Well, we're back to this.

VS.

I thought we were done with this.

I did too, but first a few clarifications.  Help me out here with some emphasis?  First things first:

Split zone does not have a read.

Thank you.  I dorfed my explanation here.  For an explanation of split zone, go here.  Or just read the latest UFR.

I called it a "zone read" play in that it's often packaged with plays like zone read in the RPS context.  (To that end, I said I didn't even really care if it has a read or not.)  I don't know if the MGoStaff have a different definition, but my understanding of RPS are a set of plays that are designed to look similar but attack defenses differently, the intent to punish overplaying a particular play.  What I was going for was like calling the ol' iso/bootleg combination a "pro-style" RPS package, the problem being that zone read is a play in and of itself.  Okay, that's gonna rankle.  Anyway, this is the point, if anyone cared:

To make a base play functional, you need the counter, and the counter to the counter's counter.  The latter keep the defense "honest", keeps them from overplaying [the base].  Well, with split zone, the counter involves the QB.  There's no "scissors", so defenses are starting to do brazenly unsound things and getting away with it.

You don't have to use zone read as a constraint for split zone, but it sure is nice, because when you're running a real zone read, where the QB might just tuck the rock and run, it compels the EMLoS guy to shuffle instead of scream toward the ball, making him easy pickings for the blocker running across the formation.  But Michigan wasn't running anything to prevent the EMLoS guy from overplaying split zone.  Which leads me to this part:

That's not split zone either.  That play is bash.  For the record, I liked the playcall, but I don't necessarily accept it as a constraint to split zone because the book on Michigan isn't, "The ball will never go outside," it's, "The QB will never keep."  The only guy moving across the field is the running back.  Nebraska's Nelson is unblocked; he has one chance to stop Corum and that's to go hell for leather.  Instead he shuffles like he's being read when the QB isn't even looking at him.  His movement makes sense in most other contexts, but his opponents that night were Cade McNamara and Blake Corum.  The ball's only going on one direction and hesitation means death.  If he's not hedging then he's surprised that the ball's going the other way for a change, which, great, but that ain't gonna work twice.

Anyway.  There was one other mistake I made that didn't get called out, but as long as there's egg on my face, might as well plant it in the frying pan and make a nose omelette.  I combined the DSR tally with the +/- chart results to generate the GMSR.  I might as well have compared apples to durian.  Then JHumich helpfully pointed out that a tally of Cade's missed reads were already in the DSR chart.  I'm on a roll, here, folks.

Is there any more to this besides you not knowing what you're doing?

Probably not, but I'll give it another go.  Since McNamara's being charged with dragging down the run game, I compared the split zone + counters RPS "package" with other runs, and passes for good measure, to compare their efficiency.  The first problem I ran into was accounting for the opponent.  Pretty much everything worked against NIU, for example.  I'm sure PFF and such have their ways of dealing with it, but my own un-scientific method was to just divide the yards-per-play by the game's overall average, creating a weird, dimension-less number that nonetheless was more useful in context than the raw numbers:

• "SZ-RPS" refers to Split Zone and its common counters that Michigan isn't executing consistently:  Arc, Belly, and Zone Read.  (I also tried to include RPOs but FWIW there were barely any of them)  There are a few I'm-not-sure ones in there like "Dble Split Zone Flip" but at this point we've established I'm an amateur so this probably won't matter anyway.
• This may look like fun with small sample sizes, but each data point includes a dozen or so run plays, more or less.  Keep that in mind.
• Also keep in mind that this is heavily weighted to the right.  WMU and UW didn't know much about the team and are less relevant now.  Each data point is more important than the one to the left of it.
• Week 2 was UW's NFL-grade pass defense; Week 5 was the other UW's NFL-grade run defense.  That's why they look weird.

The UW outliers aside, we're seeing a pretty clear decline in the efficiency of the Split Zone & Friends package.  Again, we're talking the evolution of constraints & adjustments, not commodity prices, so you need to read this graph giving week 6 far more gravity than the others.  Anyway, by Nebraska week, other runs were outperforming the SZ&F package by almost 2 ypp.

Also FWIW, I tried to break it down by QB, but the data got too noisy.  In any case, it looked like McNamara pulling the ball a few times a game wasn't enough of a deterrent to change things.  So, the real crime here might be the coaches designing the offense around an RPS package that Cade can't execute, but like I said, I wouldn't say that's FFFF's fault.

FFFF isn't really a straight-up evaluation of the players.  It's more like a scouting report. . . If they were running Denard Robinson out there in a West Coast Offense, it'd be the same deal.

I do have some good(?) news, though.

Do tell.

Seth noted in the UFR that Michigan is starting to do more things to punish defenses overplaying Split Zone.  I would like to add to that, they seem to be moving away from it anyway.  The "SZ & Friends" count over the first half of the season has been:  16, 24, 16, 15, 12, 11.  SZ itself has gone from being run 8-10 times a game to a season low of five.

Will there be a Part III?

Hell no.  I went into this with three objectives:  A) make a point about Cade's role in the offense, B) so Seth wouldn't have to, and C) not make a mess in the process.  Instead I've spent the better part of two days stirring up confusion over split zone, increasing work for everyone.  Making mistakes is human, and improving on them is noble, but full-blown negative productivity is a sign I dipped in the wrong waters.  If there's anything I gained from the last few days, it's a greater appreciation for the folks who do this for a living.