Historically, PairWise’s top 3 entering league quarterfinals are solid bets to stay as No. 1 seeds

We’re onto the quarterfinals in every conference except the Big Ten. Below are the PairWise Rankings through games of March 6, and as you can see, there are a few changes from last week, and the projected cut line has moved.

1. North Dakota*
2. Quinnipiac*
3. St. Cloud State
4. Providence*
5. Boston College
6t. Denver
6t. Yale
8t. Michigan*
8t. Boston University
10. Notre Dame
11. Harvard
12. Massachusetts-Lowell
13. Minnesota-Duluth
14. Michigan Tech*
15. Omaha
— projected cut line —
16. Cornell
— average position for outsider to make the tournament
17t. Minnesota
17t. Penn State
19. Northeastern
20. Clarkson
21. Robert Morris*
22t. Rensselaer
22t. Miami
22t. Minnesota State
25. St. Lawrence
— lowest any team has been ranked and still qualified —
26. Bowling Green
27. Dartmouth
28. Air Force
29t. Holy Cross
29t. Union
31. Bemidji State
32. Ohio State
33. Northern Michigan
34. Ferris State
35t. Mercyhurst
35t. Vermont
37. Merrimack
38t. New Hampshire
38t. Western Michigan
40t. RIT
40t. Wisconsin
42t. Michigan State
42t. Colgate
44. Lake Superior
45t. Connecticut
45t. Army
47. Bentley
48. Brown
49t. Sacred Heart
49t. Massachusetts
51. Canisius
52. Alaska-Anchorage
53t. Colorado College
53t. Alaska
55. Maine
56. Princeton
57. Alabama-Huntsville
58. Niagara
59. Arizona State
60. American International

As Michigan Tech moved up in the rankings, the cut line is dropped to 15. Minnesota is the top team in the B1G, but I deal with the PairWise, not conference standings. For those, check out Jayson Moy’s Bracketology. Since Michigan is above the cut line, I assume they will win the conference autobid. As Minnesota is the host in St. Paul and it sets up a potential North Dakota vs. Minnesota game in the first round, there are going to be strong opinions. However, for the purposes of this writing, Minnesota — or any team below the cut line — winning a conference autobid is considered an upset.

With that out of the way, onto the historical analysis.

In the 13 previous seasons:

• The team ranked No. 1 overall as the quarterfinals began finished No. 1 overall seven times and never lower than No. 3 overall.

• The No. 2 overall team finished No. 1 five times and also has never finished lower than No. 3 overall.

• The No. 3 team finished first only once (Boston College in 2005) and in 10 seasons was a No. 1 regional seed.

• The No. 4 overall team has remained in the top four in only five seasons, and never higher than third.

• Seven times, the overall No. 5 team has moved into the top four.

Overall, the top four will remain in the top four 77 percent of the time. Put another way, the top three teams typically remain and the No. 4 and 5 teams trade places.

The worst performance by any of the top four came in 2012 when UMass-Lowell dropped from No. 3 overall all the way to No. 12 after being upset by Providence in the quarterfinals of the Hockey East tournament.

Conversely, the 2013 Notre Dame squad rose from No. 14 to finish No. 4 overall. 2012 is the only season two of the top four teams finished worse than fourth as Minnesota-Duluth joined UMass-Lowell in the exodus. No. 13 North Dakota leapt up the rankings after sweeping through the WCHA Final Five to finish fourth.

The top nine teams are statistically safe. The 2005 Dartmouth Big Green squad is the only team to drop from No. 10 entirely out of the tournament after losing to Vermont in the ECAC Hockey playoffs. Teams that are ranked 1-12 now make the tournament a whopping 94 percent of the time. Overall, 90 percent of teams above the projected cut line remain above the cut line. Only the 2010 Michigan Wolverines and 2013 Wisconsin Badgers have jumped from No. 20 or worse to land in the tournament. The bubble is getting smaller.

Looking at the schedule coming up, the opportunity still exists for teams like Clarkson and Miami to play their way in, but realistically, people should look to Cornell, Minnesota and Penn State to see a team move above that cut line.

54 COMMENTS

  1. The Sioux blasted Michigan Tech 8-0 tonight and Frattin and Gregoire both had Hat tricks for UND. Say hello to the new leading scorer in the country.

    • must have missed the part where the drunk fratboy had 65 points on the season. he can’t score against anyone except for horrible competition.

  2. The Sioux blasted Michigan Tech 8-0 tonight and Frattin and Gregoire both had Hat tricks for UND. Say hello to the new leading scorer in the country.

    • must have missed the part where the drunk fratboy had 65 points on the season. he can’t score against anyone except for horrible competition.

    • The drama associated with his departure from the team last year will be used against him. Not saying I agree with it, but it is what will happen.

    • The drama associated with his departure from the team last year will be used against him. Not saying I agree with it, but it is what will happen.

  3. I think Frattin deserves it as well. But it’s so hard to use anything against Tech as a measuring stick. The game last night looked like an exhibition between Ivan Drago and Apollo Creed. 2 completely different levels of competition.

  4. I think Frattin deserves it as well. But it’s so hard to use anything against Tech as a measuring stick. The game last night looked like an exhibition between Ivan Drago and Apollo Creed. 2 completely different levels of competition.

  5. I think if people look at fratts situation with an open mind he won’t seem so bad… the maple leads offered him a contract the day after he was dismissed… he goes home and does hard labor instead so he can come back and wear the white n green… the kid is currently in treatment and college and is scoring at an absurd pace… maybe give him some credit for overcoming adversity and fighting his way back…

  6. I think if people look at fratts situation with an open mind he won’t seem so bad… the maple leads offered him a contract the day after he was dismissed… he goes home and does hard labor instead so he can come back and wear the white n green… the kid is currently in treatment and college and is scoring at an absurd pace… maybe give him some credit for overcoming adversity and fighting his way back…

  7. I actually feel better about the Huskies this weekend seeing these predictions. It’ll be a good game. Home ice will give the Dogs a boost.

  8. I actually feel better about the Huskies this weekend seeing these predictions. It’ll be a good game. Home ice will give the Dogs a boost.

  9. The Pairwise needs to be improved or just used by committee. It is certainly imperfect and every year a deserving team is left out in favor of an obviously undeserving team.

    • Any suggestions on how PWR can be improved? What do you mean by “just used by committee”? If you think PWR is bad, I remember the “old” days when totally biased “experts” chose who made the tourney. The problem with the “committee” just getting to select, is that they NEVER see all the decent teams play during the year. Eastern “experts” follow eastern teams, Western “experts” follow western teams, and they both are biased.

      • I don’t know that a formula can properly work. The “old way” with just a committee had it’s flaws, but I think the Pairwise is also flawed. Let the Pairwise pick the top 10 then have a committee look at the rest after the automatics go to the weaker leagues with champions nowhere near the Pairwise top 16. Seems senseless to go strictly by the pairwise and then allow teams not in the top 25. Either take the top 16 and too bad for the “league champions” that are weak, OR use the Pairwise and a committee. For another thing, I think when a league has more than one team in the tournament, all teams should be on a path NOT to play other teams from the same league. A strong league with 4 teams should have a chance to have them all in the Frozen Four, not knocking each other off…….

        • I am totally opposed to any committee deciding who gets in, for the biased reasons I already stated. As long as the computerized algorithm used pertains to every team equally, I fid it hard to complain about. As for your other thing, it is physically impossible for a league that has five teams (HEA this year, NCHC last year) to not play each other in a regional.

          • I guess my point is this: IF you have faith in the Pairwise, then ONLY the top 16 teams belong, and additionally, it should be bracketed 1 vs. 16 etc. My point about 4 teams was that unless/until one league gets MORE than 4 teams, the 4 teams or less should be spread out in different regionals so a good league isn’t eliminating their own teams until it cannot be avoided. As far as the Pairwise rankings, I will forever believe that 6 teams from one league, one with a 19-18 w/l record, was ridiculous, so I find the computations flawed. How BU is three teams higher than Lowell, when they had the same exact league records and BU lost the tie breaker, is beyond me. They may prove the better team this weekend, and good for them if they do, but they don’t deserve to be ranked higher.

          • BU is ranked higher than UML due to SOS. BU SOS is 11, UML is 35. That is a very large difference. I have much more faith in PWR than humans, and totally agree with you that they should be bracketed 1/8/9/16, 2/7/10/15, etc. The only exceptions should be host teams get to stay home and #1 seeds should have priority to play closest to their place. It is pure garbage that the NCAA uses attendance as a reason to place teams where they don’t belong.

          • “I have much more faith in PWR than humans…”.
            Ummm, I’m pretty sure that the PWR was designed by at least one or more humans, and programmed by at least one other human. Unless you’re trying to say that an AI designed it, then programmed itself. Geez, maybe SkyNet’s gone online earlier than the Terminator movies predicted.
            I think I understand what you’re really trying to say, but you just haven’t been able to say it yet. I won’t say what that is, since I don’t want to put words in you mouth, plus I get in enough trouble trying to express my own thoughts…

          • Please don’t play dumb. You know exactly what I meant. Unless the Pairwise programming changes, depending on which team is being assessed, every team has an equal chance of being evaluated by the same standards. Human “experts” cannot possible evaluate every team equally, they do not see every game played. Does that explain what I meant to your satisfaction?

          • Heh. Of course I knew what you meant, and I said so, didn’t I? It ain’t my fault that you don’t trust humans, but are totes fine with computer programming created by the very humans that you distrust so.
            But enough of that. I take it from your timing that you might have attended the DU-UNO game. Following the out-of-town scores, it seems that UNO kept things close, with DU pulling away at the end. Is that how it went down, and does UNO have a realistic shot at making this a series?

          • You are correct, I have season tickets and attend all DU home games. UNO outplayed us for all but 15 minutes of the game. Our goalie made some great saves, and we scored on two breakaways and an ENG. Heinan continues to be a stud, plays well every game. UNO are still a very good team and will definitely make it tough tomorrow night. Ortega and Guentzel played well all three recent games. Very hard to beat a decent team four in a row. Unfortunately, they probably have to beat us the next 2 games to get into regionals.

          • If I’m reading the current PWR right (I think it’s up to date now), I’m thinking UNO will probably have to do more than that. The teams that they’d be capable of passing with just winning this weekend are all leading in their series (NE, Duluth, Harvard, Lowell), and if they win the pecking order will stay the same. I didn’t include Tech because if they lose to AK, they’d drop out of the top 16, then the WCHA autobid becomes the 15 PWR team and UNO still gets bumped. Finally, the three teams above UNO that are actually down in their series (BU, Yale, Domers at 8/9/10) are probably too far ahead for UNO to catch them this week, meaning UNO would have to win at least once next weekend. Sound about right?

          • What you said sounds logical and makes sense. In the past playing the #1 team, which they would if they went to our Frozen Faceoff, does not drop you with a loss. Beating DU twice would probably put them at 13/14, and they might stay there even if they lose to UND. I agree that they would have to win the consolation game to stay at 14, or even move up. Don’t forget, they would play another high ranked team in the consolation game. BTW, Pairwise is updated about 10 minutes after the score becomes final. That is one of the benefits of computer rankings, humans don’t have to do the math. :-)

          • Something working in UNO’s favor right now is that teams below their position like Cornell and Clarkson are facing elimination tonight, but that still leaves Penn State, St. Lawrence and Mankato in position to pass UNO. I think Minnesota has crapped the bed once too often – at this point I think an autobid is their only chance. Of course, a second autobid from outside the top 16 probably dooms UNO anyway unless UMD, or Northeastern loses. If Tech loses, then that means there could be a 3rd autobid from outside, and I think that definitely dooms UNO unless nearly everyone above them loses. Enjoy your game tonight.

          • Think you are correct about the teams facing elimination, and Gophers are doomed without autobid. If UMD beats Miami tonight they are probably safe. Big NCHC advantage this year is that all teams at Frozen Faceoff probably will be ranked high enough that a split ensures keeping PWR ranking. Thanks, I will enjoy the game tonight and hope the next time I see DU will be Friday on CBSSN.

          • Please explain how BU LOSES to Lowell last night and MOVE UP from t-9 to 8…. When Football did computer rankings, teams didn’t move up by losing late in the season. System is flawed. Perhaps a good tool to use, but flawed.

          • i know this is odd and confusing but i have seen it happen – the only explanation is that there are so many factors being used in the formula(s) and so many games each night that contribute to those factors – i suppose the only thing that can be done is change the formula to avoid this but i am not sure if they know what to change – there must be some brainiac they use to derive the formulas – i tried making my own one year and i changed it multiple times but never saw a credible difference – it was fun and insightful – try it

  10. The Pairwise needs to be improved or just used by committee. It is certainly imperfect and every year a deserving team is left out in favor of an obviously undeserving team.

    • Any suggestions on how PWR can be improved? What do you mean by “just used by committee”? If you think PWR is bad, I remember the “old” days when totally biased “experts” chose who made the tourney. The problem with the “committee” just getting to select, is that they NEVER see all the decent teams play during the year. Eastern “experts” follow eastern teams, Western “experts” follow western teams, and they both are biased.

      • I don’t know that a formula can properly work. The “old way” with just a committee had it’s flaws, but I think the Pairwise is also flawed. Let the Pairwise pick the top 10 then have a committee look at the rest after the automatics go to the weaker leagues with champions nowhere near the Pairwise top 16. Seems senseless to go strictly by the pairwise and then allow teams not in the top 25. Either take the top 16 and too bad for the “league champions” that are weak, OR use the Pairwise and a committee. For another thing, I think when a league has more than one team in the tournament, all teams should be on a path NOT to play other teams from the same league. A strong league with 4 teams should have a chance to have them all in the Frozen Four, not knocking each other off…….

        • I am totally opposed to any committee deciding who gets in, for the biased reasons I already stated. Your sentence about “weak” league champions proves my assertion, who gets to determine which teams are “weak”? As long as the computerized algorithm used pertains to every team equally, I fid it hard to complain about. As for your other thing, it is physically impossible for a league that has five teams (HEA this year, NCHC last year) to not play each other in a regional.

          • I guess my point is this: IF you have faith in the Pairwise, then ONLY the top 16 teams belong, and additionally, it should be bracketed 1 vs. 16 etc. My point about 4 teams was that unless/until one league gets MORE than 4 teams, the 4 teams or less should be spread out in different regionals so a good league isn’t eliminating their own teams until it cannot be avoided. As far as the Pairwise rankings, I will forever believe that 6 teams from one league, one with a 19-18 w/l record, was ridiculous, so I find the computations flawed. How BU is three teams higher than Lowell, when they had the same exact league records and BU lost the tie breaker, is beyond me. They may prove the better team this weekend, and good for them if they do, but they don’t deserve to be ranked higher.

          • BU is ranked higher than UML due to SOS. BU SOS is 11, UML is 35. That is a very large difference. I have much more faith in PWR than humans, and totally agree with you that they should be bracketed 1/8/9/16, 2/7/10/15, etc. You cannot merely take the top 16, it would kill other conferences and be terrible for college hockey overall. The only exceptions to the seedings should be host teams get to stay home and #1 seeds should have priority to play closest to their place. It is pure garbage that the NCAA uses attendance as a reason to place teams where they don’t belong.

          • “I have much more faith in PWR than humans…”.
            Ummm, I’m pretty sure that the PWR was designed by at least one or more humans, and programmed by at least one other human. Unless you’re trying to say that an AI designed it, then programmed itself. Geez, maybe SkyNet’s gone online earlier than the Terminator movies predicted.
            I think I understand what you’re really trying to say, but you just haven’t been able to say it yet. I won’t say what that is, since I don’t want to put words in you mouth, plus I get in enough trouble trying to express my own thoughts…

          • Please don’t play dumb. You know exactly what I meant. Unless the Pairwise programming changes, depending on which team is being assessed, every team has an equal chance of being evaluated by the same standards. Human “experts” cannot possible evaluate every team equally, they do not see every game played. Does that explain what I meant to your satisfaction?

          • Heh. Of course I knew what you meant, and I said so, didn’t I? It ain’t my fault that you don’t trust humans, but are totes fine with computer programming created by the very humans that you distrust so.
            But enough of that. I take it from your timing that you might have attended the DU-UNO game. Following the out-of-town scores, it seems that UNO kept things close, with DU pulling away at the end. Is that how it went down, and does UNO have a realistic shot at making this a series?

          • You are correct, I have season tickets and attend all DU home games. UNO outplayed us for all but 15 minutes of the game. Our goalie made some great saves, and we scored on two breakaways and an ENG. Heinan continues to be a stud, plays well every game. UNO are still a very good team and will definitely make it tough tomorrow night. Ortega and Guentzel played well all three recent games. Very hard to beat a decent team four in a row. Unfortunately, they probably have to beat us the next 2 games to get into regionals.

          • If I’m reading the current PWR right (I think it’s up to date now), I’m thinking UNO will probably have to do more than that. The teams that they’d be capable of passing with just winning this weekend are all leading in their series (NE, Duluth, Harvard, Lowell), and if they win the pecking order will stay the same. I didn’t include Tech because if they lose to AK, they’d drop out of the top 16, then the WCHA autobid becomes the 15 PWR team and UNO still gets bumped. Finally, the three teams above UNO that are actually down in their series (BU, Yale, Domers at 8/9/10) are probably too far ahead for UNO to catch them this week, meaning UNO would have to win at least once next weekend. Sound about right?

          • What you said sounds logical and makes sense. In the past playing the #1 team, which they would if they went to our Frozen Faceoff, does not drop you with a loss. Beating DU twice would probably put them at 13/14, and they might stay there even if they lose to UND. I agree that they would have to win the consolation game to stay at 14, or even move up. Don’t forget, they would play another high ranked team in the consolation game. BTW, Pairwise is updated about 10 minutes after the score becomes final. That is one of the benefits of computer rankings, humans don’t have to do the math. :-)

          • Something working in UNO’s favor right now is that teams below their position like Cornell and Clarkson are facing elimination tonight, but that still leaves Penn State, St. Lawrence and Mankato in position to pass UNO. I think Minnesota has crapped the bed once too often – at this point I think an autobid is their only chance. Of course, a second autobid from outside the top 16 probably dooms UNO anyway unless UMD, or Northeastern loses. If Tech loses, then that means there could be a 3rd autobid from outside, and I think that definitely dooms UNO unless nearly everyone above them loses. Enjoy your game tonight.

          • Think you are correct about the teams facing elimination, and Gophers are doomed without autobid. If UMD beats Miami tonight they are probably safe. Big NCHC advantage this year is that all teams at Frozen Faceoff probably will be ranked high enough that a split ensures keeping PWR ranking. Thanks, I will enjoy the game tonight and hope the next time I see DU will be Friday on CBSSN.

          • Please explain how BU LOSES to Lowell last night and MOVE UP from t-9 to 8…. When Football did computer rankings, teams didn’t move up by losing late in the season. System is flawed. Perhaps a good tool to use, but flawed.

          • i know this is odd and confusing but i have seen it happen – the only explanation is that there are so many factors being used in the formula(s) and so many games each night that contribute to those factors – i suppose the only thing that can be done is change the formula to avoid this but i am not sure if they know what to change – there must be some brainiac they use to derive the formulas – i tried making my own one year and i changed it multiple times but never saw a credible difference – it was fun and insightful – try it

  11. If last season has taught us anything it is that sometimes seeding does not matter. The Providence College Friars entered the playoffs last season having lost in the quarter- finals, didn’t make it to the Hockey East tournament and were a #4 seed in the NCAAS. They are now defending Frozen Four Champions.

  12. If last season has taught us anything it is that sometimes seeding does not matter. The Providence College Friars entered the playoffs last season having lost in the quarter- finals, didn’t make it to the Hockey East tournament and were a #4 seed in the NCAAs. They are now defending Frozen Four Champions.

LEAVE A REPLY

Please enter your comment!
Please enter your name here