In With A Chance

1/19/2018 – Michigan 4, Penn State 0, 11-10-2, 6-7-2 Big Ten
1/20/2018 – Michigan 3, Penn State 2, 12-10-2, 7-7-2 Big Ten

incoming [James Coller]

Here's a weird graph.

That's Michigan hockey's Pairwise ranking. For those who may have forgotten this weird thing about college hockey, the Pairwise is a computer ranking that selects and seeds the NCAA hockey tournament. Yeah, man, the robots took over like 20 years ago and hockey is fine with it.

This graph shows Michigan leaping from the mid-20s in that ranking to 15th last weekend. There are 16 teams in the tournament and one conference that will definitely occupy a spot with an otherwise-out auto-bid team. So... yeah. They'd be in if they got lucky with conference tournaments.

Hockey's made some tweaks that make some games more valuable than others but not even I, a person who knows what the Pairwise is and can give you its life history, believed that an 8-10-2 Michigan team could haul its ass from "not even on the bubble" to "maybe in the tournament" with one weekend sweep against Minnesota. Nor did I expect a sweep against another bubble team, Penn State, to do nothing further. I guess Michigan consolidated its gains?

Anyway, this caught me off guard. I hadn't even looked at the Pairwise all year since there seemed to be no reason to. I'd been interpreting the hockey team in front of me as a try-hard, much-improved outfit that had too far to go in one season. This weekend I saw... basically that, but now with a tourney bid at stake. Against the rootinest, tootinest, shootinest team in all the land. Penn State.

[Bill Rapai]

----------------------------------------------------

If there were seven seals that had to open before last year's Michigan Hockey Apocalypse, #2 or #3 was A Penn State Program The Same Age As My Son, Who Regularly Has To Be Told Not To Bite People, Gives Michigan The Business. Year 1 PSU went 3-2 against Michigan despite winning just five other games all season and knocked Michigan out of the tourney with a two OT win in the conference tourney. Year 2 PSU went 3-1.

Adding insult to injury, these Penn State teams played like Red Berenson teams, zipping up and down the ice pell-mell while trying to score goals with skill, if skill would do, or sheer erosion if it wouldn't. Penn State was (and is) prone to exploding into bits because they don't quite have the talent to do what they want to do, but from the drop they seemed to be a version of Michigan on the rise instead of the decline.

Penn State's still Penn State. At certain points their sheer aggression overwhelms. Michigan got out of the gate slow on Friday and shots were 9-3 PSU despite a Michigan power play ten minutes in. One particular shift on Saturday saw the Slaker line so gassed they couldn't even ice the puck; they ate two solid minutes of desperate defensive zone time. Lavigne got the biggest cheer of the night for swallowing a puck directly in his chest to end this shift.

That shift was the second half of the Saturday game writ large. Michigan staked itself to a 2-0 lead and seemed to have the much better chances through 30 minutes even if a bunch of point shots had PSU ahead in shots on goal. Dave had Michigan with an advantage on shots in the "house" through 20, and then the roof caved in.

The rest of the game was spent with Penn State even more amped up than they usually are. The last ten minutes of the second period was nonstop PSU attacking, with the defensemen pinching on almost literally everything. Whenever Michigan broke PSU's line they had a shot at an odd-man rush that they couldn't quite put together. (A major reason why: Yost's ice still sucks even after they redid the plant.) Michigan struggled to cope in the third, with PSU launching 31 shot attempts, almost half of them from the house.

Lavigne stood tall; puck luck bit PSU. And then Penn State pulled the goalie with three minutes left. That was finally it. PSU scored twice and made the final minute a sphincter test. They lost because Michigan did pay off on one of the opportunities PSU's hyper-aggression provided, when Dakota Raabe scored the first goal of his career on a breakaway.

If that didn't feel too different than previous PSU games, the Friday game did. Michigan outshot and significantly outchanced PSU after the rough start; that 4-0 win probably should have been 4-2 or 4-1, but it was not that scramble to hold on Saturday.

That's the team: working for it. Getting there. Still liable to get clonked. But in it. For the duration.

DON'T CALL IT A COMEBACK, CALL IT THE PAIRWISE STATUS

Hey, pairwise is relevant again! Hooray. I used to have to eyeball this stuff but there is now an excellent site called College Hockey Ranked that plays out the various scenarios over and over again and spits out graphs. Here's Michigan's graph:

To (probably) be in position for an at-large when conference tourney play starts Michigan needs to go 6-4 down the stretch. That would be about a 75% shot at the #13 spot, which is safe in all but the most apocalyptic bid-stealing scenarios. At 6-4 they would probably have to win their first round series to feel good about securing the bid. 7-3 would be pretty secure. 8-2 and they'd be close to 100% safe, but that's a tall, tall order.

Because college hockey's RPI gives you bonuses for beating highly-ranked teams and weights a road W 50% more than a home win, the most important games left by far are this weekend's series at OSU (#6 PWR) and the ND (#1 PWR) home and home, especially the road half of it. Even a split this weekend is pretty good.

Since the Pairwise is now identical to RPI unless head to head intervenes, with limited exceptions anything that's good for Michigan's RPI is good for their pairwise. The main thing to root for is for Michigan's opponents in the top 20 to do well unless there's a direct conflict.

This means you're in favor of Minnesota winning games—Michigan has already banked 3 Ws against them. Penn State is a little trickier because they still manage to win their comparison with Michigan because their 3-1 H2H deficit is made up for by their better RPI and Common Opponents, and then RPI breaks the tie. Your pecking order in conference:

1. Minnesota—Gophers literally cannot win a comparison with M so win away, Goldy. This may change if M and Minnesota are specifically competing for a last at large spot late. For now, go Gophers.
2. OSU/ND—both of these teams are big quality win bonuses waiting to happen if Michigan can pull games off of them in their upcoming series, and are so far ahead of Michigan that there's no way Michigan could win a comparison with them.
3. MSU—irrelevant.
4. PSU—Quality win bonus, on the one hand, bubble team on the other. Michigan can win or lose this comparison based on PSU does, and their QWB isn't huge.
5. Wisconsin. Currently safely out but plays their way on to the bubble with 5 wins in their last 8.

Nonconference teams you hate because they're on the bubble are Providence, UNO, Northeastern, Harvard, BGSU, Northern Michigan, Miami, and BC, but this far out it's about 90% Michigan and 10% how others play. This is especially true because the league has four bubble teams (M, Minnesota, PSU, and Wisconsin). If they finish third in the league they're probably good.