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.

2011-12 DI Rutter Computer Rankings

L

LakersFan

Registered User
Since the season started this past weekend with a pair of non-conference games, the Rutter Computer rankings are online for 2011-2012. You can find them here:

http://math.bd.psu.edu/faculty/rutter/WomensRankings.html

My ranking system is Bayesian, therefore prior information is used. In this case, the prior is the rating of each team at the end of last season, prior to the NCAA tourney. By using a prior, rankings are possible early in the season. Lindenwood, as they are a new team, started with a rating of 0. After losing two games to Wisconsin, their rating dropped to -.52. As teams play more games, the impact of the prior decreases.
 
Re: 2011-12 DI Rutter Computer Rankings

LakersFan said:
Lindenwood, as they are a new team, started with a rating of 0. After losing two games to Wisconsin, their rating dropped to -.52.
Click to expand...
Well, it's a good thing that they are meeting Mankato to decide once and for all which team really belongs in that coveted #26 spot. No? Maybe #6 vs #7 and #10 vs #11 do offer a little more intrigue.
 
Re: 2011-12 DI Rutter Computer Rankings

Rankings for games played through 10/2/2011

Code: 
Rank	Team		Rating
1 	Wisconsin 	2.0277 	
2 	Minnesota 	1.5326 	
3 	Cornell 	1.4434 	
4 	UMD	 	1.0850 	
5 	BU	 	0.7580 	
6 	Mercyhurst 	0.7517 	
7 	North Dakota 	0.7284 	
8 	Boston College 	0.6569 	
9 	Bemidji State 	0.5670 	
10 	New Hampshire 	0.5626

Still very early, so Quinnipiac does not move much based on a split with Mercyhurst.
 
Re: 2011-12 DI Rutter Computer Rankings

LakersFan said:
Still very early, so Quinnipiac does not move much based on a split with Mercyhurst.
Click to expand...
New Hampshire seemed to get a lot of mileage out of a sweep of Niagara. I imagine the bottom line is that the Wildcats are still unbeaten, therefore wins over a lesser-ranked teams can vault them higher than the Bobcats' split with a top-10 team.
 
Re: 2011-12 DI Rutter Computer Rankings

ARM said:
New Hampshire seemed to get a lot of mileage out of a sweep of Niagara. I imagine the bottom line is that the Wildcats are still unbeaten, therefore wins over a lesser-ranked teams can vault them higher than the Bobcats' split with a top-10 team.
Click to expand...

Only two games have been played, so unless you split with with Wisconsin, two wins will always be better than a win and a loss. If there was no prior, Quinnipiac and Mercyhurst would be tied, but Mercyhurst was rated higher last year, so they are still higher than Quinnipiac. If Quinnipiac and Mercyhurst are truly equal, their ratings will converge to the same value, assuming they play similar schedules (in terms of strength of opponents) and a large number of games. They likely will play different schedules however, and by end of the season we will see who benefits more from the split.
 
Re: 2011-12 DI Rutter Computer Rankings

For games played through October 9, 2011

And Maine vaults into the top 10!

Code: 
  	Team 		Rating 	
1 	Wisconsin 	2.3774 
2 	Minnesota 	1.6631 	
3 	Cornell 	1.4399 	
4 	UMD	 	0.9286 	
5 	Northeastern 	0.8997 	
6 	Maine 		0.8913 
7 	Boston College 	0.8101 	
8 	Boston Univ. 	0.7928 	
9 	North Dakota 	0.5640 	
10 	Ohio State 	0.5448
 
Re: 2011-12 DI Rutter Computer Rankings

Rankings for games played through October 16, 2011. Wiscosin drops a little, and Maine was in the top 10 for only a week.

Code: 
  	Team 		Rating 	
1 	Wisconsin 	2.1585 	
2 	Minnesota 	1.8359 	
3 	Cornell 	1.4488 
4 	UMD	 	1.1203 	
5 	Boston Univ. 	1.0596 	
6 	Northeastern 	1.0347 	
7 	Boston College 	0.9026 	
8 	Ohio State 	0.8879 	
9 	North Dakota 	0.8201 	
10 	Bemidji State 	0.6138

By the way, I just had an article about ranking Women's College Hockey teams published by Chance Magazine, not available at your local book store. You can find it here: http://chance.amstat.org/2011/09/ranking-womens-hockey/. If you would like more information, please send me an email. You can find contact info at my webpage: http://math.bd.psu.edu/faculty/rutter/. And, no, I do not know who the women in the picture from the article is.
 
Re: 2011-12 DI Rutter Computer Rankings

LakersFan: the link seems to only get you to a page that gives the first couple sentences of the article in "Chance". How can I get the whole thing? BTW, as a womens college hockey fan and a stats geek, I have followed your work since you first started posting and hope you'll keep developing it.
 
Re: 2011-12 DI Rutter Computer Rankings

hab said:
LakersFan: the link seems to only get you to a page that gives the first couple sentences of the article in "Chance". How can I get the whole thing? BTW, as a womens college hockey fan and a stats geek, I have followed your work since you first started posting and hope you'll keep developing it.
Click to expand...

Hab, send me an email I will see about getting you a copy of the paper. My webpage has my address.
 
Re: 2011-12 DI Rutter Computer Rankings

For games played through October 23, 2011

Code: 
	Team 		Rating 	
1 	Wisconsin 	2.5469 
2 	Minnesota 	2.1348 	
3 	Cornell 	1.4443 	
4 	North Dakota 	1.2607 	
5 	Boston Univ. 	1.1757 	
6 	UMD		1.0076 	
7 	Boston College 	1.0033 	
8 	Northeastern 	0.4813 
9 	Bemidji State 	0.4418 	
10 	Mercyhurst 	0.4260
 
Re: 2011-12 DI Rutter Computer Rankings

For games played through October 30, 2011
Code: 
  	Team 	Rating 	
1 	Wisconsin 	2.4624 	
2 	Cornell 	1.7077 	
3 	Minnesota 	1.6576 	
4 	Harvard 	1.1498 	
5 	UMD		1.1257 	
6 	North Dakota 	1.0533 	
7 	Boston Univ. 	0.8999 
8 	Bemidji State 	0.8057 	
9 	Boston College 	0.7866 
10 	Ohio State 	0.7648
 
Re: 2011-12 DI Rutter Computer Rankings

Curious what others are think, but I am wondering if women's D1 hockey is just a bad fit form this concept. This ranking seems extremely inaccurate. Here are a couple of examples in my opinion. 1st, does anyone really think Harvard should be number 4 on any list. Their Thanksgiving weekend series with Minny should confirm that. 2nd, Ohio State get smoked by UND twice and then squeaks by Mankato and they somehow jump up to number 10 on this list, seriously. 3rd and less glaring is the impact on Minnesota and BSU. Anyone else have an opinion?
 
Re: 2011-12 DI Rutter Computer Rankings

granddaddyscout said:
Curious what others are think, but I am wondering if women's D1 hockey is just a bad fit form this concept. This ranking seems extremely inaccurate. Here are a couple of examples in my opinion. 1st, does anyone really think Harvard should be number 4 on any list. Their Thanksgiving weekend series with Minny should confirm that. 2nd, Ohio State get smoked by UND twice and then squeaks by Mankato and they somehow jump up to number 10 on this list, seriously. 3rd and less glaring is the impact on Minnesota and BSU. Anyone else have an opinion?
Click to expand...
Regarding Harvard, they've only played two games, so it is tough to conclude a ton from those games, especially as one went to OT. I saw Harvard versus Minnesota a year ago, and the difference between the two teams was goaltending, particularly in the second game. They are a very fast team that gave Minnesota problems at times. I won't be surprised if they get points off of Minnesota on their home ice. Two years ago, the Gophers didn't score in two games at Bright. Regarding OSU, remember that they had a win and tie versus BSU, so the Beavers' win over UM probably helped the Buckeyes. These rankings don't care about margin, so a win is a win, and the Buckeyes' wins over MSU look as good as the Gophers' and much better than Mercyhurst's split.
 
Re: 2011-12 DI Rutter Computer Rankings

Rank em Grandaddy. Let's see how you think they should be rated.
 
Re: 2011-12 DI Rutter Computer Rankings

granddaddyscout said:
Curious what others are think, but I am wondering if women's D1 hockey is just a bad fit form this concept. This ranking seems extremely inaccurate. Here are a couple of examples in my opinion. 1st, does anyone really think Harvard should be number 4 on any list. Their Thanksgiving weekend series with Minny should confirm that. 2nd, Ohio State get smoked by UND twice and then squeaks by Mankato and they somehow jump up to number 10 on this list, seriously. 3rd and less glaring is the impact on Minnesota and BSU. Anyone else have an opinion?
Click to expand...

Just remember that the Ivy schools start a month later than everyone else, and since their sample size is small (1/4 or less compared to the other teams), it is statistically more error prone. My point is that the positions of Harvard and Cornell in this poll have more statistical variance due to the small sample size.

My opinion is that Cornell is pretty much in right place given their domineering results sofar, and that Harvard in reality belongs somewhere in the bottom end of the top 10.The rest of the rankings are a very good representation of where teams fit based on the results of the first 5 weeks into the season.
 
Re: 2011-12 DI Rutter Computer Rankings

WCHFan said:
Rank em Grandaddy. Let's see how you think they should be rated.
Click to expand...
No Need to ask. We know his/her answer ahead of time and it would look something like this...

Wisco, Minny, UMD, UND and then the rest of the pack. :D
 
Re: 2011-12 DI Rutter Computer Rankings

My comment is more about the process that produced this result. That said I do believe the West will produce at least 4 teams worthy of Tournament participation, I believe that the WCHA provides a better environment to prepare teams as well. So, I do think the NCAA Championship trophy will stay out west again. I will add that Wisco seems stronger than I expected, I watched significant portions of their BU and UM series and I don't think BU even with Poulin would be a UW's level of play. I also felt that UW was the better team by a fair margin against the Gophers even with the loss. Frankly I thought they would "feel" the loss of Dugan and the other 3 Seniors from last year more than they appear to.
On paper I think Cornell is deserving of their ranking, but I just don't think their schedule will prepare them for Tournament level hockey in the end.

Sorry if it sounds like bias, I feel like it is an honest assessment.
 
Re: 2011-12 DI Rutter Computer Rankings

I agree with the long-term assessment, but there's no computer ranking I know that's going to forecast teams' postseason performance based on the quality of competition within their conference.

LakersFan's ranking uses some information about prior seasons results to provide meaningful rankings when current data is limited (as with Harvard & Cornell now), but the importance of this goes to zero as the season progresses. The idea is to do what you can based on the games played this season. Ideally, Lakersfan wants to develop something that would replace the current selection system, and using anything that would use past season results is anathema to the NCAA (never mind the hypocrisy that football's national champion is decided by a beauty pageant where anything goes)

To make predictions about postseason performance you'd need some model of how in-season competition matters in terms of predicting postseason performance over the years. I don't know of anyone who's done that. Football voters are effectively using such a model when they rank LSU & Alabama above Boise State, based on the SEC's recent success.
 
Last edited:
Re: 2011-12 DI Rutter Computer Rankings

For games played through November 6, 2011

Code: 
  	Team 	Rating 	
1 	Wisconsin 	2.5710 	
2 	Minnesota 	1.8923 	
3 	Cornell 	1.7364 	
4 	Harvard 	1.0909 	
5 	North Dakota 	1.0394 	
6 	UMD	 	0.9227 	
7 	Boston Univ 	0.8671 	
8 	Northeastern 	0.7372 	
9 	Ohio State 	0.7108 	
10 	Bemidji State 	0.7088

Granddaddyscout,

I think my rankings are only useful for trying to figure out who should get the at-large bids to the NCAA tournament. In order to predict success in the tournament, the model would need to change. Due to the rules, there are no ties and overtime is extended, so I would need to remove that part of the model. But other things change as well. For example, I can show statistically that in some years, power plays are less frequent in the NCAA tournament than during the regular season. My model does not include any such effects, so it wouldn't be applicable.

If you did use my ratings to create rough probabilities of winning, obviously the number one team would have the greatest chance of winning the NCAA tournament. But if you just found the probability of any WCHA team winning, my guess for most years that would be greater than 50%, since they usually have two of the top four teams. It all depends on what question you ask. Is a team from the WCHA most likely to win the NCAA tournament? Most years the answer is yes due to quality AND quantity. Is the streak considered a rare event in a statistical sense? I will have to get back you on that.
 
You must log in or register to reply here.
Back
Top