Women's D1 Rankings/Polls 2014-15

[wwhyte]

Re: Women's D1 Rankings/Polls 2014-15

USA TODAY / USA Hockey Magazine Poll - February 3rd

1. Boston College, 190 (19) 1
2. University of Minnesota, 171 2
3. University of Wisconsin, 152 3
4. Harvard University, 121 4
5. Quinnipiac University, 111 5
6. University of Minnesota Duluth, 88 6
7. Boston University, 76 7
8. University of North Dakota, 33 n/a
9. St. Lawrence University, 32 9
10. Cornell University, 31 10
Others receiving votes: Clarkson University, 27; Bemidji State University, 8; Ohio State University, 4; Mercyhurst University, 1.

BC got every single #1 vote

Minny got every single #2 vote

Wisco got every single #3 vote

Is it really that clear-cut? It's amazing to get 19 people to agree so closely.

 

[ARM]

Re: Women's D1 Rankings/Polls 2014-15

BC got every single #1 vote

Minny got every single #2 vote

Wisco got every single #3 vote

Is it really that clear-cut? It's amazing to get 19 people to agree so closely.

The consensus isn't as great in the USCHO Poll:

USCHO.com Division I Women's Poll
February 02, 2015
Team (First Place Votes) Record Points Last Poll
1 Boston College (15) 25- 0-1 150 1
2 Minnesota 24- 1-4 134 2
3 Wisconsin 21- 4-3 121 3
...

Read more: http://www.uscho.com/rankings/d-i-womens-poll/2014-2015/poll,0202/february-2,-2015/#ixzz3Qo8OcHIn

One voter has UW ahead of UM.

 

[FiveHoleFrenzy]

Re: Women's D1 Rankings/Polls 2014-15

The consensus isn't as great in the USCHO Poll:
One voter has UW ahead of UM.

Well, if you dismiss the other 3 head to head games, there is the tie, so maybe that tipped the scales for the voter.

 

[Watson Rink]

Re: Women's D1 Rankings/Polls 2014-15

A technical question that you more statistically adept folks can ponder:

In the algorithms for the various statistical rankings, and also in whatever criteria are in the minds of pollsters, what are the effects of a truly Jekyll/Hyde seasonal performance such as BU and Harvard are experiencing? For example, last week Harvard defeats Quinnipiac and BU but loses to Princeton. Or whatever may come this week out of Harvard's matches with Yale, Brown and BC. Would it be better for H to lose one of the Yale/Brown contests but be the team that breaks up BC's streak (or, in terms of the polls, doesn't defeat BC but takes them to a few overtimes)?

I'm confining my question to the national rankings. The ECAC standings are a whole different animal, every game against every team counts exactly two points, and it's interesting that H's endemic Jekyll/Hyde performance has somehow resulted in its being at the top of the ECAC heap nationally although behind QU, and only a whisker ahead of four other teams, in a six-way ECAC dogfight. This is the result of being undefeated against the three other teams in what would be the ECAC #1-#4 bracket if the tournament were held today, while having been defeated at least once by three of the teams in what would be the #5-#8 bracket. A truly consistent Jekyll/Hyde syndrome.

 

[frozenfeet]

Re: Women's D1 Rankings/Polls 2014-15

A technical question that you more statistically adept folks can ponder:

In the algorithms for the various statistical rankings, and also in whatever criteria are in the minds of pollsters, what are the effects of a truly Jekyll/Hyde seasonal performance such as BU and Harvard are experiencing? For example, last week Harvard defeats Quinnipiac and BU but loses to Princeton. Or whatever may come this week out of Harvard's matches with Yale, Brown and BC. Would it be better for H to lose one of the Yale/Brown contests but be the team that breaks up BC's streak (or, in terms of the polls, doesn't defeat BC but takes them to a few overtimes)?

I'm confining my question to the national rankings. The ECAC standings are a whole different animal, every game against every team counts exactly two points, and it's interesting that H's endemic Jekyll/Hyde performance has somehow resulted in its being at the top of the ECAC heap nationally although behind QU, and only a whisker ahead of four other teams, in a six-way ECAC dogfight. This is the result of being undefeated against the three other teams in what would be the ECAC #1-#4 bracket if the tournament were held today, while having been defeated at least once by three of the teams in what would be the #5-#8 bracket. A truly consistent Jekyll/Hyde syndrome.

Otherwise known as Eastern Bias

 

[ARM]

Re: Women's D1 Rankings/Polls 2014-15

In the algorithms for the various statistical rankings, and also in whatever criteria are in the minds of pollsters, what are the effects of a truly Jekyll/Hyde seasonal performance such as BU and Harvard are experiencing? For example, last week Harvard defeats Quinnipiac and BU but loses to Princeton. Or whatever may come this week out of Harvard's matches with Yale, Brown and BC. Would it be better for H to lose one of the Yale/Brown contests but be the team that breaks up BC's streak (or, in terms of the polls, doesn't defeat BC but takes them to a few overtimes)?

Most of the computer rankings tend to reward the excellent result. For something like the PWR, it really isn't able to track a win over a BC-type team versus a win over Cornell. If you go 1-1 against those teams, it only matters in things like COp, and obviously, the impact it has in the comparison against those two teams, whether you defeat BC or Cornell. When I vote in a poll, I tend to reward the "good" win for teams near the bottom of the top 10, but for a team trying to move to the very top, it is a big red flag if it loses to a team as weak as Brown.

The ECAC standings are a whole different animal, every game against every team counts exactly two points...

While that's true, if you lose to the teams that are very close to you in the standings, those hurt worse because they gain points at your expense. If you lose to Brown, none of Cornell, Clarkson, Quinnipiac benefit directly.

 

[TonyTheTiger20]

Re: Women's D1 Rankings/Polls 2014-15

For something like the PWR, it really isn't able to track a win over a BC-type team versus a win over Cornell. If you go 1-1 against those teams, it only matters in things like COp, and obviously, the impact it has in the comparison against those two teams, whether you defeat BC or Cornell.

Well, I think you're forgetting about RPI. Beating a team like BC will help a lot more than beating a low-level TUC like Cornell.

It won't make a difference in TUC but the vast majority of PWR follows RPI anyway given that RPI is the tiebreaker.

As for the different computer rankings, it depends on the formula.

PWR is basically an adjusted RPI, so if you want to understand PWR you can figure out 90% of it by looking RPI. Since RPI protects top teams that beat bad teams by adding RPI points for a win that would otherwise hurt you, PWR *REALLY* hurts you for losing to a bad team.

For example when BC beat Providence it lowered their RPI -- but the different was added back in (idk what it was, call it like 0.005). If BC had lost to Providence, we would have had the corresponding losing value, but wouldn't have had the benefit of adding back in the 0.005.

With the GRaNT ratings, teams that are jekyll and hyde will be rewarded more than a team that wins and loses the games it 'should'.

If BC beats a great team (eg. rating of 1.200) and loses to a bad team (eg. rating of 0.200), the net effect would be something like (1.200+0.000)/2 = 0.600, divided out over two add'l games.

If BC splits against an average team (say, rating of 0.700 -- right in the middle of the two teams above), the net effect would be (0.700+0)/2 = 0.350, divided out over two add'l games.

Not sure about some of the other rankings.

 

[ARM]

Re: Women's D1 Rankings/Polls 2014-15

Well, I think you're forgetting about RPI. Beating a team like BC will help a lot more than beating a low-level TUC like Cornell.

No. If you go 1-1 versus BC and Cornell, RPI doesn't know care which one you beat and which one beat you. The win adds to your winning percentage, the lost hurts it, but you are going to pick up opponents winning percentage and opponents-opponents winning percentage either way.

 

[TonyTheTiger20]

No. If you go 1-1 versus BC and Cornell, RPI doesn't know care which one you beat and which one beat you. The win adds to your winning percentage, the lost hurts it, but you are going to pick up opponents winning percentage and opponents-opponents winning percentage either way.

Sorry yeah -- I think we're talking about different things. I meant that beating BC will help more than beating Cornell -- independent of each other, which I guess was actually not the point was it? haha

Oops

 

[driventoit]

Re: Women's D1 Rankings/Polls 2014-15

Re Watson Rink's Jekyll/Hyde syndrome:

It seems to me that Harvard is uniquely positioned to make regular late-season hay, whatever the algorithm, since in addition to its conference foes usually being well represented in the polls, its Beanpot opponents are perennially strong. Thus this year the Crimson play six current top-ten teams (and seven games---Cornell x 2) in the last month of the season. I don't know how routinely other teams face this potential gold mine/perfect storm of possible outcomes.

 

[TonyTheTiger20]

Re: Women's D1 Rankings/Polls 2014-15

Re Watson Rink's Jekyll/Hyde syndrome:

It seems to me that Harvard is uniquely positioned to make regular late-season hay, whatever the algorithm, since in addition to its conference foes usually being well represented in the polls, its Beanpot opponents are perennially strong. Thus this year the Crimson play six current top-ten teams (and seven games---Cornell x 2) in the last month of the season. I don't know how routinely other teams face this potential gold mine/perfect storm of possible outcomes.

There's two ways of looking at it -- one that Harvard has the *opportunity* to climb with a hot end to the half, but then one that says that they might be looking at a couple losses in there which might drop them lower than if they played a bunch of Hockey East cupcakes.

 

[driventoit]

Re: Women's D1 Rankings/Polls 2014-15

There's two ways of looking at it -- one that Harvard has the *opportunity* to climb with a hot end to the half, but then one that says that they might be looking at a couple losses in there which might drop them lower than if they played a bunch of Hockey East cupcakes.

Yeah. The couple losses scenario is one I thought I had covered in my "perfect storm" side of things.

But it all goes to making the Beanpot something special: nary a cupcake in sight, and it comes complete with its own Hall of Fame. (Talk about the Hub of the Universe!)

 

[TonyTheTiger20]

Re: Women's D1 Rankings/Polls 2014-15

Oh I know it -- it's harder to win the women's beanpot when you're good than it is to win the men's beanpot when you're good I think.

 

[Watson Rink]

Re: Women's D1 Rankings/Polls 2014-15

Don't get me started on the Good Old Days of the ECAC prior to its lamentable fission into HE and ECACHL.....BUT suppose that Joey Bertagna's legacy were to be a healing of that split and we saw a reconstituted league composed of:

BC BU NU UNH H Y Pr D Br Cornell

27 games, 3x round robin (1x home, 1x away, 1x neutral site as part of a doubleheader with a regular hometeam's home game). Almost as much rivalry concentration as in the Original Six NHL. ECAC tourney to consist of 2 bye teams, 4 other teams, three rounds.

National tourney to consist of 3 ECAC teams, 3 WCHA teams, 2 teams from a reconstituted CHA that included all the NY and NE teams not included in the ECAC. Guaranteed no intraconference games in the QFs.

What say?

 

[robertearle]

Re: Women's D1 Rankings/Polls 2014-15

The consensus isn't as great in the USCHO Poll:
One voter has UW ahead of UM.

For the Feb 9th poll:
(I think) four voters have Wisconsin ahead of Minn for 2nd, and then four voters have Harvard ahead of Wisconsin for 3rd.

 

[D2D]

Re: Women's D1 Rankings/Polls 2014-15

USCHO.com D-I Women's Poll
February 9, 2015

1 Boston College (15)
2 Minnesota
3 Wisconsin
4 Harvard
5 Quinnipiac
6 Minnesota-Duluth
7 Boston University
8 Clarkson
9 North Dakota
10 Cornell
Others receiving votes: St. Lawrence 8, Princeton 1

 

[Golden Tuuk]

Re: Women's D1 Rankings/Polls 2014-15

If BC beats a great team

Got a chuckle out of that.

 

[D2D]

Re: Women's D1 Rankings/Polls 2014-15

USCHO.com D-I Women's Poll
February 16, 2015

1 Boston College (15)
2 Minnesota
3 Harvard
4 Wisconsin
5 Quinnipiac
6 Boston University
7 Clarkson
7 North Dakota
9 Minnesota-Duluth
10 St. Lawrence
Others receiving votes: Ohio State 6, Cornell 5, Princeton 4

 

[KTDC]

Re: Women's D1 Rankings/Polls 2014-15

USCHO.com D-I Women's Poll
February 16, 2015

First time 1st Place isn't unanimous for a while

…..Team...(First Place Votes)…Record…..Points..Last Poll
1..Boston College ....(13)………..30-1-2……148……….1
2..Minnesota……..…….(2)…………29-2-4.……139…..….2
3..Wisconsin…………………………….24-6-4…….112……...4
4..Harvard……………………………….21-5-3…….106……...3
5..Clarkson……………………………..22-9-3………72………..7
6..Quinnipiac…………………………..24-7-3……..64………..5
7..Boston University……………….21-8-3……..63………..6
8..North Dakota……………………..20-11-3…….54………..7
9..Minnesota-Duluth……………...19-10-5…….40………..9
10..St. Lawrence…………………...19-10-5….…22………..10

Others receiving votes: Cornell 4, Ohio State 1.

 

[robertearle]

Re: Women's D1 Rankings/Polls 2014-15

USCHO.com D-I Women's Poll
February 16, 2015

First time 1st Place isn't unanimous for a while

…..Team...(First Place Votes)…Record…..Points..Last Poll
1..Boston College ....(13)………..30-1-2……148……….1
2..Minnesota……..…….(2)…………29-2-4.……139…..….2
3..Wisconsin…………………………….24-6-4…….112……...4
4..Harvard……………………………….21-5-3…….106……...3
5..Clarkson……………………………..22-9-3………72………..7
6..Quinnipiac…………………………..24-7-3……..64………..5
7..Boston University……………….21-8-3……..63………..6
8..North Dakota……………………..20-11-3…….54………..7
9..Minnesota-Duluth……………...19-10-5…….40………..9
10..St. Lawrence…………………...19-10-5….…22………..10

Others receiving votes: Cornell 4, Ohio State 1.

This is the 2nd week in a row that the points for 1st and 2nd total more than (150 plus 135 equals) 285. How can that be?

Fifteen voters, 10 points for first, 9 points for second, etc etc. 150 available for first place votes, 135 for second place; So how can BC's total plus Minn's total be more than 285?