URL for this frameset: http://slack.net/~whelan/tbrw/tbrw.cgi?1999/pairwise.990315.shtml
To see how the final pairwise numbers might look, try the conference tournament bracket or non-tables "what if" interface to the interactive "You Are The Committee" script.
In just six days, the NCAA selection committee will seed the men's division I hockey tournament. With only 19 games left to be played, let's see how the Tournament selection procedure plays itself out pending those results. First of all, we know for certain four of the teams who will be in the tournament: New Hampshire, Clarkson, Michigan State and North Dakota each receive an automatic bid for winning the regular season titles in their respective conferences. Up to four more teams will receive automatic bids for winning their conference tournaments, and the remaining four to eight at large bids will be given out on the basis of pairwise comparisons among teams that finish with Division I records at or above .500. As of now, those comparisons look like this (with US College Hockey Online down at the moment, I have supplemented their Division I Composite Schedule by adding this past weekend's results by hand):
Team PWR RPI Comparisons Won 1 North Dakota 21 .647 NHMeMSCCCkDUQnBCSLOSMiNMPnNtRPCgMkCtNiPvHC 2 New Hampshire 20 .633 __ MeMSCCCkDUQnBCSLOSMiNMPnNtRPCgMkCtNiPvHC 3 Maine 19 .616 ____ MSCCCkDUQnBCSLOSMiNMPnNtRPCgMkCtNiPvHC 4 Mich State 18 .603 ______ CCCkDUQnBCSLOSMiNMPnNtRPCgMkCtNiPvHC 5 CO College 17 .586 ________ CkDUQnBCSLOSMiNMPnNtRPCgMkCtNiPvHC 6 Clarkson 16 .584 __________ DUQnBCSLOSMiNMPnNtRPCgMkCtNiPvHC 7 Denver U 15 .558 ____________ QnBCSLOSMiNMPnNtRPCgMkCtNiPvHC 8 Quinnipiac 13 .548 ______________ BC__OSMiNMPnNtRPCgMkCtNiPvHC 9 Boston Coll 13 .581 ________________ SLOSMiNMPnNtRPCgMkCtNiPvHC 10 St Lawrence 12 .555 ______________Qn__ OS__NMPnNtRPCgMkCtNiPvHC 11 Ohio State 11 .534 ____________________ MiNMPnNtRPCgMkCtNiPvHC 12 Michigan 9 .557 __________________SL__ NM__NtRPCgMkCt__PvHC 13 Northern Mich 9 .539 ________________________ PnNtRPCgMkCtNiPvHC 14 Princeton 8 .534 ______________________Mi__ __RPCgMkCtNiPvHC 15 Notre Dame 6 .537 __________________________Pn RPCgMkCt__Pv__ 16 RPI 6 .533 ______________________________ CgMkCtNiPvHC 17 Colgate 5 .530 ________________________________ MkCtNiPvHC 18 MSU-Mankato 4 .528 __________________________________ CtNiPvHC 19 Connecticut 3 .527 ____________________________________ NiPvHC 20 Niagara 3 .480 ______________________Mi____Nt________ __HC 21 Providence 2 .513 ______________________________________Ni HC 22 Holy Cross 1 .493 ____________________________Nt____________
(In addition, Minnesota or St. Cloud State could also qualify for the NCAAs by winning the WCHA Final Five.) Before charging ahead into the awarding of at-large bids, we must note that the ratings percentage index upon which the pairwise comparisons are largely based, has a weakness which is very significant this year: it does not judge a teams' strength of schedule accurately when that team's opponents have themselves played weak schedules. Since the six division I members of the new MAAC conference play 20 games each against each other, plus a few non-conference games against Division I independents, a team like Quinnipiac can rack up a high winning percentage against weak competition without the weakness of their schedule being reflected in the RPI. Anticipating this, the selection committee, as reported in the NCAA News, "noted that it reserves the right to evaluate each team based on the relative strength of their respective conference." The best way to gauge that relative strength is via the conferences' performance against the four Division I independents:
vs Indies vs Army vs Niagara vs AFA vs Mankato Avg RPI PF-PA Pct PF-PA Pct PF-PA Pct PF-PA Pct PF-PA Pct HE .525 14- 2 .875 12-0 1.000 0- 2 .000 2-0 1.000 0- 0 .--- WCHA .504 32-10 .762 0-0 .--- 0- 0 .--- 10-0 1.000 22-10 .688 CCHA .504 5- 5 .500 0-0 .--- 2- 4 .333 0-0 .--- 3- 1 .750 ECAC .496 28-10 .737 10-0 1.000 10-10 .500 2-0 1.000 6- 0 1.000 MAAC .453 8-22 .267 5-5 .500 0- 4 .000 3-5 .350 0- 8 .000
MSU-Mankato's surprise victory over North Dakota in game one of the WCHA quintafinal series makes the Mavericks look a bit stronger, but since no MAAC team actually beat them, it does little to change the conclusion that the MAAC has not reached competitive equity and it is reasonable to assume that the committee will exclude Quinnipiac, UConn and Holy Cross from consideration for at-large bids.
Maine, Colorado College, Denver, and Boston College all win comparisons with all of the remaining teams, and are thus easy choices for at-large bids. The awarding of the remaining bids is very tricky, and could be done in different ways with at least two different results. Here are the teams still in contention for those bids:
1 St Lawrence 9 .555 OS__NMPnNtRPCgMkNiPv 2 Ohio State 9 .534 __ MiNMPnNtRPCgMkNiPv 3 Michigan 7 .557 SL__ NM__NtRPCgMk__Pv 4 Northern Mich 7 .539 ______ PnNtRPCgMkNiPv 5 Princeton 6 .534 ____Mi__ __RPCgMkNiPv 6 Notre Dame 5 .537 ________Pn RPCgMk__Pv 7 RPI 4 .533 ____________ CgMkNiPv 8 Colgate 3 .530 ______________ MkNiPv 9 MSU-Mankato 2 .528 ________________ NiPv 10 Niagara 2 .480 ____Mi____Nt______ __ 11 Providence 1 .513 __________________Ni
The committee is supposed to compare teams which are "on the bubble" but the question is how to define that set of teams. However you slice it, Princeton will be a bubble team, which means that SLU and OSU will enter the tournament ahead of Michigan. Using the algorithm of the "automatic" button on my "You Are The Committee" script, which removes teams from the top and/or bottom of the table and then recalculates the number of comparisons won, we would remove Providence, Mankato, Colgate, and RPI from contention. At each turn, the bottom team has won a comparison only with Niagara, while the Purple Eagles have won comparisons with both Notre Dame and Michigan. This leaves us with
1 Northern Mich 3 .539 PnNi__Nt 2 Princeton 2 .534 __ NiMi__ 3 Niagara 2 .480 ____ MiNt 4 Michigan 2 .557 NM____ Nt 5 Notre Dame 1 .537 __Pn____
Notre Dame is dropped off the bottom of this bubble, and NMU and Princeton have won two comparisons (out of three) each with the remaining teams.
If, on the other hand, the committee leaves out Niagara, who lose comparisons to four teams directly below our ultimate bubble, we find ourselves deciding among the following:
1 Michigan (C) 2 .557 NMNt__ 2 Northern Mich (C) 2 .539 __ NtPn 3 Notre Dame (C) 1 .537 ____ Pn 4 Princeton (E) 1 .534 Mi____
which would put Michigan in the tournament instead of Princeton. We can't tell for sure what the committee would actually do, but if I had to guess I'd say that they would not think of Niagara, who win comparisons only with two teams (albeit obvious bubble teams) as themselves on the bubble. So let's proceed assuming that Michigan is in the tournament. That leaves us with seven Western teams and only five from the East, so we declare Northern Michigan, the lowest-rated Western team, to be honorary Easterners:
West East 1 North Dakota 5 .647 MSCCDUOSMi | 1 New Hampshire 5 .633 MeCkBCSLNM 2 Mich State 4 .603 CCDUOSMi | 2 Maine 4 .616 CkBCSLNM 3 CO College 3 .586 __ DUOSMi | 3 Clarkson 3 .584 __ BCSLNM 4 Denver U 2 .558 ____ OSMi | 4 Boston Coll 2 .581 ____ SLNM 5 Ohio State 1 .534 ______ Mi | 5 St Lawrence 1 .555 ______ NM 6 Michigan 0 .557 ________ | 6 Northern Mich 0 .539 ________
Both regions are nicely ranked by the pairwise comparisons. North Dakota and Michigan State are in line for the two Western byes, with New Hampshire and Maine in the East, although if Clarkson wins the ECAC tournament, they will receive an automatic bye. We need to swap the bottom two teams from each region, but in each case three of the top four teams come from the same conference, which leads to potential intraconference matchups in the second round. With the East Regionals being held in Worcester, Massachusetts, it seems pretty safe that attendance considerations will lead the NCAA to keep BC there anyway, and a possible intra-conference matchup in the West is inevitable with seven Western teams in the tourney. Going strictly by the numbers, we get the following teams in the regionals:
West East 1 North Dakota (W) 1 .647 MS | 1 New Hampshire (H) 1 .633 Me 2 Mich State (C) 0 .603 | 2 Maine (H) 0 .616 3 CO College (W) 3 .586 DUSLNM | 3 Clarkson (E) 3 .584 BCOSMi 4 Denver U (W) 2 .558 SLNM | 4 Boston Coll (H) 2 .581 OSMi 5 St Lawrence (E) 1 .555 __ NM | 5 Ohio State (C) 1 .534 __ Mi 6 Northern Mich (C) 0 .539 ____ | 6 Michigan (C) 0 .557 ____
The question here is whether attendance considerations would lead to NMU or DU trading places with Michigan or OSU. DU is in the WCHA along with host school Wisconsin, but NMU was recently in that league as well, and of course the other two CCHA schools in question are in the Big Ten. Very tentatively, though, let's leave the regions as they are. There is one avoidable second-round CCHA matchup between MSU and NMU, so we swap Northern Michigan and SLU, also swapping the two Colorado teams to preserve first-round pairings, and obtain
5W Northern Mich (C) 6E Michigan (C) 4W CO College (W) 3E Clarkson (E) 1W North Dakota (W) --+--2E Maine (H) | 2W Mich State (C) --+--1E New Hampshire (H) 3W Denver U (W) 4E Boston Coll (H) 6W St Lawrence (E) 5E Ohio State (C)
You can also see a detailed accounting of all the pairwise comparisons.