This is correct. We're talking about small sample sizes, so the buckets you pick can have a large effect on how the numbers look. I just wanted to show that in Graham's case it doesn't really matter how you slice them, he was consistently productive in every role.

Indeed: http://sports.yahoo.com/nfl/news?slug=ms-trippintuesday033010

You're conflating two different types of reads here. Vick's blitz recognition is abysmal, he can't see dropping linebackers, etc., etc. That's all taken as read. But when you look at what Kelly has said about his run game, it's not that complicated. He even says "it's not a read." Just hand the ball off, every time, unless the DE (or scraping LB) gives you absolutely no choice. As a final point, we should be explicit about the ground we're fighting on here. You're saying "replace Vick immediately" and I'm saying "replace Vick as soon as you can possibly get someone better." By the time the draft is done, there may no longer be any space between our positions.

Dixon is here, so that's covered. If he's better than Vick, he'll play. As for Smith, it's the same thing as above -- he's better, but he's not free. San Francisco isn't just going to give him away. Anything we trade in return takes away from resources we could use to address other needs. Why do that for a stopgap when you have a stopgap already? As for "teaching other players in the offense," that's not his role. Again, you're tied up in looking at this from Vick's perspective. Break away from that for a minute. Foles *can't* run the read option. He's too slow. Defenses will just crash on the back and then if he keeps it -- which he should, every time, if he's reading correctly -- they'll just run him down and demolish him before he even gets to the line. Vick may or may not be able to run the read option *well*, but he *can* run it. You have to play him straight or else he will get the corner on you. Will he run 60 yards and score a touchdown? No, not anymore. But he at least gives: * Kelly the chance to see how his concepts work at this level. * Everyone else on the offense a chance to practice the things they'll be doing going forward with the new QB. I'm arguing, in effect, that this one qualitative difference mitigates whatever other quantitative differences may or may not exist between the two players.

And the point argued above is that all of these things can be true, but it's still better to run with Vick, because the unit of analysis is the *other* 10 guys, not the quarterback. Will *their* performance running Chip Kelly's offense in 2014-? with Future Unnamed QB be better or worse if they spend 2013 playing with: 1) A better QB (for same of argument) who can't actually run Kelly's offense. 2) A lesser QB (for sake of argument) who can run the offense and therefore give people live reps. Also, let's not assume Vick is the final answer. We still have a draft and free agency to go. Finally, when talking about cost, even just cutting Vick was going to be expensive against this year's cap: http://web.archive.org/web/20130203011502/http://eaglescap.com/Players/MichaelVick.html If the reports of the new deal are to be believed, he's basically costing us the same ~$7M cap hit this year but at least we get to keep him in the mix -- and he's now tradable. There were a few different pieces of this decision that need to be considered together.

Kelce and Peters played 1.5 and 0 games respectively. They didn't help much.

And Peters would easily best Thomas. I'm just talking about guys we've seen on the field recently.

In response to the last few comments -- I think our defense was like a 100-year-old piece of machinery that creaked and rattled and shook but basically stuck together and did the job if you talked to it softly and maniacally patched every issue that cropped up. So to say Juan borrowed pieces and cobbled them together, yeah, he did. This was a frankenstein defense from the get-go -- on purpose once Reid decided to build everything around Washburn's fronts. That was the problem with Castillo. He was always one step behind, doing things that worked but then being slow to adjust when opponents would attack him. I think Bowles came down from the server room and expected things to just work when you plugged them in. Not so much. The reason I haven't done breakdowns is because it's too annoying to watch, but also because it would just be an exercise in passing out blame. Look, here's who ****ed up on this play. Now it's this guy on this play. Who wants to read a detailed breakdown of, say, every Ryan Howard strikeout when he's in the midst of one of his slumps?

We're also not talking about some marginal stuff happening on the edges. This has been a collapse of literally historic proportions.

Neither guy really made the system. Andy dictated it, hired Washburn to install and run it on the front end, then tasked Juan/Todd with figuring out how to make everything else fit. Would Bowles do better with a full offseason, his own system from the ground up, etc.? Sure. Did Castillo have any of those advantages? Nope.

Juan was in the same position. His results were better.

Yes, I think that was exactly what was happening. Remember we rarely blitzed, so they should have been able to count to 4+7=11 most of the time. The system didn't work when Castillo wanted to bring pressure, which I believe is what we saw in 4Q vs. Detroit.

It was a well-executed play, but they tried it again later with the CBs on different sides and Flacco didn't even bother trying to throw it. DRC was all over it.

There's also not a huge difference between sitting in a deep half vs. doubling CJ. In fact, the latter is probably easier in some ways, because you don't have to figure out what your responsibility is. Just stay around the really big dude as Nnamdi handles him short.

Depends who you blame for Jarrett.

I think the charitable explanation is that he didn't think Colt could execute the game plan.

No receiver in the history of flag football, at any level, has ever believed he was covered on the last play.

In Kurt's defense, what they're asking him to do is really, really hard. He makes mistakes, but I have to say, given degree of role difficulty, I'm not convinced he isn't playing a little better than Allen. "Kurt, here's your job. If it's a run, come screaming towards the line of scrimmage as quickly as possible to fill this open gap. If it's a pass, run screaming away from the line as quickly as possible to stop the bomb." "Play action? Just don't ever be wrong and you'll be fine."

He's not Allen, but yes, there were things we could have done with him that would have been more useful than marooning him 20 yards off the ball. Consider this: When we started blitzing a safety, we used Coleman. Why not blitz the guy who's lost in coverage and keep Coleman back?

In response to the two previous comments, I thought about this a lot earlier this season. Clearly AR/MM have made a ton of adjustments to the offense this year. It just took a couple games. So I don't think we can say they won't adjust. Perhaps they just need a bit more results-driven persuasion to do it :) Anyway, I think the difference between coaching offense and coaching defense is that in the former case the playcaller has an illusion of control. They start each play, while the other team is forced to react to what they do. It seems like they should therefore be able to do whatever they want, as long as their guys execute. Defensive coaches don't have that luxury. You can't say, "Well, look, we'll just firing and sooner or later we'll stop the big play." You have to adjust, right away, as the head coach is screaming at you to "COVER THAT $#@!ING GUY!" I do wonder about the placement of coordinators. If Marty were in the booth and had a first-hand view of what's happening in the B gap, does that change how he does things, as opposed to hearing from the offensive line coach, "Grrrrrrrrowl we'll get that taken care of."

That's the essence of the Castillo discussion for me. It's not a question of if he can implement a scheme, if he's given talent and capable assistants. Clearly he can do that. What we don't see, however, is adjustments on the fly. If whatever we're trying to do isn't working -- or if it *is* working but the situation has changed and now we need to try to force something -- our adjustments are either slow or non-existent. AND THIS MAKES SENSE. Castillo has been a defensive coach for 1.25 years. If he wasn't behind the curve, he'd be a bloody genius, on the order of Bill Belichick, and if that were the case, he'd have been given a chance something like a decade ago. This is what I find frustrating about all this. If the defense plays well for a couple games, people (sportswriters) rush to praise him, because "clearly they're not coaching themselves." Or they look at top-level stats and say, you know, he can't be that terrible. These statements are true. It's just that they have approximately 0% relevance to the issue of what we'll do in January when Mike McCarthy or Jim Harbaugh is on the other sideline doing something we're not ready for and we can't plug the leak until it's too late. On the zone question, where things break down is generally with Boykin and Kendricks, with Ryans alternating between looking really solid and oops, he blew that one, so yeah, I'd say it's mostly the young guys. Everyone on the interior seems more comfortable in man. DRC and Nnamdi look fine either way.

Very. He has no margin for error anymore. If he does everything right, the coverage is terrific. But any mistake and he doesn't have the recovery speed to catch back up.

DTs did not have a strong game at all. But you coach what's actually happening, not what you wish would, you know? Also, a lot of these runs were cutbacks. Harder for linemen to go one way then the other.

Thanks, poetx. Are you planning any more write-ups yourself?

+1