I thought that myself when I read that article that talked about a Data Scientist(tm)
META: Okay hivemind, I think I'm gonna break the Tuesday column formerly known as Hokepoints (and Museday) into different columns, alternating between stat stuff and Xs and Os stuff. Gimme name ideas. Jimmystats and Jimmypens? I'm at a loss.
Lynn Sladky, AP via Freep.
In his profile in heroism on Harbaugh yesterday, Brian mentioned that THIS pro-style offense at least sometimes does things other than what the defense wants/expects/prepared for:
Those 5.2s [rushing YPC in 2009, 2010 and 2011 at Stanfard] are crazy given the context—on par with Rodriguez's Denard-era run games minus, you know, Denard. This is not a scheme that's just "run it until you stop it"—Harbaugh is trying to screw with your run fits every play.
I thought I'd get into that screwing just a bit because one thing we haven't seen much of at Michigan is someone who knows how to run a power offense correctly. We've seen DeBord and Borges run it poorly, and we've seen Nussmeier try to mold it onto a horizontal spread while still fulfilling Hoke's mandate that at least one tight end must be doing something he's bad at every play.
For something approximating what I expect Michigan will run I went back to the last game Harbaugh coached in college, the 2011 Orange Bowl. Stanford faced Bud Foster's quarters defense, which is helpful since VT's scheme is from the same tree that Michigan State and Ohio State now run.
This was an evisceration. Stanford called 27 running plays and got 300 yards (8.6 YPC) from them. You can remove garbage time (optional since Harbaugh was still running his offense full-bore at 34-12) and it's still 182 yards on 22 carries, for 8.3 YPC. A lot of those were deep gashes—60 yards, 26 yards, 56 yards—which is what you'd expect against a defense that usually gives its safeties gap assignments. I'll show you how the first of those gashes was set up.
Step 1: Scissors.
The first play from scrimmage Stanford came out in a standard I-formation: inline fullback, tight end off the line. Anyone who's scouted Stanford would guess they're going to run their bread 'n butter play: Power-O. We've been over that one before: the backside guard pulls, everyone else has to pin defenders in their spots, and then that pulling guard and the fullback and the RB all come downhill at the MLB, and the resulting yardage is determined by the resulting collision.
Harbaugh showed it without running it with a clever counter that sold the defense on Power-O then had the FB reverse direction and head into the flat, where the rolling out Andrew Luck had essentially a vertical option play on the isolated DB (I labeled him the BCB but I think I got him and the FS confused on the diagram):
He ran a West Coast play from the shotgun on the ensuing 1st down, and then on 2nd and 4 tried to put a power run on the backside with a pass look:
This was blown up by the MLB shooting the gap abandoned by the pulling right guard. A third down pass attempt was blown up when a blitzer wasn't picked up.
[After the jump: how to make the defense eat rock.]
Nagelvoort rides to the rescue
Due to some recording snafus I ended up catching only the final two periods of Friday's game and the third period Saturday, along with the overtimes. Also, the feed FCS picked up looked like an internet stream and it was really hard to figure out who anyone other than Kevin Lohan was even though the announcers tried their damndest to keep us informed. (Seriously, they were great.) I didn't actually see any goals until the Motte winner on Saturday, though I saw replays of some of them. Not enough to write a column, but here are various bullets:
That was probably a good UNH team. The Wildcats were 20-12-7 last year, 13-8-6 in Hockey East, reaching the second round of the NCAA tournament. They lost a couple of their better forwards but returned the vast bulk of their scoring—10 of their top 12—and both goalies. They opened with a solid win over Clarkson in the Icebreaker and then lost 3-2 to Minnesota. By the end of the year that's going to be another quality scalp for Pairwise purposes. Michigan's done a lot of work in just two weeks here.
So far so good for Nagelvoort. Man, when Racine went down with what was obviously a groin issue that I'd be lasts a month or maybe longer (he's definitely out this weekend, and not practicing), dark thoughts flitted through my head. Nagelvoort comes out, my former goalie buddy remarks on how enormous his pads are, and he proceeds to shut UNH out through a rampant third period in which they outshoot Michigan 14-2, with one of those stops an impressive recovery on a penalty shot.
The next night he holds UNH to two goals through an entire game and overtime. Four games in Michigan's save percentage is .937 as a team and Nagelvoort is at .949. Massive sample size disclaimers are of course warranted. It's still the best possible start you could have hoped for minus the Racine injury. Hopefully it keeps up.
Power play: extant. Michigan's 6 of their first 16, a 38% strike rate, and that feels like a sustainable thing since Michigan's been going up against good teams and has been setting up in the zone for extended periods of time. The puck movement is night and day from last year, when their single idea was "get the puck to Trouba." It's too early for me to tell you much else—I get my mind around hockey things slowly.
Recovery. Michigan scrambled their lines for the first time this year after they got pinned in their zone for disturbingly long stretches of the third period on Friday night. They ended up getting outshot nearly 2 to 1 and that was a fair reflection of the play on the ice, if aided by buckets of penalties—UNH had eight power plays. The next night the script flipped and Michigan was better in the last 25 minutes.
Buddies. Michigan's line scramble affected almost everyone but did leave two forward pairs joined: Copp/DeBlois and Motte/Compher. I expect those pairings are untouchable with the success the former has had since its formation at midseason last year—Copp also leads the team in points with 6—and the success the latter's had since their NTDP days. Motte and Compher have already connected on a number of plays that show great understanding of each other and seem like they're more than the sum of their parts when they're on the ice together.
The defense is about what we expected. Bennett is far more aggressive with his puck rushes, Clare's slow speed of thought on the ice gets Michigan trapped in their own zone too often, and Serville continues to make scary mistakes. The freshmen have been a pleasant surprise, especially Lohan, who I figured would mostly ride the bench but has been in the way of a lot of scoring plays. Judgments here are still extremely tentative—ask me again after the upcoming four-game homestand.
Michigan's going to need to get some more playmaking from these guys. Successful passes to set up rushes have been lacking. Four games in the defensemen have four points between them, all of them assists, three of them Clare's.
Nieves stands out. Nieves had the proverbial jump over the weekend; on Friday his line was the primary one generating chances in the final two periods. The shuffle put him with Guptill and Hyman and while they didn't score the line got Guptill seven shots. That is a good guy to get shots; Nieves seems to be emerging. Di Giuseppe, too, seems to be more active this year.
I'm feeling quite a bit better about Michigan's DL performance now that I'm actually going over the tape. They're not doing much more than it seems like they did live, but since no actual NT types are getting much time and a lot of the problems rest squarely on the shoulders of things like "let's see if Mario Ojemudia is a 3-4 DE" and "let's see if Frank Clark is a three-tech." They turn out not to be.
Hopefully we can file this under experimentation and things won't be so bad when the big boys are actually in there. If Michigan goes long stretches without Washington, Campbell, or Pipkins on the field against Notre Dame I'll be surprised. And possibly catatonic.
Not everything can be waved away by calling it mad experimentation, unfortunately. Michigan's linebackers, be they beardy veterans or baby-fresh newcomers, are not making plays. One particular example leapt out because I'd just seen the UMass LB read Michigan's sprint counter, shoot past a blocker, and fill.
Thing I'm talking about == watch Mealer and the MLB
UMass gave up seven yards because all their guys ended up downfield but that's not on the LB.
On UMass's next drive they'd run a play that's very close to that sprint counter. It's just the plain ol' counter, but it's got a pulling tackle that leaves for the wide side of the field on the snap, a linebacker who could be but must not be looking at that, and positive yardage for a team that has struggled to find any.
Late first quarter, second and ten, UMass comes out with trips and a TE to the boundary (short side). Michigan is in the nickel look they spent almost the whole day in. Your DTs are Brink and Black; your ILBs are Bolden and Ross. Ryan is the DE who gets run at.
The tackle at the top of the field pulls.
My great and powerful desire in the above frame is for a Michigan linebacker to read that pull, bug out for the frontside, and hit whatever hole the tackle shows up in. I've been thinking of Notre Dame's linebackers this week since Notre Dame is the next team on the schedule, and they do this. If you zone your line one direction or pull a guy, they're gone. They go so hard it seems they leave themselves open to misdirection and counters, but that seems preferable to the steady drip drip of not getting off blocks.
Ross doesn't do this. He's moving, but the wrong way. Everyone else has taken two steps here; he's gone a half yard and drifted slightly to what would be the playside if this was the standard inside zone. Bolden, by contrast, sees what's going on and gets on his horse.
A moment later, Ross is kaput, Bolden is moving at the LOS, and there's a pretty big hole because Black is not a nose tackle.
Bolden makes contact at the LOS. This is a good place to make contact, but the thing that bugs me here is something I can't show you in a still.
Here's a still anyway. Bolden's got to the LOS and he's got this tackle and he forms up. Okay. But even if Ross is here, the RB is going outside of Bolden. All he does is make the gap somewhat small. He has not MADE PLAYS.
As I watched this I started getting frustrated with Bolden's approach. This is a technique thing and I may be wrong, but don't you want this contact to be less dainty?
I want some BOOM at the end of that approach. Bolden just kind of catches the guy, which has two negative impacts. One: he does not go BOOM. If Bolden really whacks this guy he has a good shot at giving Cox no crease, or at least forcing him to slow down and pick another one. Two: he cannot make a tackle because he hasn't hit him hard enough to set up outside. No tackle, no funnel, no point. His ability to impact the OL at the LOS is essentially irrelevant because he didn't turn it into the Situation BOOM [tm shutdown fullback].
Like, I want to insert a little fireball when Bolden makes contact here. Instead, crumpets. There is some amount of control that must be deployed to prevent you from not impacting the play. Here the control makes you… not impact the play.
Anyway. Cox bursts through the hole…
…and is hewed down after six yards.
Things and Stuff
Once you've committed to the run you should COMMIT TO THE RUN. Whenever you're hitting a blocker in the backfield you get a check-plus for your read. But because Bolden just impacts the guy softly, he does not force Cox into a new hole. He doesn't even get the diving arm-tackle attempt Ryan puts in, and Ryan has contain responsibility.
Bolden needs a little Ross in him on this play. Not the Ross on this play. The Ross on other plays. The ones where he meets a guy at the LOS and that guy ends up on his back, antennae flailing in the air.
I don't get what Michigan's reading. You can't chalk this up to Ross being a freshman since he's a freshman who seems in the process of displacing Kenny Demens and Michigan linebackers have been frustrating like this for a year-plus now. Are they supposed to be looking in the backfield? Are they making Mattison chew his lip in frustration? Does Michigan require their DL to fill a bunch of these holes and want to use LBs as a cleanup crew?
I don't know. I hear Alabama LBs talk about what they see before a play even starts…
…and I'm like whoah. It doesn't seem like Michigan's getting much of that.
Big dang hole here. Black gets put away, but I'm not sure that's a problem with him. He doesn't know a tackle pulled. He sees the guy in front of him start inside zone blocking. He wants to get in his gap. He does. This goes back to the questions about Michigan's line slants against Alabama. If the DL controls his gap and you've got the extra guy who knows where the line is going, you should have a free hitter somewhere. Michigan has not gotten that much this year.
Ryan: active. Here he almost makes a great play by coming upfield of his guy and making a tackle attempt without giving up the outside. He did this late enough that his attempt did not open the hole any wider. He's a quality player.
Nice fill from Gordon. This is only six yards despite a tailback running untouched through the LOS because he comes down well and tackles in space.
3/23/2012 – Michigan 2, Cornell 3 (OT) – 24-13-3, 15-9-4 CCHA, season over
Shawn Hunwick first stepped on the ice for a 18-16-1 Michigan team that had seen its at-large NCAA hopes evaporate during a dismal road sweep at the hands of Nebraska-Omaha.
No one wanted him out there, but at least it didn't much matter. This year's team was in danger of missing the tournament in November and recovered to finish second in the Pairwise. Two years ago they had missed it, period, until they lost their starting goalie and inserted a guy who came to Michigan with no illusions he'd play.
That was the catalyst for a change in Michigan's fortunes. Involuntarily pulling Bryan Hogan was another outlet for the dread everyone was feeling at the near-certainty that Michigan would break its tourney streak. Those in the stands reacted by assuming that every shot at or in the general vicinity of the net would either go straight in (in the case of shots that needed no assistance) or be deflected into the dead center of a wide open goal (in the case of shots that were not already on net).
The team felt the same way. They responded by swarming into the slot in a great mass to sweep away the fat, glistening rebounds Hunwick seemed to give up on every shot, no matter how harmless. Their certainty that Hunwick would be overrun led to a 4-0 shutout.
The next night they'd finish the regular season by giving up five goals in an untelevised road loss. Did they relax? I don't know.
Michigan entered the playoffs the next weekend and went on a rampage. They continued to patrol their own slot with feverish intensity, and this translated into the "jump" hockey coaches and commentators are always using to define that ineffable quality a hockey team has when its passes are going tape to tape and the opponents keep finding inconveniently located defenders.
The jump lasted three games. They swept Lake State out of Yost, then bombed Michigan State 5-1 at Munn. The second night they leapt out to a two-goal lead and then bled it back. The first goal was just one of those things. Tristin Llewellyn took an insane elbowing penalty to put Michigan down two men and MSU passed it around until they got a slam dunk.
The second and third goals were the end of the ride. They were both power play goals—Llewellyn would watch State score from the box three times in three minutes—but they were pillowy soft ones. This was the moment at which it all came screeching to a halt and Hunwick was revealed as the walk-on he was. Michigan went to the locker room down 3-2 after one, certain that anything they let on net was going in. The jump had left Michigan's step.
Michigan State got one shot in the second period. It did not go in. That period was twenty minutes of battering a door until it hung by the barest sliver of a hinge. Three minutes into the the third, it gave way.
State managed 22 shots for the game but no more would get past Hunwick; Michigan tilted the ice decisively in the second, tied it, and finished the job in the third. The next weekend at the Joe, Michigan allowed 22 shots to Miami and 18 to Northern Michigan as they secured a streak-extending bid with the most rousing CCHA playoff run they'd had since the days when Michigan was looking up at the Lake States of the world.
They played like banshees. They died like Vikings. They did so because they didn't know what the hell was going to happen when someone threw a puck at the net.
Two years later, Shawn Hunwick is possibly the best Michigan goalie of all time and it's overtime because Michigan had a goal disallowed because Michigan always has a goal disallowed.
Michigan wins a faceoff and gets a shot off that is saved and caroms to Cornell. Cornell turns the play back against a third line of Luke Moffatt, Derek Deblois, and Travis Lynch. Moffatt is there to provide a third man back against the rush.
The defenders can't handle the rush that well and end up giving up a scary shot from a Cornell forward cutting left to right in front of the net. Hunwick's way out of the blue ice, because he's always way out of the blue ice because he's 5'6". He gets his right pad on the shot. He's 5'6". He has limited options when it comes to leg angles that kick pucks places. His choice here is between letting the thing into the net and kicking his leg as straight as he can so that there's no angle for the thing to go in. He's got a save percentage above .930. He's a Hobey Baker finalist. He kicks it out into the slot, like he did against Notre Dame, over and over again.
Moffatt's there, but in a bad position. His check is crappy, he doesn't tie the guy's stick up sufficiently, the guy puts it in the net, and Hunwick is over. All that's left for him to do is take the puck that was in the slot and is now in the net and hand it to Cornell. Deblois and Lynch are cruising into the defensive zone still. They don't look much like banshees, and they're not there in the slot. They're sophomores—juniors now—and don't remember what it was like when Shawn Hunwick was a 5'6" walk-on and not a Hobey Baker finalist.
The Horrible Horrible Power Play
For the third straight year Michigan's season ends 3-2 in overtime thanks in part to a disallowed goal. The rage factor on this one is lower than the other two because it came with 58 minutes to play, was not disallowed because the ref blew his whistle, and there's not enough rage to go around this year thanks to the power play.
Michigan's terrible awful power play entered the NCAA tournament 46th nationally and leaves it 48th, where they'll stay since everyone else around them is done for the year. Michigan spent half the
third second period up a man, almost three minutes of that time up two, and achieved a –1 goal differential in that time. That was the game right there. Michigan finished 0/7 on the power play, gave up a power play goal on one of Cornell's three opportunities, and conceded a shorthanded goal for the first time all year.
It's clear there's something wrong with the power play that can't be explained away by pointing to a lack of talent. Michigan hasn't had a power play you could actually call good in four years despite consistently putting up a lot of offense:
|YEAR||PP RK||Goals per G||Goal RK|
You can say '09-'10 is slightly above average, but that's all. Meanwhile Michigan continues to finish around the top ten in scoring despite not getting much production out of their power play. If their ability with a man advantage roughly corresponded with their 5x5 scoring this year* Michigan would have put up 13 extra power play goals and leapt into the top five in scoring.
It's hard to take the argument that Michigan just doesn't have the talent seriously when outfits like Bemidji State, Western Michigan, Northern Michigan, and Ferris State all finish 20+ spots ahead. Zero of those teams have NHL draft picks littering the roster, let alone a set of offensive defensemen like Merrill, Moffie, and Bennett.
This is a coaching issue. Watching Michigan cluelessly bat it back and forth from one covered guy to the other one on the five-on-three should make that clear. No one moves, no one has a plan, and the most common thing to do is fling a point shot into a defender's pads. Red is the king of all he perceives but this is a major problem that doesn't look like it's going away.
*[The #10 power play, North Dakota, converted at a 22% rate compared to Michigan's 14.6.]
The disallowed goal. I don't think Moffatt's impact changed the outcome of that play. The goalie was already sliding away from the puck and had no idea where it was. That said, Moffatt did impact the goalie in the crease, and it didn't look like his defender had anything to do with it. I don't think it's an outrageous injustice. It's very frustrating, of course, but if the ref screwed that up he more than made up for it with the avalanche of Cornell penalties Michigan could do nothing with.
The penalty shot was a terrible call, but at that point I think I preferred it to the alternative since Michigan was down, had a power play, and was playing a team without a ton of offensive skill.
Merrill: WTF? Also Moffie. The biggest reason Michigan lost other than its power play was the Merrill-Moffie pairing. Moffie initiated the sequence that led to the shorthanded goal with a suicide pass to Merrill; Merrill screwed it up at the line and the two-on-one started. Then Merrill took a swipe at the Cornell saucer pass with his stick instead of getting his body into the passing lane, leading to a slam dunk.
On the winner it was Merrill and Moffie who combined to let that rush turn into a dangerous shot; Merrill got too far outside and again out of the passing lane. Moffie also added a stupid crosschecking penalty seconds into Cornell's dubious major; it was Merrill who ended up giving up the (admittedly ludicrous) penalty shot.
Merrill has not played well over the last month. He was responsible for goals against Northern Michigan, Bowling Green, Western Michigan, and Cornell and hasn't been as superb with the puck as he usually is. I'm not sure what's going on there but he doesn't seem focused.
CCHA: not so much. The conference got almost half its membership into the tournament this year but saw four of its five teams flame out in the first round. Ferris State got past injury-riddled Denver and Cornell to make its first Frozen Four, and congrats to them.
Everyone else went out in game one. Takeaways from this:
- A conference where no one can score that was won by a team without an NHL draft pick on it is not that good at hockey.
- Non-conference games are hugely important because they are so sparse and provide the basis of comparisons between conferences.
That latter issue should evaporate after next year. Western college hockey will reform itself into three conferences from two and Michigan will have 14 nonconference games instead of six. Hopefully those aren't all home series against Bentley during football season.
A glance at next year. It's hard to predict without knowing the results of the NHL draft and whether Michigan will suffer early departures. A hypothetical no-defection defense corps looks pretty good:
That's light on sandpaper but should have no problems moving the puck. The only problem is that Michigan could lose the first three guys listed above. Bennett came in saying outright that he would not be a four-year player, Trouba is good enough to be signed immediately by an NHL club, and who knows what Merrill's attitude will be towards a hypothetical junior season after the rollercoaster he went through. Losing one guy is survivable. Two is worrying.
Michigan really needs a big leap forward from Serville. He's a lot younger than Chiasson, has a decent NHL draft pedigree, and seemed to be moving forward late in the year. If he can develop into a solid second-pairing type it'll be okay.
At forward, Red will put them through the blender but one man's rough guess:
- Moffatt-T. Lynch-PDG
- Random assortment including Rohrkemper, Sparks, Other Lynch, and freshmen Daniel Mile and Justin Selman
It's possible Nieves comes in and forces himself onto the top two lines but I'm guessing Red will go with a defense-oriented player over the freshman. Defections here are also possible, of course: Guptill, PDG, and Brown are all potential departures. People keep talking about PDG leaving but I'd be surprised if an NHL team is eager to sign him just now. His 26 points are good for a freshman but not Pacioretty good. The kind of guys who have left after one year have driven play more than PDG did.
The biggest change will be in net, where NTDP goalie Jared Rutledge replaces Hunwick with Junior A vagabond Steve Racine backing him up. Rutledge's Pointstreak page is a little scary—a drop in games and performance from year to year—but the embarrassingly primitive spreadsheet the NTDP uses to track its stats shows that over the course of the year Rutledge has a .902 versus teammate (and Ohio State commit) Collin Olson's .893. NTDP save percentages can be pretty ugly since a big chunk of their games are against college teams, so that's fine. Rutledge is a small, aggressive, technically-sound goalie who sounds a lot like Hunwick.
BONUS SPREADSHEETIN': Michigan's 3 NTDP U17 commits are #1, #4, and #5 in scoring on their team. JT Compher is the guy at #1 and has played 7-8 fewer games than the rest of the team. He's the only guy with a PPG. Tyler Motte is neck and neck with Miami commit Anthony Louis and UNH commit Tyler Kelleher for #2; Evan Allen is a half-dozen points back of that group. With those three guys and Bryson Cianfrone, a Canadian Junior A player who was projected as a first round OHL draft pick before committing to Michigan, Michigan looks like they'll have a dynamite 2013 class. Pending defections, of course, Always pending defections.