In terms of what we're talking about now, that's indeed the issue, because you could very well base your personnel strategy on maximizing the performance of Tua specifically, in a way that would actually undermine the team if Tua can't improve enough himself. Again consider the weak impact of offensive line play, relatively speaking. The team could believe Tua needs the best offensive line in the league to function adequately, and then pour lots of resources into the line, only to find out that Tua's individual ceiling is too low and his performance isn't a function of offensive line play. At that point the team has lots of resources tied up in the offensive line and can't put them toward more important areas. The same could happen with a running back, a dime-a-dozen position that has no relationship with QB performance unless we're talking about only Derrick Henry perhaps. What I would do is get a solid tight end who can play inline, block well, and also be a big reliable target over the middle. A Mark Andrews type. That will translate across QBs and offensive systems and not be hooked to Tua specifically for its merit. Same thing at receiver -- go get the go-to dominating type the best teams have (Adams, Chase, Hill, Kupp, Diggs). That translates beyond Tua and whatever system benefits him. As far as the offensive line goes, I would do whatever makes it simply average. Go get a quality (good but not exceptional) center and a right tackle. Leave the rest alone. They've put enough resources into Hunt, Eichenberg, and Jackson. Continue to pay those players peanuts on their rookie contracts while they develop, and direct the salary cap money they save elsewhere. Whatever QB this team has, whether it's Tua or someone they move on to, he should be able to function fine behind a merely average offensive line.
I think this deserves its own thread. If it’s under the frame of Tua, it’s going to continue to stay in the Tua isn’t the answer vs he doesn’t have pieces around him debate that won’t go away until we find out either way.
I am willing to spend a couple years of us saying "man our offensive line is actually too good" just to mix things up a bit
You sound like someone who used to post here often. I believe that the idea that QB play is unaffected by the oline is ludicrous. Of course it's affected.
That makes you potentially the Eagles or Browns for example -- teams with great pass blocking in 2021 and insufficient QBs, who likely aren't going anywhere. With Tua, I'd rather aim toward being the 49ers -- average pass blocking, an above-average but not exceptional QB whose performance is maximized by the system, great targets in the passing game (Kittle and Samuel), and enough of a run game to make defenses at least think about it. That system has gotten a ton out of Jimmy Garoppolo -- he was actually the 5th-best QB in the league in EPA per pass dropback in 2021, and here the 49ers are playing for the Super Bowl tomorrow. I think that's probably our team ceiling with Tua.
Certainly it's affected, but it's affected across teams in such a way that the variation among teams in offensive line play is far smaller than the variation in other areas important to QB play. Again, if every offensive line played exactly the same way (hypothetically), they'd certainly affect QB play, but the degree to which they caused variation in QB play across teams would be nil. However QBs varied under those conditions would be entirely a product of other variables. The reality of the league isn't all that different from that, and all it takes to acknowledge that is the realization that best offensive line in the league isn't all that much better than the worst one. And all it takes for that realization is the awareness that when you're watching the Dolphins' offensive line play, the best offensive line in the league -- the one you're not watching at that time -- isn't playing all that much better.
I get what you guys are saying about the return of investment in the OL isn't as good or significant as what can be had investing in WR's for example. However, at this point, that's just a high-level strategy. If all of this is just to say that Miami should invest in WR's/TE's etc. then there's really not much more to say about that. The question becomes, in what way should Miami invest in WR's etc. Is there any insight into that, or can the data only go so far as to say, invest in WR's? EDIT: Just saw your post above. "What I would do is get a solid tight end who can play inline, block well, and also be a big reliable target over the middle. A Mark Andrews type. That will translate across QBs and offensive systems and not be hooked to Tua specifically for its merit. Same thing at receiver -- go get the go-to dominating type the best teams have (Adams, Chase, Hill, Kupp, Diggs). That translates beyond Tua and whatever system benefits him. As far as the offensive line goes, I would do whatever makes it simply average. Go get a quality (good but not exceptional) center and a right tackle. Leave the rest alone. They've put enough resources into Hunt, Eichenberg, and Jackson. Continue to pay those players peanuts on their rookie contracts while they develop, and direct the salary cap money they save elsewhere. Whatever QB this team has, whether it's Tua or someone they move on to, he should be able to function fine behind a merely average offensive line." We've got decent cap space this year, so we've got a bit of freedom, but in general, I think the team should seek out at least one decent veteran OL, maybe a C, who can help bring experience and leadership, as well as talent. Maybe one more guy lower down the draft of in FA for depth. Heck, two veterans might even work. If we can nab a great WR in FA that's okay but the good ones are really expensive. Much better value if you can get one in the draft on a rookie contract. Haven't looked at prospects this year so no idea if there is one. Based on what you're saying about the data though, the advice would be to spend our first round pick on a WR above all else, if there's a great one to be found. So maybe one or two picks on WRs this year?
Feel free to start one. The point here is that Tua is going to be our QB for one more year and so this is a space for talking about the direction of the team with that in mind. It's not necessarily about building specifically for Tua. The best moves to improve this team may well be the same regardless of who is at QB.
Please stop it with the oline is not important crap. Your analytics relationship between passing efficiency and oline play cannot be applied universally to every situation and every QB. Tua is not the type of QB that can overcome a bad offensive line. He's doesn't have the elusiveness, athleticism, arm strength or durability to scramble and play off schedule with pressure bearing down on him while taking hits left and right like a Mahomes, Allen, or Herbert can. Maybe we should try something new for a change and actually have good offensive line for once. We tried the go get a big time receiver route before with Tannehill and other non elite QBs before him, and it was still not enough. Tannehill didn't have any real success until he went to the titans and had a good oline. Non elite, none scrambling QBs need good offensive lines. It's common football sense, lay off the numbers a bit please.
It should be noted that Mahomes didn't overcome his offensive line last year so they went out and made moves
Why settle for a non-elite QB? This really shouldn't be about what's best for Tua, even if that's the aim of this thread. What's best for the team is paramount. The better solution is to find that Mahomes, Allen or Herbert type QB even if that's much harder to do. That shouldn't be a "it would be great if we could do that but we can settle for less". No, it should be the primary strategy. Also, the only reliable way to put together a good OL given how bad we've been at drafting (and we've drafted quite a few OL with higher round picks in the past) and how incapable we are at finding a good Ol coach is through FA. Spending a lot on OL in FA is a bad idea. The FA WR market is really good this year. We have a serious chance at adding an elite WR opposite Waddle. That would provide us with one of the best WR units in the NFL, which would benefit any QB including Tua. So given that Tua is most likely our QB for 2022 I'm fine with using high picks on OL and RB, but there's no expectation on my part that we're going to build a good OL through that. The goal is to just improve things a bit more — average is fine — and either Tua someday shows he can overcome adversity or we find a QB that can.
It should also be noted he won a SB with only an average OL before they made those moves. KC has an elite QB so they have the luxury of pouring resources elsewhere. We're not in the same situation. Oh, and their offensive production was elite both before and after that. It didn't change much in terms of production.
You've said Watson is elite. Okay, great. Houston is one of the worst teams in football...and they became progressively worse after drafting Watson? How's that possible with an elite, top-3 QB?
You can certainly try something new and see how it works. Just realize that the statistics relevant to what's most strongly associated with pass efficiency don't support it. It's about like saying, "let's try to get a great punter and see how much it improves our QB and passing game -- just to try something new." If it's indeed common sense that non-elite, non-scrambling QBs need good offensive lines, it shouldn't be difficult to produce a laundry list of such QBs whose performances were elevated sufficiently by them. Who are they? If you're an advocate for that approach, let's see you support it with actual data. And you'll have to start with a QB other than the one you nominated above -- Ryan Tannehill. The correlation between his EPA per play and PFF pass blocking grade, year-by-year of his career, is -0.27 (yes that's a negative sign). As his offensive line has pass protected worse, he's passed the ball better. EPA = nflfastR expected points added per play (efficiency) CPOE = nflfastR completion % over expected (accuracy) DVOA = F.O. defense-adjusted value over average (efficiency) DYAR = F.O. defense-adjusted yards above replacement (gross production) PFF = individual PFF offense grade QBR = ESPN total QBR; measure of win probability added per play Sack% = sack rate (sacks/total dropbacks) AirYds = Air yards per attempt CP = nflfastR completion probability based on situation & air yards PFF Recv = team PFF receiving grade PFF Pblk = team PFF pass block grade PA/G = team points allowed per game
Because the second-strongest predictor of winning in football is pass defense, and Houston's has been routinely poor. It isn't just what your own QB does -- it's also what you let the other team's QB do.
No they didn't. They were 1-9 without Watson the first year, while he was 3-3 so that was an immediate improvement. Then they went 11-5 and 10-5 with him starting and made the playoffs. It was only the final year they were bad at 4-12 with one of the worst defenses in the NFL. Watson still performed elite that year coming in 2nd in passer rating, which shows he wasn't the problem. Obviously win% depends more than on just the QB. The question here is what to do regarding the QB: build around Tua or get an elite QB. There's no expectation you can give Tua as good a surrounding cast as Tannehill let's say and see him match Watson's production. As I said before you need a QB that can overcome adversity in the playoffs. You're not getting around that except in maybe 1 out of 7 SBs won in the last 20 or so years. Playing the odds says you need to prioritize getting an elite QB.
Wait what? That doesn't make sense. Can you post PFF's pass blocking grades for each year? Either PFF's rankings make no sense or there's some mistake made in the calculation. Tannehill has definitely had better pass protection in Tennessee than in Miami. Also, it's well known that passer rating takes a pretty big hit under pressure (on average) so something isn't right here.
Definitely counterintuitive. The PFF pass blocking grades are in the image at the bottom of the post you quoted. Something that may explain it: https://www.pff.com/news/pro-z-quar...pff-data-study-of-who-controls-pressure-rates The correlation between Tannehill's PFF QB grade and his EPA per play, year-by-year, is 0.86. The EPA per play/PFF pass blocking grade correlation year-by-year is indeed -0.49 however (EDIT: corrected to -0.27). The correlation between his EPA per play and his PFF team receiving grade, year-by-year, is 0.67. All of those stats again are in the image at the bottom of the post you quoted.
Well.. the reason I asked is because when I take your "EPA" column divided by the "PLAYS" column, i.e., EPA per play (if that's what that's showing) and calculate the correlation to the "PFF PBLK" column I'm getting -0.0691 so basically no correlation, which is at least believable. You sure you didn't input something incorrectly?
The "EPA" column is EPA per play. Should've posted this along with it: EPA = nflfastR expected points added per play (efficiency) CPOE = nflfastR completion % over expected (accuracy) DVOA = F.O. defense-adjusted value over average (efficiency) DYAR = F.O. defense-adjusted yards above replacement (gross production) PFF = individual PFF offense grade QBR = ESPN total QBR; measure of win probability added per play Sack% = sack rate (sacks/total dropbacks) AirYds = Air yards per attempt CP = nflfastR completion probability based on situation & air yards PFF Recv = team PFF receiving grade PFF Pblk = team PFF pass block grade PA/G = team points allowed per game
OK thanks. Well now I'm getting -0.267 which is still different from the -0.46 you got. Anyway, if that's the case I'd say this is a reason to be skeptical of how PFF grades pass blocking. They're saying that in 2020 and 2021 pass blocking was far worse for Tannehill in Tennessee than in any year he played in Miami?? Who seriously believes that. I think this hurts PFF's credibility more than anything else.
OK after getting -0.267 in Jamovi I figured out the issue -- copying and pasting the table into Excel made the -0.17 EPA in 2018 translate as something other than a negative sign, like a dash or whatever. Consequently it read that as no value.
It's very simple: the more time to throw, the better (unless you're talking a 3 man rush scenario). So, yeah, we all understand, better QB don't need AS MUCH help from the oline. They don't need as much time to be in their comfort zone. But worse QBs, more average QBs, need more time to throw to be in their comfort zone. So improving their oline can make a HUGE difference, for a guy like Tua. If you want Tua successful next season, you've got to give him AT LEAST a statistically average oline, an average run game, and some receiving help. That's a lot to invest in a guy for one season. Can't believe DolphinGreg is sucking me back into this argument again.
Just for some perspective - is there anyone here, at all, who doesn't agree with Miami pursuing an average OL this season? If not, what are we arguing about?
The part I bolded and underlined above actually isn't true. See here for example: https://www.sharpfootballanalysis.c...ata-patrick-mahomes-josh-allen-lamar-jackson/ The percentage of plays with positive EPA falls below 50% just before 3 seconds after the snap. If teams throw the ball beyond that point, they're more likely to have a negative EPA play than a positive EPA one. Combine that with this: https://www.pff.com/news/nfl-the-perfect-timing-a-deeper-dive-into-time-to-throw-data ...and you start to see why the relationship between offensive line play and passing efficiency isn't so commonsensical.
Can't go by rankings because so many teams have similar percentages near the mean. The average PBWR in 2020 was 57.4% and 53% isn't overly bad: z-score = –0.67. In 2021 the average was 59.6% and 56% corresponds to z-score = –0.62. PFF's scale is very different and I'd bet a 53.5 is really low for them. If you can find team PFF PBLK ratings for all teams from 2020 or 2021 we can compare. Not sure if you have a subscription for PFF, but I don't and I haven't easily found team OL ratings from PFF by googling. But it's z-scores we need to compare.
PFF stopped listing the team pass blocking grades after 2018, but from 2021 we do have this -- note the highlighted portion: From 2020: https://www.pff.com/news/nfl-final-2020-offensive-line-rankings
This was PFF's final OL breakdown for the regular season this year: 32. MIAMI DOLPHINS (NO CHANGE) LT Liam Eichenberg | 50.7 LG Austin Jackson | 49.9 C Michael Deiter | 60.6 RG Robert Hunt | 67.4 RT Jesse Davis | 52.5 Miami’s offensive line surrendered a league-leading 235 pressures this season and recorded the worst pass-blocking efficiency score in the NFL. The unit did this despite being well protected by a quarterback getting rid of the ball quickly and by the team running the third-most RPOs in the league. The offensive line was run blocking on almost one out of every five passing plays, removing the chance of being exposed in pass protection. Miami gambled that their young players would develop this season and the line would improve, but that unquestionably backfired. https://www.pff.com/news/nfl-final-2021-offensive-line-rankings
OK, if 53.5 is 31st, then assuming a normal distribution with 32 teams the average z-score would be –1.7 (based on 1000 simulations). That's over 1 standard deviation below where PBWR puts it. Like I said, to me this strains PFF's OL ranking credibility.
Look, you can give whatever stats you want, Greg, but you will never be able to convince me that average QBs don't benefit from better oline play than elite QBs. Taking ONE sentence of my post is incredibly disingenuous, because the entirety of my post is talking about average QBs benefiting from more time to throw. Also, I'm not saying that having 6 seconds to throw is better. Also, when those QBs you cited have MORE than 3 seconds, you need to know why. Is it a 3 man rush with 8 in coverage? Well then the Qb will have forever to throw, but will have a difficult time finding an open receiver. This is where dependence on stats is faulty. So, for a QB LIKE TUA, the SINGLE most important thing to do is to improve his oline. He needs more than 1.5 seconds to throw. Tua would probably kill someone to get an average of 3 seconds to throw.
DolphinGreg? His posts are very different IMO, if you look at some by searching. I think Dolfanalyst was formerly The Guy, at least going by his writing style. Very different from DolphinGreg who I also haven't seen post here in awhile.
Here's a little bit of PFF on OL - section descriptions/time stamps are off. OL talk in general starts at 56:20: Being a brilliant OL player is not important, being a reliable player, keeping a clean sheet, is what's important. Teams need to have decent players across the board, a weak player will be attacked. There's a short Miami section earlier starting at 10:31.
OMG. Please. Someone bump up that Statistical Analysis Methods For Football thread. This is getting bad again.
LMAO, you can't aim to build an "average NFL line." It doesn't work that way. The QB is one person- so you could sign an average QB. Or RB, WR, whatever...you have stats on that one player to figure that out. But the line is five players and what the guy next to you does directly effects what you'll do. Line is about chemistry and working together, more so than any other position in football. Coaching is critical. Scheme is critical. Even the other offensive players help or hurt the line. Think about bobsledding for a moment- that's a lot like offensive line. Four guys in a hollowed out rocket sliding downhill at 100+ MPH. If one messes up, they're pretty much knocked out of medal contention. If two or more mess up, they might all be dead. Line is the same thing- you can't sign 5 average players, slap them on the butt, and they magically become an average group. There's so much more involved and the end goal isn't to be "just okay". The real goal is five guys working in harmony and we don't have anything near that at the moment due to coaching, chemistry, etc. This discussion is just ridiculous- that's not how football works. And you can't learn football by looking at a stat sheet.
Some other stats of note: Although the Dolphins spent the second-least money in the league on its offensive line personnel in 2021, the correlation between pass block win rate and offensive line salary cap expenditure, team-by-team in 2021, was only 0.29. Over 91% of the variance in pass block win rate in the league is unexplained by salary cap money spent on the offensive line. The Chiefs for example had a great offensive line in terms of pass block win rate in 2021, while spending only $1.7M more than the Dolphins on its offensive line personnel. Compare that to Jacksonville, which spent $34.5M more than the Dolphins on its offensive line personnel (most in the league) in 2021, yet had only a middling pass block win rate. The Dolphins spent more money than any team in the league at two positions in 2021: WR and CB. Obviously they didn't get the return on the WR money. Again this was a big issue in my opinion. What was meant to be a formidable group of Fuller, Parker, Waddle, Wilson, and Gesicki ended up being only a crippled version of that. https://overthecap.com/positional-spending/