on the previous thread about whether Tannehill's reduced performance when behind was due to him succumbing to pressure or whether it was due to poor playcalling by Lazor. Finster's opinions were based on what he saw on film and mine were based on statistics. Ultimately statistics alone or film study alone won't resolve that dispute, for something like that it has to be done hand in hand together.
I don't mean to ramble but this has got me thinking a bit more. It explains why Lazor going pass wacky whenever he was behind bugged me so much. Yes, passing is the best way to score points in the NFL. However passing and failing to score points gives control of the clock to your opponent. They can use that extra time to score more points, or simply to bleed away time to deny you the opportunity to use it. There are times it makes sense to give up the run and go all out passing. When you're behind by 3 in the first quarter isn't that time.