First Look Inside the Pairwise for 2010
by Adam Wodon/Managing Editor
Boy has this year's NCAAs snuck up on me. This is likely a sign of getting older. Time flies.
With the NCAAs set to be bracketed just 3 1/2 weeks from now, this is our first analysis of the season — the 14th straight year of doing these kinds of articles. Since CHN's inception in 2005, we have been consistently the most accurate in projecting the brackets once all the games have been played, and in explaining why.
Note the caveat there — "once all the games have been played." We don't try, at this point, to project the brackets. Making projections, while fun, is not the intent of these articles. There are so many millions of possible combinations of outcomes in the coming weeks, that it's impossible to account for them all, even with a bit of educated guesswork. Even with this very objective system the NCAA has in place, what happens to each team is affected not just by themselves, but what happens in other games they're not even playing, and for a variety of reasons.
Also, even if you can narrow down the seed of a team, or a range of likely finishes in the overall Pairwise — projecting where those teams will play in the Regionals is also next to impossible, because it relies on other external factors. Teams are bracketed — as best as possible — according to a strict serpentine ordering, such as 1-16-8-9 and 2-15-7-10 and so on. And No. 1 seeds are placed in their region first. So before we can know where No. will be placed, for example — even assuming we can project which team No. 10 will be — we need to know which teams will be at places 1, 2, 3, 4 and 7. Then you have to account for the potential first-round matchups against teams from your own conference, which the committee avoids, thus creating more potential shuffling.
That's not to make excuses for not analyzing things. You can still very much analyze the current state of the Pairwise, and we do. The point is, this is not about projecting but about trying to narrow down as best as possible each team's chances for certain seeds, or to make the NCAAs at all. And to pinpoint the areas of possible upside or downside for a team, and the potential pitfalls the committee will face when trying to seed the teams.
Something to keep an eye on: How much the men's ice hockey committee will be forced to "Regionalize" the bracket. Apparently, not at all. This was discussed here and here last year, if you want to see more on what the means. According to current NCAA men's ice hockey committee chair Bill Bellerose, the NCAA — after tabling the issue last year — never revisited it, and as far he knows, there will be no regionalization mandate whatsoever.
So, that said, let's get on with it. And again, if you're unfamiliar with how this very precise and objective system works, it's best to read up on our CHN NCAA Tournament Primer. It has thorough and easy-to-understand explanations of all of the intracacies of the process that's come to the be known as The Pairwise, and lots of historical reference material and links to old articles.
1-2. Denver, Miami
These two are so tight in every criteria, that it could flip-flop six more times before we know which team is which. Ultimately, it probably doesn't even matter. Both of these teams also have a significant edge over all other teams, including everyone from Wisconsin at No. 3 and down. So no matter how you slice it, it's pretty apparent that Denver will play in St. Paul, and Miami will play in Fort Wayne.
The only question for Denver is whether the committee takes the "a flight is a flight approach" to placing it. In other words, if St. Cloud State or Bemidji State, two schools very close to St. Paul, is a No. 1 seed, would that team host in St. Paul while Denver goes East to Worcester? Possible. Perhaps probably.
Funny addendum ... if Holy Cross wins the Atlantic Hockey tournament, and is thus the No. 16 overall seed, it is hosting the Worcester Regional, meaning the No. 1 overall seed would go to Worcester anyhow.
Splitting with St. Cloud State last weekend was sufficient for Wisconsin to maintain this position. The Huskies are really the only team Wisconsin is in much danger of having a comparison flipped on, so even if that happens, Wisconsin would still be No. 4 in that sceanario. Either way, you're looking at the Badgers playing in the East — which means Albany or Worcester this season.
4-5. St. Cloud State, Bemidji State
These two teams are making the NCAAs, that is not a question at this point. That's another milestone for Bemidji State, which would make it at large for the first time, one year after its thrilling run to the Frozen Four, and one year before moving to the WCHA. It's very possible these two could be bracketed with each other somewhere out East. BSU has very little opportunity to move up, since it won't be playing any teams of note the rest of the way (one more regular-season weekend, and then the four-team CHA tournament). Also, the Beavers have only played seven games against Teams Under Consideration (TUC). Ten is needed for that criteria to have any effect. So really, the only thing propping up Bemidji in this slot against teams right below it, is the RPI, and that is close enough in the case of Yale and Boston College, in particular, where the Beavers could easily slide if things break wrong for them.
6-7. Yale, Boston College
Two more teams that are already locks for the NCAAs. Yale has a chance to make a move by winning out in the ECACs, the problem is, even with that, it might max out at nine games vs. TUC, which won't be enough for that criteria to matter. That will put all its hopes on the RPI criteria, and a lot would have to break right to surpass teams like St. Cloud State in that area.
Boston College, on the other hand, has much more significant potential to rise. Surpassing Yale and Bemidji State is a simple matter of winning and improving the RPI vis-a-vis those teams. Against St. Cloud State, even if the Eagles can't catch the Huskies in RPI, the Record vs. TUC and Record vs. Common Opponents are very close. The one common opponent is Denver. If St. Cloud State loses to Denver in the WCHA tournament, while BC wins out or even comes close, then Boston College could easily find itself in the No. 4 overall slot — which would mean hosting in nearby Worcester.
8-9. North Dakota, New Hampshire
You start to get into some tentative territory with these teams making the NCAAs, but for UND and UNH, it's looking pretty good. It's pretty hard to see them shooting near a top seed at this point, but it doesn't appear likely they could fall enough to drop out of the tournament. North Dakota's record vs. TUC is already not great — but that's a positive, in a sense. It means they can't get worse in that area, and yet are still No. 8. The interesting thing to watch will be the matchup issues with WCHA teams in the first round, something — as most of you are probably aware — the committee is mandated to avoid at all costs. Depending on where Minnesota-Duluth and St. Cloud State may or may not fall, it could get tricky for the committee, but having two, or maybe three, No. 1 seeds from the WCHA, should actually help avoid the problem.
10-11. Michigan State, Alaska
This is where things start to get quite messy. First off, if things ended today — which they don't, but hey — Alaska would make the NCAAs at large for the first time in its history. That's a great accomplishment in and of itself. Helping prop up Alaska are some wacky comparison wins over St. Cloud State and New Hampshire, based precariously upon some interesting Common Opponents. But those could easily flop the other way.
One way to do these bracket analyzations is to see where the upside is vs. the downside. Teams like Alaska are already maxing out their upside potential, with two comparison wins against teams significantly higher in the Pairwise than itself, and with little chance to make gains with others right around it — especially since the next four games Alaska plays will be against teams that aren't TUCs. There are, however, four teams BELOW Alaska in the overall Pairwise that currently beat the Nanooks in that head-to-head comparison — Minnesota-Duluth, Cornell, Union and CC — and all but CC's has a good chance of going the other way.
So that's what we mean by the messiness that envelops the teams in this vicinity. It's very volatile.
Alaska's potential seems to be in the 10-17 range. Michigan State's potential seems anywhere from 9-15 or so.
12-13-14. Minnesota-Duluth, Ferris State, Cornell
Things are so bunched up and so transient in this group, that it's difficult to make heads or tails of any of it. But let's see. Minnesota-Duluth has some good upside potential, but its loss during the season to Vermont — one little non-conference game — is causing it to lost four comparisons to Hockey East teams below it in the overall Pairwise, and those will be difficult to earn back. So the upside is limited without a major run. ... Ferris State has a poor record vs. TUC, but with most teams around it, that's not coming much into play. But that means only the RPI is propping up Ferris at this point. The comparison with Union is hinging on future results of the two teams against Lake Superior and Yale, respectively. ... Last week's loss to Dartmouth was a killer to Cornell, and it has a huge game coming up this weekend against Union at Lynah Rink. If Cornell wins that game, and gets to Albany, it should be fine. Otherwise, it will be trouble.
15-16. Union, Maine
The comparisons of both of these teams are a scrambled mish-mosh. At this point, "just win baby" seems to be the best you can say. Union will have a chance to do major damage against Cornell this weekend. Maine doesn't have any Hockey East teams directly around it. If Maine doesn't get in, Hockey East — which has won the last two national titles — would probably only get two teams in.
17. Colorado College
The Tigers win many comparisons with teams above them, thanks to a strong RPI. But it loses numerous comparisons to teams around them because of a 1-2 record against Northeastern and Maine this season.
18-19. Vermont, Boston University
It's hard to believe these two teams are even in the picture, especially Vermont, which is currently ninth in Hockey East and wouldn't even make the playoffs. But an early-season win over Minnesota-Duluth, and a split with Denver, are seriously propping up the Catamounts right now. It's hard to see much upside potential, however. ... BU, having tied Vermont in its only meeting so far this season, gets the residual benefit of the Vermont win over Minnesota-Duluth. The low RPI, however, makes it hard to surpass many other teams without a huge end-of-season run. It's possible, however.
20-21. Northern Michigan, Nebraska-Omaha
NMU's RPI is not as bad as BU's right now, so there is still some decent upside potential for the Wildcats, if they can continue to win some games. UNO is also not that far off from winning many comparisons above it, if the right teams lose to the right other teams. Of course, this requires both NMU and UNO to win a bunch of games coming up, and they can't both do that, probably. And this coming weekend, UNO has two games with Bemidji State — toughies.
This seems to be the Southern edge of feasibility in the Pairwise right now — meaning Minnesota and Michigan are on the far outside, and not even able to look in.