Nice data dig, Seth. I guess Dileo as "The Threat" was right all along ... Just by putting him in a simple crossing route or slant the defense has to acknowledge his presence and that opens up the TEs and the run.
Hokepoints Wins a Medal for Participation
Fuller didn't get a shot of Dileo that he put on Flickr, but he got this 6-yard catch by Butt on 2nd and 5, when Dileo was busy running off two defenders.
The primary complaint with Michigan's offense, rightly, has been with the blocking dudes' problems with blocking dudes. While gathering data on personnel changes throughout the Northwestern game I got an opportunity to look hard enough to have an idea where the UFR will lay blame for 9 points in regulation. Preview: Bosch didn't have a good game. However the freshman guards are a problem solved mostly by experience, i.e. we can't fix it this year.
But if Michigan is looking for an offensive boost it might find one by improving which parts they deploy among the five eligible receiver positions. Which personnel and how they're aligned come with various strengths. Generally the smaller and more spread out, the better to make space for you to operate; conversely the larger and tighter the better to block dudes. I put forth that our blocking dudes are currently pretty bad at blocking dudes, thus it's worth moving some of their snaps to 3rd and 4th receivers.
MANBALL isn't Borgesian
Here's Borges's offense being run at UCLA in 1998, a time when the spread offense was something that won games at Tulane:
Note the 3WR sets pop up plenty. I believe the goal here is to be multifarious, not just very large and good at something. He wants to be impossible to prepare for because at any moment you might put in your 4-4 personnel when you see him trotting out 3 tight ends, and then he'll spread them out and put a 6'6 monster on your tiniest cornerback. This is why they're recruiting Fifty Shades of Shea.
But That's a Long Way Away
Today, they have precious few developed parts to play these "skill" positions. The running backs can't block, either because they're really spread nutrinos (Toussaint, Hayes, Norfleet) or true freshmen (Green, Smith) who didn't need blocking lessons to run over high school fools. The fullbacks are a walk-on they've been developing for awhile but who still misses 1 in 5 blocking assignments, and a RS freshman they recruited out of Utah who needs work.
|Off. Performance vs. NW'ern When Player is On Field
(Only normal downs counted)
From a Borgesian perspective, the tight ends are in even worse shape. Funchess became a receiver because despite all that size he's not much of a blocker. That leaves his classmate A.J. Williams at the top of the depth chart despite the fact that he's not been a very good blocker, and his threat as a passing target fizzles out about three yards downfield. They've got Jake Butt, who like Funchess is more of a receiver at this stage in his career. And just so they have another body there, positional vagaband Jordan Paskorz has been getting a few drives here and there; after him it's burning a redshirt and air.
It would make sense, then, for the receivers to pick up the slack. If you can't block a guy with Williams, you can get that same block by putting a receiver far away from the play, so long as you threaten to go out there if a defender doesn't follow. But there's another problem with the receivers: Gallon is great but tiny, Funchess is great but still raw. Chesson is coming along. Dileo is himself.
And…? The coaches seem to have put every other receiver on the shelf: they've played Jeremy Jackson a lot and gotten little returns. Joe Reynolds seems to be not an option. So every time they go 4-wide, effectively the whole depth chart is out there. Exhaust those guys and the passing game goes away. Or at least this is the best reason I can imagine.
I'm not sure it's a good reason. It seems to me that they're pretty effective the more they spread 'em out, because you're essentially replacing a mediocre-to-bad FB or TE with a slot receiver who is pretty good at that job.
Did You See Dileo's Number in that Chart?
I spent much of yesterday and all night last night charting the personnel moves during last Saturday's game to be able to pull those numbers. The whole thing is here:
There's no way I can go back and do the whole season, unless Brian has a secret code hidden in the UFRs or something. Anyway: 9 YPA when Dileo is out there, and 4.5 to 5.5 when he's not. Here's some other things I found in there.
[After the Jump: What We've Learned]
- A few tendencies, like Houma coming in was 100% run (so was tackle-over but you knew that).
- Smith in the backfield on a normal down and distance was 78% run. Hayes was 13% runs. Green was 53%. Hayes also played the entire 2-minute drill. I find this good news: Hayes needed to learn to block to justify the 3rd down back role we envisioned for him, and the freshmen have plenty of time to catch up in that department.
- Jake Butt split time between Y and U, and Michigan was more effective with him at Y. Butt lined up at U-back 20 times. All but one were on normal downs, on which Michigan averaged 4.2 YPA. He had 23 snaps at Y but just 14 were on normal downs—on those Michigan averaged 7.2 YPA to Williams's 5.1 on 38 snaps. Butt was also open on at least two of Gardner's shoulda-been-INTs. At this point I think his blocking is close enough to A.J. for the added pass threat to be important.
- However A.J. did a lot of lining up as a virtual extra tackle when Michigan went heavy to the strong side. Northwestern started blitzing at Williams's side and he handled it poorly.
- Joe Kerridge downs: 19 for 4.7 YPA on normal downs. Non-Kerridge normal downs: 41 for 5.8 YPA. Fullback hatred tags unlocked.
- When Kerridge was in at running back Michigan always passed. I'm totally fine with this so long as it's on obvious passing downs since he's as much a threat out of the backfield as the other RBs and a better blocker.
- That's not saying much; he wasn't a very good blocker either running downhill or at picking up protections.
- When Michigan covered a guy or had an OL shirt on an end, they averaged 1.3 YPA. When they didn't they averaged 4.6 YPA. Tackle over is dead to me.
The Dileo Effect
|"The most effective weapon is the one you never have to use, but you still have to pull it out and wave it at people!" –Masamune|
Here's ten "normal" plays with Dileo on the field:
- 11-yard PA in to Funchess
- A zone read that Gardner biffed the read, ends with 1 yard.
- 13-yard Funchess reception on a smash route
- 6 yards to Butt on a PA curl on 2nd and 5
- 14-yard pass to wide open Chesson
- Incomplete skinny post to Gallon on 2nd and 23
- Incomplete sideline pass on one of those Illinois rollout things.
- Gardner scrambles for 22 yards out of a pistol
- Quick hitch to Gallon for 11 yards
- 12-yard keeper on a zone read in 3rd overtime
Note the lack of any Dileo targets, or anything else that shows up on a scoreboard, but also a distinctly different experience than what you saw for most of the game.
Borges did finally trot him out for the 2-minute drill and a bunch of long conversion attempts, but on most of those Northwestern was undercutting those deep flag routes that used to get Dileo open.
What We're Asking: Be Scott Linehan, Not Rich Rodriguez
Not necessarily going back to the spread. I don't think Borges knows any more than he did the last two years, and they're transitioning away from that anyhow. But I think it's very reasonable to ask that he run his West Coast offense from a spread, using the width of the field and the horizontal and vertical spaces to make room for his young interior OL and developing Shea-bots.
Yeah, there's no "hand" in our future.
But I think it's very reasonable to ask that he start spreading to pass, using the width of the field and the horizontal and vertical spaces to make room for his young interior OL and developing Shea-bots.
This has been my question for most of the year. Even when Borges does bring on multiple receivers, it seems like he keeps them close to the linemen. It feels like all that does is still allow the defense to stack the box, so that if we do run, they are still there for run support. Gardner has shown he has the arm to throw to the sideline from the opposite hash, so why not spread those guys out as much as possible?
I have to admit that I'm just a run of the mill fan. I don't understand most of the X's and O's strategy or scheme and what little I do understand isn't apparent in real time, I see it when watching a replay. Even I know that some sort of magic happens when Dileo is on the field.
Didn't we burn DeMario Jones' and Csont'e York's Redshirts? Why can't they be involved in the WR depth chart a little more?
We are 11 weeks in. You can't tell me they haven't learned enough to run routes and block at least for a second. WR blocking is what turns 5 yard runs into 25 yard runs. But right now, I would settle for the 5 yard run if the WR is taking an extra defender out of the box.
Running those numbers for every play this season would be very time consuming. But I would be interested to see the mix of run/pass when Funchess used to line up as a TE vs. Williams for the first half dozen games. Gotta be 20/80 & 80/20 respectively.
I also wonder about other teams' tendancies as well, if they would tip their hands with personnel on the field decisions. If I were an OC one of my main goals would be to keep all as close to 50/50, or 60/40 worst case, splits just in case other teams hired Seths to do such time-consuming research.
I imagine, all big teams do personnel play selection break downs, at least for big opponents. They certainly do it on themselves. That's what GAs and such are for.
The problem with trying to make it 50/50 is you end up running a lot of suboptimal plays with the wrong personnel. You end up asking Funchess to block too much for instance. Then it becomes a balance of putting your personnel in a position to suceed, by asking them to do what they are good at and not what they are bad at, against fooling the defense.
Of course, very difficult to do with success. Becomes a cat vs mouse game at times.
That's backwards though. I daresay half of Borges' problems is his insistence on running the ball.
What the guy SAYS in pressers makes more sense, that the game determines the play selection. But if that's the case and the defense has 9 in the box, throw 30, 50, 80 times a game if the defense refuses to back off. Punish them for overplaying until they learn, then move to something else the instant they do. It's not about run/pass ratio so much as:
A) Running different plays from the same look so the defense doesn't know what's coming. This is an arbitrary personal preference but with the exception of and-long plays which are obvious passing downs, I want at least four different plays from the same formation & personnel group. A single formation running several plays is way more effective than twice as many plays that each give the defense a pre-snap read. We have way, way too many of the latter.
B) Taking what the defense gives you. I don't care if you're the #1 offense in the nation; if the defense is concentrating on one part of the field, that's the one place the ball shouldn't go. And IF you are the #1 offense, odds are you're not doing something that stupid anyway.
And FWIW while Hoke is responsible for the offense as HC, I don't buy the theory that he's holding Borges back -- at least, it's just speculation to me. Borges runs the offense from a box Hoke isn't even headset to, and the unanimous feedback from former players, current players and coaches is that everybody loves working for the guy. He pulled Mattison from a successful gig in the NFL FFS. The same guy is lording over Borges' playbook with an iron fist? I need evidence.
Bring up these numbers in the next press conference....
"9 YPA when Dileo is out there, and 4.5 to 5.5 when he's not."
There would be no point to bringing this up to Borges. He already knows. Anyone could have predicted this result--although having the hard numbers is nice (good work Seth.) Dileo is generally not out there in the traditional running sets (80/20 pass/run split.) Everyone knows that Michigan gets their yards through the air and goes backwards on the ground. With the exception of Hayes, who probably got most of his snaps on the last drive (sample size blah blah), Dileo is the most pass heavy on that list by a significant margin.
If you did pass splits and showed a significant increase in YPA with Dileo, that would be something different all together.
After the Nebraska debacle, I imagine Michigan will continue to move away from the run heavy gameplans for now. I'd also like to see more spread formations. I just don't think confronting Borges with this particular stat will do anything other than make you look naive.
Uh, increased offensive productivity due to Dileo's presence is racist.
but then there is the Kerridge factor as well...
Um, so, I'm not sure that is something Masamune ever said. In fact, I think you are more likely remembering this scene from iron man.
everywhere it makes it a little easier to understand. Nobody can block! Time to get back to raw fundamentals I guess.
In the meantime, I'm gonna dance to the Michigan VICTORY!
Problem is we have some turkeys that can’t swim and only fly when threatened, but are pretty good at walking. We are now asking the turkeys to swim some and fly once in a while and we have some ducks that can swim and fly pretty well but it’s dicey if they have to walk. So now you are asking this group to swim, fly and walk, either way some of them are not going to perform well because…well they just can’t! I think once we get more ducks and move to the swimming and flying program it will be ok.
Hmmmm...I'm not the sharpest tool in the shed, but is the implication here that Borges is not running Borges's offense? Instead, Hoke is forcing Borges into an offense that he hasn't exactly run, and that if Borges had his choice Michigan would be running something much more wide open?
Recalculating "Fire Borges" meme...
I would think there was something to this except...didn't Borges run a more wide open attack at SDSU under Hoke?
This is what I have believed from Day 1.
Borges made a comment somewhere to the effect of "Sure, I'd like to open it up and fling the ball all around the field. But that's not what we do here. Gotta protect your Defense." Those look like words that came out of Hoke's mouth, not Borges.
The exact quote is somewhere on MGoBlog, about a year ago or so, if someone can find it. Borges sounded a little frustrated in it, but he was being a good troop and spouting the company line.
I have a feeling that if you let Borges do what he wanted in a vacuum, it would look a lot different than what we have been seeing. Firing Borges is not going to change what we have been seeing, becuase he's not the full reason we've been seeing it.
I don't think that's the implication here, but your conclusion is still good. The implication is not that Hoke is forcing Borges into this offense. Rather, I think the personnel and their development is forcing Borges hand and he is running something he isn't quite comfortable with. Whether he should adapt better or not ultimately will fuel whether people think he should go. Nonetheless, I think the question also has to be what your OC hiring policy is. Are you hiring them knowing they will have middling success with the wrong personnel during a transition but that when they can fully impliment their system with their personnel they will be very successful; or, are you hiring them expecting them to be able to be like a swiss army knife and be able to have good success regardless of the system or personnel? While I would love an OC would is able to adapt to all personnel and schemes, this is likely a very hard to find, especially given that they have to fit the demands/personality of the head coach. If, on the other hand, that OC is skilled enought that you will be very successful once he gets all his pieces into place, then you can still have a great team at the cost of some middling performances until then. I think the football programs probably plan for the long-term and ultimately, I favor this approach as well.
I've been asking around a lot the last few weeks to see if anyone can figure out whether Borges is trying to embiggerate this offense all on his own or if he feels he has a mandate from Brady to do that.
The consensus so far is it's just Borges.
Probably the best way to think of it is that Hoke is trying to maximize his chances of winning. He is not concerned with YPA, YPC, etc. unless it creates a win.
You can see this in terms of how they have played different teams this year. Against Penn State, it seemed that they calculated that minimizing offensive mistakes would give them the best chance to win, so they pulled back the reins on the offense. Against Indiana, it seemed that they calculated that they were going to need to score a lot of points, so they opened up the throttle.
They also are probably basing these decisions on personnel, and right now they seem to figure that a more reserved offense is going to give them the best chance to win, especially when they have a pretty decent defense.
I also think that the overall goal for Hoke is a strong defense, and a low-risk offense. It may not win the "shiniest" award, but it will get you wins consistently.
assured me that Joe Kerridge was a very good fullback. So what the hell happened? Now he can't block?
Please more info on Houma vs. Kerridge vs. no fullback at all. That would be interesting.
Nice find although limited sample size, but I could see how it has some truth to it.
What I really liked from this game was how many times I feel like they threw out of the I-formation. Someone had posted last week that I-formation naturally was very high running % so I thought that was good to mix up. Oh and bubble screens, obviously. I don't think there's a play I associated more with Michigan as a kid than the bubble screen.
So, if we take as an assumption - which of course we can't take because we don't know, and won't ever know - that Hoke is forcing Borges to run some offensive concepts he doesn't like because RAWR MANBAWL and RAWR TOUGHNESS, then Michigan is screwed, at least for a while.
Hoke won't fire Borges, because Borges is doing what he's told. This makes some sense (note: I am saying SOME) if it's taken in context with Hoke telling Borges to try to score more to keep up with Indiana rather than RAWR PHYSICALITY. We also surmise that Brandon won't fire Hoke until at least after 2015. Thus, buckle up for a rough to very rough 2014, and maybe 2015 and beyond, folks.
because none of the young talent on either side of the ball will improve. I see your Ugh and raise a GTFO.
Or, get ready for the growth of the OL, the young big backs being able to pound the ball up the middle with increasing success, Devin Gardner getting more protection so that those PA passes that everyone hates actually work as designed, and the big young WR's start to stretch the field vertically.
I think Hoke's insistance on Manball is more a ploy to get the OL going than anything else. I think he is well aware of the fact that spread concepts have been incorporated very successfully into the power offense, but without the OL, nothing works. Once the OL pipeline is totally flowing, I think we'll see a much more varied offense that is able to switch from one "scheme" to the next when need, but the staple will ALWAYS be pounding the ball. If you pound the ball successfully, play great defense, you're not going to lose very many games.
Hoke won't fire Borges, because Borges is doing what he's told
This is questionable logic. All assistants do what they're told. It's not like insubordination is rampant in the college football world.
If Hoke does not feel that his assistants are doing their jobs well, then he may let them go. Lloyd went through five OCs, not because they rebelled and wouldn't execute his ball-control offense, but because he wasn't satisfied with the results.
because they share a common understanding about the difficulty of the task and the timeline to accomplish it. You may disagree with that understanding and think it is wrong, but as long as they share it Borges isn't going anywhere.
This is not among the "Hoke won't fire his buddies" or "Hoke is to loyal to his staff" or "Hoke is out of touch" memes that sprout all over this and other sites. It is rather a type of contract. Borges will attempt to install the offense that Hoke wants with the understanding that Hoke will give him the time and cover to have a reasonable shot at success.
You can also be sure that they saw the possibility of a season like this at this point, before Al took the job. They also knew it might get this bad before they decided, apparently at the end of last season, that 2013 was the season to bite the bullet for installation of this offense. The decision to change philosophy this season was certainly jointly made by both men.
So Borges has at least 2014 to improve the offense's perfromance (note I did not say have a better record) and barring a complete melt down probably through 2015.
When Michigan covered a guy or had an OL shirt on an end, they averaged 1.3 YPA. When they didn't they averaged 4.6 YPA.
Is there a typo here? Our gamelong YPA was 5.3.
G/T/TE--WR and G/T/T alignments. Tackle over and covered otherwise-eligible ends. Not included would be G/T/TE or G/T--WR formations.
It seems like you're looking at overall offensive production, i.e. both run and pass plays. Wouldn't the conventional term for that be yards per play (YPP) not yards per attempt (YPA)?
Great work, Seth.
Using our Dileo tendencies to our advantage seems like a good idea. If we're passing 80% of the time he's out there and we're STILL successful with it, it seems like his presence could open some options for us. And--without looking at the full season of data--my guess is that about 90% of the runs with Dileo on the field are inverted veer/option looks.
I would like to see some single back, three wide sets. We had some good success with the I-form twins vs. Northwestern; spreading the field a bit more might be advantageous.
Everyone knows Dileo is a security blanket; Northwestern almost got a couple of picks out of that knowledge. It's time to use our tendencies to our advantage.
I'm with you in "we should be spreading the field out because we can't block" department. My initial thought was wondering whether this trend held up all season, but you mentioned that'd be too time consuming. I was thinking of this as an alternative...
Could you look at YPP, YPA, YPC, and Run% for 2 wides, 3 wides, and 4 wides for the season? I'm thinking you might be able to pull that from the UFR data. I know someone looked at YPP from all the different formations, but in a sense I think that's a bit under-aggregated. This might also help isolate the causal variable as Dileo vs the formation.
Also, have you thought about changing your name to Misopogon in real life? That would be awesome.
Yeah. That's a future thing.
I hope you're referring to changing your name. But either way I think the results should prove interesting.
I like what you've tried to summarize here. It would be interesting to know what the YPC and YPA are with Delio on the field versus Williams/Kerrige/etc. I think the data will be skewed for YPP in general since we pass more with him out there and generally suck at running. But do we actually get more yards per carry with him on the field? Now THAT would be interesting to ask of Borges in a press conference.
- Normal downs: 10 plays (8 passes 2 runs) for 9.0 YPA
- 2-minute drill: 12 plays (all passes), 3.8 YPA
- Short conversion: zero plays
- Long conversion: 5 plays (4 passes), -2.6 YPA
- Normal downs: 50 plays, (23 passes), 4.7 YPA
- 2-minute drill: 1 play (pass), 6.0 YPA
- Short conversion: 7 plays (1 pass), 1 YPA
- Long conversion: 4 plays (all passes), 1.8 YPA
Perhaps what worries me most about this analysis is the near future. I'm sure the coaches hoped Funchess and Williams would be better at blocking by now, but they aren't. They are both pretty terrible. It seems like the best case scenario for next year is minimal improvement. Butt is the most likely to show significant improvement as he gains size and experience, but he will always be a receiver first. Compounding matters, Delio graduates and for some unknown reason his logical back-up (Norfleet) has yet to see the field. On top of that, we are redshirting an H-back in Khalid Hill that is an unknown commodity but not really an in-line blocking TE. We also have an incoming freshman in Ian Bunting who is a converted and skinny WR coming off an injury plagued (and unproductive) season. We seem light years away from the day when we can be like Stanford and trot out three TE and destroy people even when they know what's coming. Our only hope seems to be that Borges/Hoke realize this and we see a healthy dose of Norfleet/Jones/Canteen in the slot next year along with Funchess/Chesson/Darboh on the outside.
I was looking for something along these lines. Less Argh, Borges an idiot and more here is something we could do that is not a huge change but would work marginally better and that could add up to an offense that moves the ball more consistently.
I agree that this spread them out more as a way to back off the pressure could work. I remember when on one of the Steelers Super Bowl runs they couldn't run the ball until they had success with passing early on. Against both the Colts and Broncos they came out passing more agressively against stacked fronts and then were able to go to more ball control offense with a lead and the defense backed off. The Steelers have never had a great offensive line but try to find ways to make it work and hope the line can gel at some point. Obviously, Gardner isn't neccesarily Ben Roethlisberger but since I am a fan of both teams the similarities struck me.
Gallon and Dileo gone.
Amara Darboh back.
I also think RB ustice Hayes could be a nice receiver option for Michigan, if needed.