What's new
By:
USCHO Fan Forum

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

  • The USCHO Fan Forum has migrated to a new plaform, xenForo. Most of the function of the forum should work in familiar ways. Please note that you can switch between light and dark modes by clicking on the gear icon in the upper right of the main menu bar. We are hoping that this new platform will prove to be faster and more reliable. Please feel free to explore its features.

pairwise -- post jan 8th

Re: pairwise -- post jan 8th

JimDahl said:
Here's the underlying data on the remaining possible scenarios, broken down by outcome, that I summarized above (9 in, 7 competing for at-large):
http://test.collegehockeyranked.com/forecast/pwrtournament.php
Click to expand...

Thanks for the work you do on this. Always a pleasure to read.

Gotta admit, I miss the lack of real "bubble action" this year. Realistically, we're looking at only a couple of games that can change the field. Seedings are, of course, in flux, but the field itself is essentially set.
 
Re: pairwise -- post jan 8th

purpleinnebraska said:
Thanks for the work you do on this. Always a pleasure to read.

Gotta admit, I miss the lack of real "bubble action" this year. Realistically, we're looking at only a couple of games that can change the field. Seedings are, of course, in flux, but the field itself is essentially set.
Click to expand...

What lack are you talking about? Big Ten has 6 teams alive...only one of them already in so any of the other 5 winning moves the line. ECAC has 2 teams that can move the line (while Harvard isn't guaranteed of being in, I would think most fans of bubble teams are rooting for Harvard to eliminate St. Lawrence), Northeastern and UMD are hot but they both can be out if they lose Friday (and saturday for UMD). There aren't any crazy moves anymore with the removal of the TUC cliff, but I think we can all agree thats good.
 
Re: pairwise -- post jan 8th

Shirtless Guy said:
What lack are you talking about? Big Ten has 6 teams alive...only one of them already in so any of the other 5 winning moves the line. ECAC has 2 teams that can move the line (while Harvard isn't guaranteed of being in, I would think most fans of bubble teams are rooting for Harvard to eliminate St. Lawrence), Northeastern and UMD are hot but they both can be out if they lose Friday (and saturday for UMD). There aren't any crazy moves anymore with the removal of the TUC cliff, but I think we can all agree thats good.
Click to expand...

Of course until the tournaments are complete, you can always speculate about a Wisconsin making the tournament, but realistically, that isn't happening. But almost no one has to watch the results of any game but their own. Win, you're in. Lose, you're done (Or lose, and you were already in anyway.). And the teams you mention will all know their fates after their semifinal game. By Friday evening, the field will be set. That's simple and merit-based, but not nearly as fun from a "hoping for pure chaos" perspective.
 
Re: pairwise -- post jan 8th

purpleinnebraska said:
Of course until the tournaments are complete, you can always speculate about a Wisconsin making the tournament, but realistically, that isn't happening. But almost no one has to watch the results of any game but their own. Win, you're in. Lose, you're done (Or lose, and you were already in anyway.). And the teams you mention will all know their fates after their semifinal game. By Friday evening, the field will be set. That's simple and merit-based, but not nearly as fun from a "hoping for pure chaos" perspective.
Click to expand...
Do you really think PSU or OSU can't win the B1G? If St. Lawrence or Dartmouth win, the cut line could move 2 spots on the final day, but I do understand what you mean. The TUC cliff change has certainly flattened things.
 
Shirtless Guy said:
Do you really think PSU or OSU can't win the B1G? If St. Lawrence or Dartmouth win, the cut line could move 2 spots on the final day, but I do understand what you mean. The TUC cliff change has certainly flattened things.
Click to expand...

I tend to agree with him to the extent that this year no one can really play themselves into an at-large bid. We have about four teams sitting in at risk at large positions. They themselves have a chance to win an AQ spot, or they could get knocked out by someone else getting an AQ, but it's not like we have a bunch of bubble teams who can just play themselves into or out of at-large spots.
 
Re: pairwise -- post jan 8th

purpleinnebraska said:
Of course until the tournaments are complete, you can always speculate about a Wisconsin making the tournament, but realistically, that isn't happening. But almost no one has to watch the results of any game but their own. Win, you're in. Lose, you're done (Or lose, and you were already in anyway.). And the teams you mention will all know their fates after their semifinal game. By Friday evening, the field will be set. That's simple and merit-based, but not nearly as fun from a "hoping for pure chaos" perspective.
Click to expand...

To move the cut line all that has to happen in the B1G is for Minnesota - the #1 seed and regular season champion - to win the conference autobid. It doesn't have to be Wisconsin or Ohio State.

The field isn't going to be set after Friday night. Even then there will be enough variables to keep things interesting. It isn't as chaotic as it has been in the past, but it also isn't cut and dried.
 
Re: pairwise -- post jan 8th

The parley is a term I coined a few years ago to describe the confluence of events where conference autobids are won by teams outside the top 16 and thus move the cut line - and cause the elimination of a team.

This year the parley is:
Northeastern winning Hockey East
Minnesota-Duluth winning the NCHC
Dartmouth or St Lawrence winning the ECAC
Any team except Michigan Tech winning the WCHA
Any team except Michigan winning the B1G

The AHA is not involved in the parley.

The parley is why Yale is not yet safe. If the parley hits, the Bulldogs are knocked out. This is not Yale's first time at this particular rodeo.

In 2013 Yale was a mortal lock to make the tournament, and competing for a regional #1 seed. Then it lost to Union and Quinnipiac in the ECAC tournament and had to wait as results of the Hockey East and CCHA Championship games decided their fate. If either Boston University or Michigan won a conference championship that night Yale would be eliminated.

They came up short and Yale survived as the last team into the tournament. The Bulldogs then went on a tear and won the NCAA Championship in Pittsburgh.

Yale shouldn't need that kind of luck again to gain a spot.

This also is not new territory for another team of Bulldogs. In 2009, the University of Minnesota-Duluth was sitting at 18th in the PairWise and decidedly on the outside of the tournament. Further, they would have a gauntlet to run at the WCHA Final Five, playing the Thursday play-in game for a chance to play Friday for a chance to play Saturday. UMD beat the odds, won the Broadmoor Trophy and entered the tournament as the #7 overall seed.

The field for this season's tournament will again not be decided until the last games go final. Both the "east" and "west" #1 seeds (and thus favorable geographic locations) are still up for grabs and there will be plenty of movement within the field itself. Plus who knows...that parley might still be alive come Saturday...
 
Last edited:
Re: pairwise -- post jan 8th

ticapnews said:
The parley is a term I coined a few years ago to describe the confluence of events where conference autobids are won by teams outside the top 16 and thus move the cut line - and cause the elimination of a team.

This year the parley is:
Northeastern winning Hockey East
Minnesota-Duluth winning the NCHC
Dartmouth or St Lawrence winning the ECAC
Any team except Michigan Tech winning the WCHA
Any team except Michigan winning the B1G

The AHA is not involved in the parley.

The parley is why Yale is not yet safe. If the parley hits, the Bulldogs are knocked out. This is not Yale's first time at this particular rodeo.

In 2013 Yale was a mortal lock to make the tournament, and competing for a regional #1 seed. Then it lost to Union and Quinnipiac in the ECAC tournament and had to wait as results of the Hockey East and CCHA Championship games decided their fate. If either Boston University or Michigan won a conference championship that night Yale would be eliminated.

They came up short and Yale survived as the last team into the tournament. The Bulldogs then went on a tear and won the NCAA Championship in Pittsburgh.

Yale shouldn't need that kind of luck again to gain a spot.

This also is not new territory for another team of Bulldogs. In 2009, the University of Minnesota-Duluth was sitting at 18th in the PairWise and decidedly on the outside of the tournament. Further, they would have to a gauntlet to run at the WCHA Final Five, playing the Thursday play-in game for a chance to play Friday for a chance to play Saturday. UMD beat the odds, won the Broadmoor Trophy and entered the tournament as the #7 overall seed.

The field for this season's tournament will again not be decided until the last games go final. Both the "east" and "west" #1 seeds (and thus favorable geographic locations) are still up for grabs and there will be plenty of movement within the field itself. Plus who knows...that parley might still be alive come Saturday...
Click to expand...
I don't think Michigan Tech winning should be required in your parley? If they win, they would still bump Yale because they can't get higher than 13th, unless somehow them winning bumps Yale enough because MTU was on Yale's schedule?
 
Re: pairwise -- post jan 8th

Shirtless Guy said:
I don't think Michigan Tech winning should be required in your parley? If they win, they would still bump Yale because they can't get higher than 13th, unless somehow them winning bumps Yale enough because MTU was on Yale's schedule?
Click to expand...

Partially correct. Different teams have different parleys. You are correct in that for Yale the winner of the WCHA is irrelevant. The conference is getting an AQ and taking a spot. For other teams it matters a great deal as Michigan Tech is ranked high enough to lose the WCHA Final and still make the tournament.

For Notre Dame, the parley is all but one of the AQ going to upset winners (Yale gets in, ND does not). The trick is that there is a moving part since Harvard can theoretically also knock out ND by winning the ECAC SF.
 
Re: pairwise -- post jan 8th

I really want to spend some time developing a tool to rank every team based on importance to a given team. So for Team A rankings if Team 27 and Team 28 play, Team A would want Team 27 to win. To help in what you're doing ticapnews. Granted you also have to know the rankings for the other teams you're trying to flip comparisons for...and as you flip games, it changes the calc because so much is based on weighted games played throughout the calculation. I tried to do it for MTU this year, and came close but it still wasn't quite right, or method of proving it wasn't quite right.

With the RPI 25/21/54 its very hard to predict what is important to a team other than that team winning its own games. Obviously conference opponents are very important, but most of the year, they play each other...
 
Re: pairwise -- post jan 8th

Atlantic Hockey Semifinal #1: Robert Morris def. Army
Atlantic Hockey Semifinal #2: Air Force def. RIT
Atlantic Hockey Championship: Robert Morris def. Air Force
Big Ten Semifinal #1: Ohio State def. Minnesota
Big Ten Semifinal #2: Michigan def. Penn State
Big Ten Championship: Michigan def. Ohio State
Big Ten Play-in #1: Penn State def. Wisconsin
Big Ten Play-in #2: Ohio State def. Michigan State
ECAC Semifinal #1: Dartmouth def. Quinnipiac
ECAC Semifinal #2: Harvard def. St. Lawrence
ECAC Championship: Dartmouth def. Harvard
Hockey East Semifinal #1: Boston College def. Northeastern
Hockey East Semifinal #2: Providence def. Mass.-Lowell
Hockey East Championship: Providence def. Boston College
NCHC Semifinal #1: North Dakota def. Minnesota-Duluth
NCHC Semifinal #2: St. Cloud State def. Denver
NCHC Championship: St. Cloud State def. North Dakota
NCHC Consolation: Denver def. Minnesota-Duluth
WCHA Semifinal #1: Michigan Tech def. Ferris State
WCHA Semifinal #2: Minnesota State def. Bowling Green
WCHA Championship: Michigan Tech def. Minnesota State

Because I hadn't seen it mentioned yet here, this scenario gets 5 ECAC teams in. Cornell last in at 15.
 
Re: pairwise -- post jan 8th

For those that would like another nice <a href=http://www.collegehockeynews.com/ratings/probabilityMatrix.php>reference</a>. This link shows the College Hockey News breakdown of odds making the tournament and odds at being a given overall seed. For example, this indicates based on 20,000 Monte Carlo simulations BC cannot achieve a overall #1 seed but has a 36% chance of getting a #1 seed. Or that Yale, ND, and Harvard are all virtual locks. Northeastern, Michigan Tech, and UMD are the teams on the bubble as far as having a chance to get an at large bid without winning their tourney.
 
hockeyplayer1015 said:
Atlantic Hockey Semifinal #1: Robert Morris def. Army
Atlantic Hockey Semifinal #2: Air Force def. RIT
Atlantic Hockey Championship: Robert Morris def. Air Force
Big Ten Semifinal #1: Ohio State def. Minnesota
Big Ten Semifinal #2: Michigan def. Penn State
Big Ten Championship: Michigan def. Ohio State
Big Ten Play-in #1: Penn State def. Wisconsin
Big Ten Play-in #2: Ohio State def. Michigan State
ECAC Semifinal #1: Dartmouth def. Quinnipiac
ECAC Semifinal #2: Harvard def. St. Lawrence
ECAC Championship: Dartmouth def. Harvard
Hockey East Semifinal #1: Boston College def. Northeastern
Hockey East Semifinal #2: Providence def. Mass.-Lowell
Hockey East Championship: Providence def. Boston College
NCHC Semifinal #1: North Dakota def. Minnesota-Duluth
NCHC Semifinal #2: St. Cloud State def. Denver
NCHC Championship: St. Cloud State def. North Dakota
NCHC Consolation: Denver def. Minnesota-Duluth
WCHA Semifinal #1: Michigan Tech def. Ferris State
WCHA Semifinal #2: Minnesota State def. Bowling Green
WCHA Championship: Michigan Tech def. Minnesota State

Because I hadn't seen it mentioned yet here, this scenario gets 5 ECAC teams in. Cornell last in at 15.
Click to expand...

Nice find. Here is the <a href=http://pwp.uscho.com/rankings/pairwise-predictor/?uniq=pwp_56e8b34186b27>link</a> for those interested.

Correction: Jim Dahl said this is 1 in 1700 scenarios based on his simulations. I previously thought Jim missed it because his table says 0%, just a very small percent chance.

On the flipside, <a href=http://pwp.uscho.com/rankings/pairwise-predictor/?uniq=pwp_56e8b52986255>here</a> is an example (multiple exist) of only two ECAC teams making the tourney.
 
Last edited:
ticapnews said:
St Cloud to 5th
Am still trying to get BC to #1. Have gotten them within .0003 RPI but have not gotten them over the hump yet.
Click to expand...

I stand corrected to my previous post. BC as the #1 overall: <a href=http://pwp.uscho.com/rankings/pairwise-predictor/?uniq=pwp_56e8beb4c1696>http://pwp.uscho.com/rankings/pairwise-predictor/?uniq=pwp_56e8beb4c1696<a>
 
Re: pairwise -- post jan 8th

In the tournament:
Boston College
Boston University
Denver
Massachusetts-Lowell
Michigan
North Dakota
Providence
Quinnipiac
St Cloud

Need AQ:
Air Force
Army
Bowling Green
Dartmouth
Ferris State
Minnesota
Minnesota State
Ohio State
Penn State
RIT
Robert Morris
St. Lawrence

Alive, Still Playing
Harvard
Minnesota-Duluth
Michigan Tech
Northeastern

Not playing, still alive:
Cornell
Notre Dame
Yale

The scenario that gets Boston College to #1 overall is still active. Keys today: Quinnipiac upset by Dartmouth, Minnesota upended by Ohio State.
 
Re: pairwise -- post jan 8th

Shirtless Guy said:
Check this one out...a 3-way tie for last team in (at least to 4 decimals for RPI):
http://goo.gl/bcQDvH
How much would that suck to be the team left out in a virtual tie with 2 other teams?
Click to expand...

ran that scenario in my excel file and got RPI values as follows:
Michigan Tech - 0.535734
Minnesota-Duluth - 0.535725
Cornell - 0.535719

Do we really trust this method enough to rank MTU first by RPI in that scenario? Granted MTU has a better COp record against both UMD and Cornell but we're talking about 2 common opps with UMD (BSU/NMU) and 1 for Cornell (Yale).
 
You must log in or register to reply here.
Back
Top