starrymessenger Posted November 13, 2019 Posted November 13, 2019 4 hours ago, BarleyNY said: Just going to hit some points here: - DVOA is a much better gauge of performance than yards or points for offense and defense. You seem knowledgeable enough to know this but have chosen to pretend otherwise to try to make your point. - Whether or not it would’ve been a viable strategy to run Singletary a lot more against Cleveland’s stacked box is pretty irrelevant to the main point here. We can debate whether that would’ve worked all day, but thats not the issue. - The real issue is that doing so should not have been necessary. It’s that Allen could not beat simple, undisguised man coverage. This wasn’t Allen getting beaten by Belichick, it was him being beaten by a bad DC running a basic scheme that he’s seen since high school. And if it had been some bad reads here and there against a tougher scheme I wouldn’t have an issue. But he just couldn’t make the throws. I don’t see how that improves and I don’t want to waste another season watching it happen again. - Lastly, people keep saying he’s improving. I see a few areas where he has, but overall he is still poor and it’s about exactly as poor as last season. I’d guess that a big reason for that is that teams have learned how to play him. Let’s take a look. His Total QBR has dropped from 52.0 in ‘18 to 36.0 this season. His Defense Adjusted Value Over Average (DVOA) went up from -35.9% to -23.0%. That’s still terrible for a starting QB. There have been 41 QBs this season that have performed better according to that statistic. So if he really is improving overall, it’s not by nearly enough. If he doesn’t make a big jump during the rest of this season I think the Bills should look for an upgrade. https://www.footballoutsiders.com/stats/qb/2019 I think that was true in the past, but that the rookie salary cap has made it easier to move on from high picks. Still, you might be right. I'm not an expert by any means, but I think that DVOA, DYAR, QBR as statistical performance measures are not a basis for prudent and informed decision making by NFL teams and indeed they don't serve that purpose (though of course they may more or less in a given case align with the team's QB evaluation). They are generalized/ comparative references that can provide some insight into how well (or not) a QB is playing over the relevant period, though without direct correlation to important inputs, for example for supporting player and coaching talent. According to these measurements in 2019 Mayfield, Darnold and Allen are all pretty bad. Maybe all three teams should move on from their young QBs. And some pretty good QBs, or at least QBs having had some success, are similarly bad in one or more of these statistical catagories. So QB evaluation is a lot more complicated than that. So Allen didn't do well enuf for you in beating man coverage (I assume you might be more forgiving if you felt the opposition was better at running a man scheme). So the Bills should cut their losses. In New Jersey many Jets fans are convinced that Darnold cannot read a defence to save his life and should be discarded. To me Darnold and Allen are talented prospects who have been drafted by teams that are for perhaps different reasons somewhat dysfunctional on the offensive side of the ball. They have also been thrown directly into the action without the grooming benefits that Garropolo or Rodgers were able to enjoy. So to me you have to show patience and absorb the downside consequences in the meantime. My take is that there is a reasonable chance that Allen (and Darnold) will improve as I hope and expect his teammates and coaches will improve.
BarleyNY Posted November 14, 2019 Posted November 14, 2019 2 hours ago, starrymessenger said: I'm not an expert by any means, but I think that DVOA, DYAR, QBR as statistical performance measures are not a basis for prudent and informed decision making by NFL teams and indeed they don't serve that purpose (though of course they may more or less in a given case align with the team's QB evaluation). They are generalized/ comparative references that can provide some insight into how well (or not) a QB is playing over the relevant period, though without direct correlation to important inputs, for example for supporting player and coaching talent. According to these measurements in 2019 Mayfield, Darnold and Allen are all pretty bad. Maybe all three teams should move on from their young QBs. And some pretty good QBs, or at least QBs having had some success, are similarly bad in one or more of these statistical catagories. So QB evaluation is a lot more complicated than that. So Allen didn't do well enuf for you in beating man coverage (I assume you might be more forgiving if you felt the opposition was better at running a man scheme). So the Bills should cut their losses. In New Jersey many Jets fans are convinced that Darnold cannot read a defence to save his life and should be discarded. To me Darnold and Allen are talented prospects who have been drafted by teams that are for perhaps different reasons somewhat dysfunctional on the offensive side of the ball. They have also been thrown directly into the action without the grooming benefits that Garropolo or Rodgers were able to enjoy. So to me you have to show patience and absorb the downside consequences in the meantime. My take is that there is a reasonable chance that Allen (and Darnold) will improve as I hope and expect his teammates and coaches will improve. While it is true that DYAR, DVOA and QBR are not perfect, they do work very well to gauge how well a player (QB) is performing. They also make efforts to account for surrounding factors. I am not sure what your point in saying you don’t think coaches use them. Teams evaluate all of their players constantly. I doubt they’d need ESPN or Football Outsiders to evaluate their starting QB. They’d obviously have a finer point on that evaluation themselves, but I’d very seriously doubt that it’d be a lot different. So I think we can safely look at those metrics and understand that we have a pretty good idea of the job Allen is doing. As for Darnold and Mayfield, they are not Allen. Each needs looked at on their own. I haven’t watched enough of Darnold to form an informed opinion. I’ll say that I didn’t love him coming out of college and that his numbers suck in the NFL. They’re almost as bad as Allen’s. It certainly doesn’t look good. I’ve watched a lot more of Mayfield. He was pretty good last season, but he’s been nearly as bad as Allen this season. The numbers bear that out. So since he’s shown something I stick with him and try to get him better coaching. Dunno about that last part. I’d be happy to roll with Allen if he looked promising, but he’s failing with some pretty basic stuff. And it’s the same stuff he had problems with in college. Hell, it’s not just Lamar Jackson and Baker Mayfield from his draft class that are outperforming him. Gardner Freaking Minshew is a 6th round rookie who is. Oh and so is Kyler Murray. Allen has a year in the NFL on both of them. Bottom line: We drafted a QB who was known to have poor accuracy when he came out of college. He still looks to have the exact same issue after more than a season and a half in the NFL. Why do you think that will change with another year?
starrymessenger Posted November 14, 2019 Posted November 14, 2019 11 hours ago, BarleyNY said: While it is true that DYAR, DVOA and QBR are not perfect, they do work very well to gauge how well a player (QB) is performing. They also make efforts to account for surrounding factors. I am not sure what your point in saying you don’t think coaches use them. Teams evaluate all of their players constantly. I doubt they’d need ESPN or Football Outsiders to evaluate their starting QB. They’d obviously have a finer point on that evaluation themselves, but I’d very seriously doubt that it’d be a lot different. So I think we can safely look at those metrics and understand that we have a pretty good idea of the job Allen is doing. As for Darnold and Mayfield, they are not Allen. Each needs looked at on their own. I haven’t watched enough of Darnold to form an informed opinion. I’ll say that I didn’t love him coming out of college and that his numbers suck in the NFL. They’re almost as bad as Allen’s. It certainly doesn’t look good. I’ve watched a lot more of Mayfield. He was pretty good last season, but he’s been nearly as bad as Allen this season. The numbers bear that out. So since he’s shown something I stick with him and try to get him better coaching. Dunno about that last part. I’d be happy to roll with Allen if he looked promising, but he’s failing with some pretty basic stuff. And it’s the same stuff he had problems with in college. Hell, it’s not just Lamar Jackson and Baker Mayfield from his draft class that are outperforming him. Gardner Freaking Minshew is a 6th round rookie who is. Oh and so is Kyler Murray. Allen has a year in the NFL on both of them. Bottom line: We drafted a QB who was known to have poor accuracy when he came out of college. He still looks to have the exact same issue after more than a season and a half in the NFL. Why do you think that will change with another year? I too doubt that football professionals need Football Outsiders or ESPN to evaluate their QBs (that was sort of my point). If they do they should be looking for other employment. They don't need these performance measurements to form an opinion on where their QB is relative to eg league averages. They independently know, or should know this because they are close to the action and their careers depend upon the accuracy of their projections, especially in evaluating young and inexperienced QBs. But more particularly the relevance of this information as a basis for making crucial decisions is by its very nature unreliable. It may give you the heads up on where your QB is now on a comparative basis but it doesn't tell you what exactly he (and indeed the offence as a whole) needs to do to improve and, more importantly, whether he (and it) can improve. I certainly wouldn't go so far as to say that the relationship of these measurements to the all important assessment is empirically accidental. Duration, with consistency of the results over time, can certainly be indicative, especially when, needless to say, a young QB fails to improve. But, y'know, that's not good enuf when it comes to projections. The past can foretell the future sure. Or not. Will Allen improve and develop into the QB we all want to see? Idk. But I don't need Football Outsiders or ESPN to know what the issue is. The Bills knew what the issue was when they drafted him. Some of the more astute posters on this board (and I don't mean me) also know what the issue is. Some say "accuracy". Some say "mechanics". Some say "anticipation". But all these are merely derivative results. What mainly drives all of them is the ability (or the lack thereof) to process with sufficient speed the live action. Allen's accuracy (including the ability to throw his receivers open) suffers when his mechanics deteriorate. His mechanics deteriorate when he has to overcome with raw physical ability a momentary lag between how the action is developing and his reaction time. But when he is on time (and it does happen with regularity if not with desired frequency) what he is demonstrably able to do places him immediately above the average of starting calibre NFL QBs. He made some throws against the Browns that not all starters can do, or at least not nearly as well. That's what having a sky high ceiling is all about. I don't know that "boom or bust" quite captures it because I think there may be a large performance range between these extremes, but that's what people are getting at when they use these catchwords. It's a big problem and an unanswered question (although in Allen's case it looks like it has been answered to your satisfaction, albeit negatively). We often hear that accuracy is something that a prospect either has or has not, and that it really cannot be developed, and there is certainly evidence for that. What that means is that a rare and not often found chip, processing speed, is missing from Allen's computer and that his is a model that cannot be refurbished to insert it. My own view is that he may, but not necessarily will, improve. Give him a clean pocket and he performs a lot better. He.has improved his short/intermediate range accuracy. Overall he manages his progressions better than he did last year. I hope to see continued improvement. If he (and the O as a whole) can get there the payoff will be well worth the growing pains. I'm not a fanboy and I'm not making excuses for him (or Darnold) but Lamar was lucky to land with Harbaugh, Roman and a much better complement of skill position players and O-line, and a much more cohesive unit overall. In Allen's draft year I actually had Lamar graded ahead of him (I pat myself on the back). I also preferred Rosen (I kick myself in the butt). But I was, and remain, enthusiastic about our QB, fully realizing that he may not develop the way we would like. I thought, and I continue to think, that he fully justified his draft position (I thought Lamar almost faling out of the first round was ridiculous - clearly Ozzie knew better when he made sure he got that fifth year option). I agree that the Bills should draft a developmental QB if they see one they like. There should always be one that the team has some commitment to on the PS. If you feel you need to you carry three. Gardner is a hoot and he looks like he can play.
frostbitmic Posted November 14, 2019 Posted November 14, 2019 If Daboll is worth his weight in Spam, Singletary will get 9 carries.
Recommended Posts