mgoblog - i'm an actor, not a reactor (mgoblog_syn) wrote, @ 2018-05-25 16:00:00 |
|
|||
http://feedproxy.google.com/~r/mgoblog/~3/r
[Paul Sherman]
Earlier this week Ace ran into an article on the Saquon Barkley pick and why, despite sabr-conventional wisdom, it might not have been such a bad idea after all. The article is Michigan-relevant for two reasons. One because he brings up the play where Barkley got manned up as a slot receiver on McCray and smoked him for the 4th quarter touchdown that officially made it a rout. Since that’s already seared into your memory and most everyone involved is now well out of Michigan’s sphere you don’t have to relive that part.
The second reason is because Michigan is stockpiling running backs again. At first blush you might dismiss that as an emphasis on running the football, but…
In 2017, according to Sharp’s data, the Patriots used “11” personnel on just 44 percent of their plays—tied for the fourth-fewest total in the league. New England’s second most-common personnel grouping? “21,” or two running backs, one tight end, and two wideouts, which the Pats used on 24 percent of their plays, second only to the 49ers’ 28 percent. Per Sharp, the Pats’ success rate out of “11” personnel was 47 percent. Out of “21” personnel, it was 60 percent.
…it also might be a sign that Harbaugh is staying at the head of the curve in the latest countermove of offensive progression.
The data say yes. Michigan returned five running backs this season and brought in three freshman. Next year they graduate one (Higdon), and are still in full pursuit of multiple targets, and not in the “we just need one of you” kind of way.
Nine running backs is over 10 percent of your scholarships. That is indeed a major investment. And when you look over Michigan’s history such an emphasis is indeed out of whack with the needs of Harbaugh’s modern predecessors. I was on the Daily when they peaked at seven in 2000 because they didn’t think they’d get both Chris Perry and Reggie Benton (and Carr held up his promise to Tim Bracken). And then it just turned out they had known, though didn’t say, that Justin Fargas and Ryan Beard weren’t going to be around for 2001; proving the anomaly, in two seasons they were down to just four.
The other example is Rich Rod’s first two years, though that includes Kevin Grady who moved to fullback. Still: Brandon Minor, Carlos Brown, Mike Cox, Teric Jones, Michael Shaw, Vincent Smith, and Fitz Toussaint were all on a roster together. We’ll come back to that one.
[After THE JUMP: Why all the backs?]
When talking about recruiting we often mention the importance of recruiting four times as many players as you have snaps for at quarterback and offensive line, and twice as many receivers. These are positions that require a high level of development and therefore projection, so you have to collect assets knowing there’s a strong likelihood that many of them won’t progress to the level you need.
Running back is not on that list. There’s absolutely progression, especially in pass-blocking and route-running, but we’ve all watched a lot of running backs for a lot of teams and rarely have any of them substantially changed their production over the course of their careers unless there’s some exigent circumstance. Chris Perry is the lone exception; by his own admission that had more to do with his level of focus as a junior than his ability (also by his own admission he doesn’t read MGoBlog).
Anyway no, if you’re grabbing lots of running backs you probably have a plan to use them.
Rather than use the “Don’t leave McCray on an island with Saquon Barkley” as our example let’s go back to the RPO I drew up last week that Michigan used to get consistent yards on Ohio State:
Michigan ran it thrice in The Game, the first and third times with “21” personnel, which means two running backs and a tight end (11 personnel is one RB and one TE, 22 is two RBs and two TEs, etc., with the remainder receivers). The sandwich attempt is above, with McDoom line up as the flanker (a receiver on the side of an inline TE) and going in a jet motion, essentially doing the same thing that Evans did.
Jet motion has been all the rage for the last five years or so as part of the natural dance between offenses and defenses. The spread put offenses in mostly 11 personnel, with three wide receivers spreading the defense horizontally. So defenses, freed from worries about fullbacks, started using hybrid space players instead of their regular linebackers to cover the extra space and forced running plays back to the meat of the line. So offenses used jet motion to mix up those matchups and threaten horizontally again. And defenses practiced various reactions to jet motion and coached their players to stay responsible.
So what’s next? Why it’s the same play just instead of all that jet motioning of a receiver who’s basically a running back, just start with another running back:
You see, the thing about putting a hybrid safety in at one of your outside linebacker spots is the other outside linebacker isn’t one (neither is your hybrid space player a thick, block-destroying linebacker). One of the goals of motion is to not give the defense the matchups they desire, and one of the most common matchup games offenses like to play is flipping the roles of the outside linebackers.
They’ll use all sorts of tricks for this. Here’s Maryland lining up in an unbalanced formation (the right tackle is technically a tight end and the receiver on the 30 yard line is technically the left tackle), then running jet motion with their slot receiver, then the tight end comes across the formation too and after all that it’s just regular old split-zone. But look who ends up LB in the gap they’re running it to: 5’9” slot safety D’Cota Dixon. Who’s blocking him? The left guard.
But there’s an easier way to threaten both sides of a formation: backs. Flank your quarterback with two running backs and it accomplishes the same threat to the field side. It also messes with defenses who set the strength (you know, strongside/weakside) by which side of the quarterback the running back lines up.
Depending on their skills, you can get a lot more than that out of your backs. The defense doesn’t want to waste a coverage guy on some dude who’s most likely running behind a wall of meat. So quite often they’ll match your back with a linebacker they wouldn’t trust on a receiver. And if your running back has the ability of a receiver, that’s a mismatch. If you have two running backs on the field and only one of the defense’s slot defenders is able to cover one, you’re guaranteed a mismatch!
That second back has to earn that matchup, however, by being good enough at other things to make up for whoever’s coming off the field for him. Chris Evans can eat into Eddie McDoom’s snaps when Higdon is on the field because McDoom hasn’t been much more than a jet pony, a trick that justifies a handful of plays per game but not more. But backs need rotation. If you have a guy with Hidgon’s skills to sub for Higdon, Evans’s passing threat can remain on the field. If Walker can be a strong lead blocker, he can siphon away fullback snaps (something we could use this year since Ben Mason’s the only one who’s seen the field).
Another trick a lot of defenses like to use these days against spread to run teams is to have an end (or a 3-4 OLB if you’re Wisconsin) take the RB out of the backfield—hey, the offense is probably optioning him anyway, right, so might as well just make that an official man-for-man. That has its downside:
This is not new stuff. It’s ooooooold stuff. Like the original stuff—those “flankers” started out as wingbacks after all. But you don’t have to go back to the 1930s to mess with defenses with a pair of backs; the ’80s will do just fine:
Somewhere Bill Walsh was smiling when Harbaugh brought this play back against Ohio State. Pure West Coast Offense run game. 19 Handoff Crack out of Pro Split Backs. Haven't seen this in a long time. pic.twitter.com/0BOIVXBV7e
— James Light (@JamesALight) May 11, 2018
Yeah that’s a fullback (yeah I’m going to draw this one up this offseason). Fullbacks are backs too. That’s the point: these guys often have skills besides runner that are on the field and not being put to use. Chris Evans is just as dangerous as a slot receiver when you get him the ball in space. Khalid Hill was a good receiver. Ohio State used Ezekiel Elliott’s great blocking to devastating effect.
data from UFRs so it’s missing some OSU and bowl games
Harbaughffense is predicated on two major principles: moving defenders where they don’t want to go, and hybrid players who create mismatches just by being on the field. By now I’m sure you’re familiar with the tight end blocky/catchy spectrum, e.g. Zach Gentry can block down on a lineman but he’s too fast to cover with a linebacker and too big to put a defensive back on him (Penn State covered him with a 6’5”/220-pound linebacker in case you’re wondering if defenses aren’t adjusting).
That extends to the backfield, although to date this has mostly been with fullbacks of all varieties. Still the chart above isn’t just an I-form and plow offense. Even including passing downs, Harbaugh’s had a second back or fullback in the formation as often as not, and he’s having them do a lot more in those formations than run or pretend to run and pass protect.
The tight end passing game is stark, and that’s not even accounting for the fact that Funchess is in the Hoke TE numbers for 2012 and 2013. It does however have to take into account that Hoke left his successor just two viable receivers, and that Mackey winner Jake Butt was around for 2015 and 2016. But those TE numbers are in line with Harbaugh offenses at Stanford and San Francisco. And though they’ve recruited a lot of receivers too it’s not surprising that Harbaugh’s tight end recruits have seen the field earlier and had more success. This isn’t his first McKeon.
He also passes five times as often to his fullbacks. But the RB targets are only slightly up from the Hoke era, and still not close to the usage Rodriguez got out of his backs in the passing game.
Michigan in 2016 was 39th in RB targets, and fell back to 84th when everything broke last year according to The Mathlete’s numbers. Also last year there was a rather sharp disparity based on who was under center.
Target | Speight | O'Korn | Peters |
---|---|---|---|
Running Backs | 4% | 10% | 16% |
Fullbacks | 1% | 6% | 6% |
Scrambles | 4% | 10% | 9% |
Wide Receivers | 74% | 50% | 38% |
Tight Ends | 18% | 25% | 32% |
The Peters numbers have a lot to do with whom he was charted against: a bunch of pushover defenses when Michigan was running on 70% of standard downs, and Wisconsin, who was mugging the receivers all day. No need to do more than dump off to an open tight end when passing. O’Korn’s RB targets would have gone up if the Ohio State game counted.
The Speight thing sticks out though. Michigan last year apparently went into the season expecting to operate a spread to pass offense like they used against Florida State in the bowl game. But come Purdue their pass protection problems necessitated more blockers, and the 3- and 4-receiver sets disappeared.
Prior to the season one of our big predictions, based on Harbaugh history, was that Michigan planned to run out their heavy personnel and multiple backs, then have a lot of those tight ends and tailbacks line up as receivers. The TRAIN formation embodied that philosophy, but it doesn’t end there.
Harbaugh loves his tight ends; Rich Rod didn’t know what to do with them. In Rodriguez’s prime years at West Virginia tight ends were at a premium in college football, and fourth stars went almost exclusively to big backs. Sub-six-foot receivers and scatbacks—even those who loved to block—were virtually ignored, a market inefficiency he could exploit.
So Rodriguez built his base offenses with two backs. One might be more of a blocker—walk-on fullbacks could do the job just fine—while the other was a dynamic scat-type who could threaten the side opposite his slot-smurfs. They all had to catch the ball. And they did so a TON:
Position | 2005 | 2006 | 2007 |
---|---|---|---|
Running Backs | 22% | 16% | 20% |
Fullbacks | 5% | 6% | 6% |
Slot Receivers | 28% | 31% | 47% |
Wide Receivers | 40% | 47% | 27% |
Tight Ends | 5% | 0% | 1% |
Rich Rod even had a position—“Superback”—separate from running back because having two in his base offense was standard. He was going that route at Michigan. Brandon Minor and Carlos Brown regularly lined up in the backfield at the same time. Michael Shaw and Sam McGuffie rotated in liberally. Mike Cox and Stephen Hopkins were the heirs apparent to Minor, while Vincent Smith and Teric Jones, and Fitz Toussaint were recruited for the Steve Slaton/Noel Devine role. All of them had different skill sets but those skills were all multiple.
Harbaugh’s still going to have more tight ends on the field than most, and use a fullback, and he didn’t recruit the greatest receiver quartet in program history to leave those guys on the bench. So where is he going to find snaps for these backs? It’s really up to the backs and how they develop. One of them will be on the field all the times, and at least four are guaranteed to get carries. If Christian Turner’s “soft hands” make him as dangerous as Vincent Smith was on screens, if Hassan Haskins can passably lead block from split backs, if Michael Barrett can develop that wiggle into lethal wheel routes, that’s good reason to put a second back on the field instead of a fullback, a second tight end, or a third receiver.
They want some guys who can threaten defenses in multiple ways and create mismatches in the downfield passing game, short passing game, blocking, and of course running the dang ball. They want to have a lot of guys involved with the offense to make them a tough scout, and force defensive coaches to cover a hundred contingencies in their short prep time, or else let Michigan dictate matchups that go exactly like Barkley-vs-any-linebacker will go.
With last year’s pass protection problems and quarterbacks who rarely made it to—let alone knew to look for—their 4th reads, Michigan more often than not had to leave Higdon and Evans in to help protect. That did not go well: they combined to allow six pressures against Michigan State alone and another four vs Wisconsin. The hope going forward is that the added RPO element and better protection from the line will open up the running back position to be more involved in the passing game. As we saw plenty in the Ohio State game, when given multiple backs with multiple skills to play with, they can really screw with a defense.