Wednesday, February 17, 2016

Engineer Bracketology: Week 7

Well, the Engineers are certainly still hanging around. The win over Harvard boosted them up as one of the best RPI-inducing wins of the year for Rensselaer, they got some good results elsewhere, and they managed to weather the loss to Dartmouth too.

1. Quinnipiac*
2. St. Cloud State*
3. North Dakota
4. Boston College*
5. Providence
6. Michigan*
7. Notre Dame
8. Denver
9. Omaha
10. Yale
11. Boston University
12. Harvard
13. UMass-Lowell
14. Penn State
15. St. Lawrence
16. Minnesota
17. Rensselaer
19. Michigan Tech*
25. Robert Morris*
In since last week: Penn State
Out since last week: Cornell
Last in: UMass-Lowell, Penn State
First out: St. Lawrence, Minnesota
League breakdown: 5 HEA, 4 NCHC, 3 ECAC, 2 B1G, 1 WCHA, 1 AHA

For the record, we've got Boston College vs. Penn State and Providence vs. Harvard in Albany this week.

Moving on to the QWB cliff. Bear in mind that when it comes to quality wins, right now the top QWB-boosted team in the nation is Boston University, whose boost is .0076. The overall impact of this boost is diminished because every other team they are competing with in Pairwise has some kind of QWB boost (only 11 teams don't have a quality win or tie, and none are in the Top 30). For instance, when compared to Boston College (.0070 QWB), it's only a net benefit of .0006 extra in the RPI, which is almost nothing.

And to fully compare, a win for Rensselaer over Quinnipiac, before considering any other result, would boost their RPI by 0.043 - exactly the amount of their current QWB from five quality wins and two quality ties. So in the grand scheme of things, wins and losses are more important than QWB - but it's nice to have and is a component worth looking at. So here's the cliff.

16. Michigan Tech .529
17. St. Lawrence .529
18. Rensselaer .526
19. Dartmouth .525
20. Cornell .525
--
21. Clarkson .524
22. Miami .522
23. Robert Morris .521
24. Minnesota Duluth .521
25. Minnesota State  .520

Union remains in "it could happen" striking range at 29th in RPI, but it's looking less and less likely as time goes by. Even a weekend sweep this weekend (including Quinnipiac on Saturday) would probably not move them up more than a couple of spots. One thing that could help Union move up farther would be teams in front of them in RPI losing. Most of them are already on our "rooting against" list anyway, but we can probably add 26-28 RPI - Bowling Green, Northeastern, and Ferris State.

Quality wins
Boston College (#4, at home)
Yale (#10, at home)
Harvard (#12, on the road)
St. Lawrence (#17, on the road)
Cornell (#20, at home)

Quality ties (half a quality win)
Quinnipiac (#2, on the road)
Harvard (#12, at home)

Missed opportunities
Michigan (#6, at home)
Yale (#10, on the road)
Harvard (#12, neutral ice)
UMass-Lowell (#13, on the road)
Dartmouth (#19, at home and on the road)

Remaining opportunities
Quinnipiac (#2, at home) - Feb 19
Cornell (#20, on the road) - Feb 27

Yes, Quinnipiac is now #2 in RPI, though they remain in first in the Pairwise entirely on the strength of their two-game sweep of St. Cloud State in October. They're still a solid QWB haul, even at home. Needless to say, an upset win on Friday would be pretty outstanding.

Looking around, there actually aren't really too many comparisons to look at this week. BC is untouchable on COp right now, the Michigan comparison is now all but unwinnable thanks to Rensselaer's loss to Dartmouth on Saturday (although it was always a huge longshot anyway), and UML is still 100% RPI until the playoffs, as are Yale, St. Lawrence, and Harvard. And, as always, teams that are not on the Engineers' schedule are always 100% RPI.

Dartmouth (2 H2H losses) - With the Big Green having swept the Engineers this season in league play, that pretty much solidifies this as an RPI/COp requirement for Rensselaer to win this comparison. They have that now, but it's by razor-thin margins on both. This, essentially, makes Dartmouth Public Enemy #1 when it comes to the Engineers and their NCAA at-large hopes.

But there's a catch - and again, it's in Dartmouth's own comparisons with other teams. While Union's win over the Big Green on Friday helped drag down Dartmouth's PWR positioning by keeping Vermont in the driver's seat on their comparison with the Big Green, Dartmouth has a comparison that they themselves are currently winning against a team ahead of them in the Pairwise - Cornell. And that's a win that's actually helping the Engineers for the time being (and possibly into the future). And wouldn't you know, they play each other this coming Saturday in Hanover. More on that in a moment.

Total COp: 7.000-6.250 Rensselaer
Quinnipiac: 0.500-0.000 Rensselaer
Princeton: 1.000-0.500 Dartmouth
Colgate: 1.000-0.500 Dartmouth
Cornell: 1.000 each

Cornell (H2H win) - The win keeps this one a 3-0 comparison victory for the Engineers, but with Cornell so close in the PWR (18th, just below Rensselaer) and both RPI and COp being so close, it's probably worth rooting against Cornell where possible.

The rub, as discussed above, is in their game with Dartmouth, and the potential impact of that game on the Cornell-Dartmouth comparison. The Big Red lose this comparison right now on COp/H2H. A Cornell victory would cancel out the H2H and return this to a straight RPI comparison, thereby flipping it back. Bearing in mind that we also want Dartmouth to lose games (Public Enemy #1, remember?), that makes this one a toughie to call.

Looking at the COp breakdown below, a Cornell win would boost the Big Red's COp against the Engineers by 0.500, something the Engineers could only hope to match by sweeping Quinnipiac and Princeton for +0.250 each. And even then, Cornell can also gain 0.500 by beating Harvard on Friday.

Now looking above at Dartmouth's COp breakdown, we see that the Big Green are utterly unable to improve their COp standing against the Engineers on their own. A weekend sweep still leaves them at 6.250. They would depend on losses by Rensselaer to improve the comparison. Even the Engineers being swept still leaves Rensselaer leading in COp 6.500 to 6.250 (although RPI would probably flip it to the Big Green anyway).

It seems like there's a lot less danger in taking Dartmouth to beat Cornell, overall. Yes, Cornell would almost certainly drop out of the top 20 in RPI, but for future positioning it's a safer bet, especially if the Engineers pick up a win or a tie against Quinnipiac at home on Friday first. Otherwise, this game becomes a complete minefield that's pretty much impossible to avoid getting blown up.

Total COp: 5.000-4.750 Rensselaer
Quinnipiac: 0.500-0.250 Rensselaer
Princeton: 1.000-0.500 Cornell
Harvard: 0.500-0.000 Rensselaer
Dartmouth: 0.000 each

Clarkson (H2H loss) - Both teams last week gained the same amount in COp (0.250), so we're basically just where we were last weekend with the Golden Knights comparison. Again, the Engineers must keep winning in both RPI and COp to claim this one, and they are right now. Extra reason to root against Clarkson, besides PWR and ECAC positioning and the obvious (being Clarkson).

Total COp: 8.750-8.250 Rensselaer
Quinnipiac: 0.500-0.250 Rensselaer
Princeton: 0.500 each
Brown: 1.000-0.750 Clarkson
Yale: 0.500 each

Miami (H2H split) - Going to talk about the RedHawks here in part because they're making a late season run and becoming relevant in the Pairwise. For the purposes of straight up Pairwise, Miami might as well be treated like a team the Engineers didn't play this year, because the series was split and the two H2H results cancel each other out. But the win Rensselaer has over Miami is a road win, and so the potential QWB bonus is heightened a bit and could come into play - even while the RedHawks start to reach the point where they're a potential rival for NCAA positioning.

It's a difficult balance - but ultimately, a Rensselaer sweep and a Miami sweep would likely still move the RedHawks ahead of the Engineers in the PWR on RPI. So it's tough to flat out root for Miami this weekend against Western Michigan (who also happens to be a past opponent). A split might work well for both sides of the ledger, but a WMU sweep would be preferable to a Miami sweep, despite the potential for improving Rensselaer's QWB.

As we get into this weekend's cheering section, it's probably worth pointing out that there's not a lot of room for growth this week. With Minnesota not playing this week and UMass-Lowell only playing one game (that they're likely to win), there's not a lot of solid potential for damaging their RPI. The only comparisons that could reasonably be flipped this week (while not seeing others below flipped against) are St. Lawrence and Penn State. That would move the Engineers ahead of both and probably the Gophers as well, but any further would be difficult. Haven't seen a scenario where the Engineers move any higher than 14th after the coming weekend.

By the way, interesting little nugget - if we get exactly what we're looking for from the cheering section this week, Quinnipiac would fall all the way to 5th in the PWR. That'd be a shakeup (and unlikely).

Just to stress once again: nothing helps the Engineers more than winning their own games against all comers, from #1 PWR Quinnipiac to #55 Princeton this weekend. We're deep enough into the season that it's basically impossible to lose and still rise in the Pairwise. But where they end up in the PWR in relation to their own results is heavily influenced on what goes on elsewhere.

The key for the cheering section:
1. Wherever there is direct COp assistance
2. For the ECAC out-of-conference
3. For improved quality win bonuses (QWB)
4. Against teams ahead of or just behind Rensselaer in the RPI or PWR
5. For non-conference opponents

Engineers Cheering Section
Friday
Brown over Clarkson - 1/4
Harvard over Cornell - 1/3/4
Union over Princeton - 3
Yale over St. Lawrence - 3/4
Colgate over Dartmouth - 1/4
UMass over UMass-Lowell - 4

Weekend series
Colorado College over Denver - 4
Ohio State over Penn State - 4
Western Michigan over Miami - 4/5
Boston College over Vermont - 3/5
Michigan over Ferris State - 5
New Hampshire over Boston University - 5
North Dakota over Minnesota Duluth - 4
Bemidji State over Michigan Tech - 4
Air Force over Robert Morris - 4
Alaska-Anchorage over Lake Superior State - 5
Alaska over Bowling Green - 5
UConn over Northeastern - 3

Saturday
Brown over St. Lawrence - 4
Yale over Clarkson - 3/4
Harvard over Colgate - 3 (the reverse is probably just fine, preferable for the ECAC race)
Dartmouth over Cornell - 1/4
Union over Quinnipiac - 3

Important COp/H2H games for currently relevant comparisons
Saturday, February 20: Cornell at Dartmouth
Friday, February 26: UMass-Lowell at Boston College
Saturday, February 27: Boston College at UMass-Lowell

No comments:

Post a Comment