The PairWise Analyzed: March 15, 2006

Update: On Wednesday night, I encouraged readers to submit PairWise scenarios that contradict my findings. I received several emails with scenarios purporting to put Denver into the NCAAs, one in particular which still works even after Thursday’s St. Cloud State win at the WCHA Final Five. Among the most difficult elements are Bentley defeating Connecticut in the Atlantic Hockey final and Northern Michigan tying Michigan in the CCHA third-place game. The final result is that Denver makes it in as the last team. Thanks to Bill Towson for this scenario.

With the pre-NCAA tournament season down to its final weekend, I’m now doing what many of you probably are too — using USCHO.com’s PairWise Predictor to run various scenarios for championship weekend. Now, despite the tantalizing notion that there just aren’t that many possibilities left, consider the following: 19 games remain to be played before the NCAA selection show Sunday, and each of them has at least two potential outcomes — in fact, three of them (the third-place games) have three each.

That means that there are still 2 to the 16th power times 3 to the third power, which equals over 1.7 million possible combinations of wins, losses and ties left out there, so everything I’ve done here is the result of considerable research, but hardly an exhaustive look. If anyone comes up with scenarios that contradict specific predictions here — especially the apparent impossibility for certain teams to make the NCAAs without autobids — I’d be interested in hearing them. (Note: you must forecast using the 3-2-1 bonus, as I have done.)

And if you get to thinking that some games don’t matter, or that certain combinations of results can be ignored since they can’t possibly matter in the big picture, I encourage you to take a look at the first section below this introduction.

Remember that the PairWise Rankings work by comparing teams against one another, one pair at a time (hence the name), in four statistical categories: the Ratings Percentage Index (RPI), record in head-to-head games, record against common opponents, and record against Teams Under Consideration (TUCs) — teams with an RPI rating of at least .500.

Note: All rankings before team names below refer to PWR rankings, not the national polls. This week, due to a number of ties near the bubble that merit attention, I have ranked teams after using the PWR tiebreakers. Also, all PWR and RPI numbers cited include a .003-.002-.001 bonus for “quality” nonconference wins. Results are through all games of March 15, 2006.

Atlantic Hockey

No. 19 Holy Cross (potential shocker): Let’s start with the obvious — the AHA tournament title puts the Crusaders in the NCAA field. But as they say on late-night TV commercials: wait, there’s more. Remember the NCAA rule I mentioned last week — the one that says any team winning its league tournament becomes a TUC by rule, regardless of its RPI? It’s about to rear its head in the most interesting way. If Bentley wins the Atlantic Hockey tournament, it becomes a TUC, and Holy Cross went 4-0-0 in the regular season against the Falcons. So if Bentley beats Mercyhurst in the Atlantic Hockey semifinals and Holy Cross advances past Connecticut to set up a Falcons-Crusaders title game, the Crusaders have the best of all possible worlds. A win in the championship game would put Holy Cross in the NCAAs with the autobid — but a loss to Bentley would tack an extra 4-1-0 onto the Crusaders’ TUC record. And since Holy Cross has played very few TUCs this season, that would send its TUC winning percentage skyrocketing — enough to potentially put the Crusaders in the NCAAs as an at-large team. You read that right. Bentley over Holy Cross in the AHA championship could give Atlantic Hockey two bids, assuming a limited number of upsets elsewhere in the conference tournaments.

CCHA

No. 3 Miami (lock for No. 1 or No. 2 seed): Miami is well-positioned for a No. 1 seed in the NCAAs — likely in Albany — but the RedHawks aren’t guaranteed that seed without the Mason Cup. Even a loss in the CCHA title game could end up pushing Miami down to fifth or sixth in the PairWise if teams like Boston University, Cornell and/or Michigan State win their conference tournaments. Conversely, a two-loss weekend pretty much guarantees the RedHawks a No. 2 seed, in which case the ‘Hawks would still probably go east, depending on who the one seeds are, and in what order.

No. 4 Michigan State (lock for NCAAs): The Spartans are in, but they could end up with a seed as low as a No. 3 with a two-loss performance at the Joe combined with tournament wins by teams like North Dakota, Cornell, New Hampshire and/or Maine. One win and one loss — in any combination — leaves the Spartans a No. 1 or No. 2 seed depending on other teams, but a Mason Cup title puts MSU in as a No. 1 pretty much regardless of what happens elsewhere. It does not look possible for Michigan State to get higher than No. 3 overall, thanks to Minnesota and Wisconsin’s big lead in the RPI, but as always, strange things can happen in games that don’t otherwise seem important in the big picture.

No. 9 Michigan (near-lock for NCAAs): Even a tie in the third-place game at the CCHA championships looks to put the Wolverines in the field for good, but things get interesting with a fourth-place finish this weekend. In that case, Michigan’s RPI and TUC record drop enough to put the Wolverines in danger, with just one upset champion capable of knocking Michigan out of the field under the right circumstances then. Michigan can rise as high as a No. 2 seed by winning the Mason Cup, but a No. 1 seed looks out of the question.

No. 14 Nebraska-Omaha (50-50 shot at NCAAs): Want a look at how convoluted the PairWise can be? Look no further than Nebraska-Omaha, and see if you can follow this chain of logic. During the regular season, UNO played Minnesota State twice and lost both games. Since Minnesota State’s RPI is currently .5037, that’s 0-2-0 on UNO’s TUC record. But for Minnesota State, .0040 of that RPI rating is bonus points from beating — that’s right, UNO, a nonconference opponent currently in the top 15 (14th right now) of the unadjusted RPI rankings. Now, Nebraska-Omaha’s CCHA season is over, but other teams such as Northern Michigan and Dartmouth can still jump over UNO in the RPI, which would push UNO down to 16th in RPI, thereby taking away Minnesota State’s bonus, making Minnesota State no longer a TUC — and therefore improving UNO’s TUC record by removing two losses, which boosts UNO back up the PairWise Rankings again. Makes your head hurt, doesn’t it? That means the worst-case scenario for UNO is upset champions like St. Cloud State or Colgate, who don’t help UNO in any way by winning. One or two of those, and the Mavericks are out.

No. 17 Northern Michigan (fringe chance at NCAAs): Here’s the best scenario I can find to put NMU in the field. Right now, the Wildcats look to be three comparisons short of a shot at a bid, but there are three comparisons NMU can turn under the right circumstances without needing to win the Mason Cup. The first is Denver, which most likely turns if Northern Michigan wins one game this weekend (I say “most likely” because the comparison win comes down to RPI, which is razor-thin and can change based on other games as well). The second is Boston College, which turns if NMU beats Michigan; since Boston College played the Wolverines this year, an NMU win there would boost its common opponents criterion to a win over BC. Also, the Eagles must not win Hockey East for this to happen. The third is Dartmouth, which NMU can overcome in the head-to-head comparison if the Big Green loses two games at the ECACHL tournament. That looks to be about it — not good odds for NMU by any stretch, but still better than nothing.

CHA

No. 30 Bemidji State (autobid): The Beavers are in. No other CHA team can make it.

ECACHL

No. 6 Cornell (near-lock for NCAAs): The Big Red are as good a bet for the national tournament as any team not named Minnesota, Wisconsin, Miami, Boston University or Michigan State — but there are scenarios that push the Big Red onto the wrong side of the bubble. Not many, and not likely, but there are scenarios. Like Michigan (above) and North Dakota (below), Cornell occupies the realm of near-locks: teams for which only a last-place finish this weekend, coupled with a pile of upset champions, can knock them out of the field. Even so much as a tie in the third-place game appears to put Cornell irrevocably in the Big Dance. Conversely, the Big Red can move as high as a No. 1 seed by claiming the ECACHL championship, coupled with two-loss weekends for a couple of teams from the group of Michigan State, Miami and Boston University — as long as North Dakota doesn’t win the WCHA title in the process.

No. 7 Harvard (lock for NCAAs): I can’t find a combination of results to put the Crimson out of the NCAAs, thanks to Harvard’s sterling record against TUCs. On the other hand, Harvard can rise as high as the No. 2 overall seed by winning the ECACHL championship over Cornell (that’s the win that gives the Crimson its biggest boost in RPI), combined with Boston University failing to win Hockey East. It’s actually very easy to get Harvard into the third or fourth overall seed, regardless of what most of the favorites do in other tournaments. Failures by BU, and to a lesser extent Miami, appear to be key to Harvard’s shot at a one seed.

No. 15 Dartmouth (all but out?): The schedule, as I said last week, has decimated the Big Green’s at-large chances. Unfair as it seems, Dartmouth, which tied for the Cleary Cup as the ECACHL regular-season champion, could also fail to make the NCAAs even if it gets as far as losing in the league’s tournament championship. Going 1-1 in the ECACHLs this weekend does Dartmouth little good overall, potentially raising its RPI and TUC record a bit, but not enough to climb over the Hockey East teams down near the bubble, which isn’t going to get it done.

Hockey East

No. 5 Boston University (lock for No. 1 or No. 2 seed): Boston University can turn the comparison with Miami by winning the Hockey East championship, no matter what the RedHawks do. That’s because Miami must play Northern Michigan in the CCHA semis, while BU gets a tougher — from a PairWise perspective — New Hampshire squad en route to the title. That’s enough to put BU up into a No. 1 seed in most scenarios. At the bottom end, meanwhile, Hockey East’s lack of a third-place game means the Terriers can go no worse than 0-1 this weekend, which would leave BU in as a No. 2 seed. Either way, BU plays at Worcester as the regional host.

No. 11 New Hampshire (strong odds for NCAAs): Hockey East can get four teams in the NCAAs if enough favorites win in other tournaments (keeping at-large bids from disappearing to upset champions). The best odds belong to New Hampshire, which can get as high as a No. 2 seed by winning the tournament, and which is guaranteed a bid by winning its semifinal, thanks to a superior record against TUCs when compared to Maine or Boston College. Even the wildest series of upsets don’t appear to push UNH out of the field of 16 if it reaches the title game, although a loss in the semis makes the Wildcats fair game for upset champions. One deadly outcome for UNH: Northern Michigan winning the CCHA not only takes away an at-large slot, but it can turn the UNH-NMU head-to-head comparison, costing the Wildcats a critical PairWise point. If UNH drops its semi and NMU does win the CCHA, the Wildcats’ hopes are suddenly slight.

No. 12 Maine (favorable odds for NCAAs): Unlike UNH, Maine can fail to make the field even if it wins its Hockey East semifinal, though it requires a lot of underdogs to win elsewhere. More likely, a loss in the title game puts the Black Bears in the NCAAs as a No. 3 seeds somewhere, and a championship could push Maine up to a No. 2. A loss to Boston College in the semis, though, leaves Maine hoping for a dearth of upsets — in particular, the Black Bears would be rooting against BC in the final, especially if the Eagles’ opponent turned out to be New Hampshire. Still, even a 0-1 weekend leaves Maine plenty of chances to make the NCAAs.

Chris Collins' Eagles can sink or swim this weekend in the Hockey East tournament (photo: Melissa Wade).

Chris Collins’ Eagles can sink or swim this weekend in the Hockey East tournament (photo: Melissa Wade).

No. 13 Boston College (favorable odds for NCAAs): Like Maine, the Eagles can win a Hockey East semifinal and still fail to make the NCAA field. And also like Maine, BC has plenty of chances to get in without a single win this weekend. The Eagles have a better TUC record than Maine, but not as good as UNH, and Maine is currently winning the head-to-head comparison against BC, so turning that in BC’s favor would help its cause. Unfortunately for the Eagles, that can only happen by beating Maine in the semis, making a loss there a double whammy for Boston College.

WCHA

No. 1 Minnesota (lock for No. 1 seed): Minnesota head coach Don Lucia said earlier in the season that the Gophers’ first goal was to make the NCAA tournament, after which they could worry about a top seed. Minnesota no longer has to be concerned about either goal. The Gophers have been a lock to make the field of 16 for some time, and after last weekend’s sweep of Alaska-Anchorage, a No. 1 seed looks to be guaranteed. In fact, I can’t find a way to push the Gophers down lower than the second overall seed. Either way, Minnesota is likely still headed to Grand Forks as the top team in the West Regional.

No. 2 Wisconsin (lock for No. 1 seed): The Badgers seem to have recovered from their late-season swoon, sweeping Michigan Tech last weekend to join the Gophers in No. 1 seed lock-land. While two losses at the Final Five could push Wisconsin down a spot or two in the PWR, UW still gets a No. 1 seed, and it’s more likely that the Badgers and the Gophers will be the top two overall seeds in the NCAAs, with the easiest way for the Badgers to jump over the Gophers being a head-to-head win — doesn’t seem to matter where — in the WCHA Final Five. Alternatively, Wisconsin can also get past Minnesota for the top seed by winning more games in St. Paul this weekend, although that may depend on exactly which teams both teams face in the process, due to the impact on RPI. In any case, the Badgers are probably your top seed in Green Bay.

No. 8 North Dakota (near-lock for NCAAs): I say “near-lock” to mean “all cases except a rash of upset champions.” Every big-time upset champion — meaning a league tournament champion that would not otherwise have been at at-large bid — has the simple effect of taking away a slot from an at-large team. Thus, if Colgate wins the ECACHL title, it’s one less bid for an at-large team, since the Raiders cannot make the NCAAs otherwise. The only way North Dakota can fall out of the NCAAs is to lose two at the Final Five and then watch at least a couple of upset champions crowned. It also hurts the Fighting Sioux if a team other than Boston University wins Hockey East, since Maine, New Hampshire and Boston College would all be jockeying with UND for bubble positioning if the Sioux lose two this weekend. Still, add it all up and UND has perhaps a 5 or 10 percent chance of sitting out the NCAAs. Incidentally, on the other side of the coin, the Broadmoor Trophy and poor results by teams such as Miami, Michigan State and Boston University can move the Sioux as high as a No. 1 seed, so the gamut of possibilities is still out there for UND.

No. 10 Colorado College (near-lock for the NCAAs): The Tigers can be moved out of the field of 16, but it’s no easy task. With Colorado College’s WCHA season over after last weekend’s defeat by St. Cloud State, the good news for CC is that its PairWise credentials are essentially fixed, meaning that only unlikely outcomes in conference championships — e.g., Northern Michigan as CCHA champion, SCSU or Minnesota-Duluth as WCHA champions, and so forth — can push the Tigers out of the picture. Like other at-large hopefuls, CC also must beware the possibility laid out at the top of this article, the Bentley over Holy Cross scenario that could give Atlantic Hockey two bids. But still, the Tigers are in barring a rash of upsets this weekend.

No. 16 Denver (all but out?): I’ll say it right now: the collective efforts of the USCHO.com editorial staff cannot find a way for the Pioneers to make the NCAA tournament. Two losses to Minnesota-Duluth last weekend crippled DU’s RPI, and that, coupled with the Pioneers’ mediocre TUC record, appears to have killed any chance of an at-large bid. If anyone can find a scenario — no matter how unlikely — that puts Denver in the field using the 3-2-1 bonus, please let me know and I’ll publish it. But as of now, it seems that the Pioneers’ hope of an historic three-peat is DOA.