Colgate vs. Cornell home-and-home 1/19-20

[Jim Hyla]

the creativity of the passing has been much improved this year.  since the team is young that points at better recruiting.

our big guys can carry the puck. like, really carry the puck. it's been nice to have a team that can beat a forecheck. we used to get spooked by even halfhearted efforts from mediocre teams. now we bring the puck out.

Agree with all of this. In past years we've seemed tentative and now we seem to skate it out with authority. As the year has progressed I've also noticed that, when they can't skate the puck into the offensive zone,  instead of always dumping the puck in, sometimes they skate back to regroup and try again. I can't say I remember them doing that, with some consistency, in forever. It speaks to having confidence in your puck control and passing,

But I'm still hesitant to think this years team is going to end the year near the current position. We and Clarkson are #2 & 1 in national shooting percentage. While some of that may be because of the rest of the league being relatively softer than in past years, I just don't know that we can keep it up, especially in post-season play.

But this year is sure a lot of fun.

[margolism]

Note that two of our [uninjured] draft picks - Tshantz and Cairns - are not regular players in the lineup.  So we aren't even consistently playing with all our NHL draft picks.

Similarly, look at all the blue chip 1st and 2nd round draft picks on the [unranked!] BU team.  On paper, BU should be a top 4 team.

[Trotsky]

But this year is sure a lot of fun.

It is and we should savor it because whatever the x factor is that this team may have it is fleeting.

This is a very good team playing very well.  We all know that doesn't happen very often.

Make it count.

[adamw]

I'm pretty sure Cornell was #1 in PWR at the start of the 2003 NCAA tournament. Not sure if there's a reliable week by week history anywhere, but you could probably dig up end of season ratings from USCHO or something.

If the links are relatively consistent you might be able to use archive.org to get some of that data.

We have most of this historical stuff on CHN. Follow the variety of links from the Pairwise page.

[adamw]

I believe a lot of the credit goes to Ben Syer.  It's since his arrival that the recruiting has changed.  Hopefully he sticks around awhile, but my gut says there is a head coaching opportunity coming his way in the near future.

Syer is a good coach, but to whatever extent recruiting has changed - if it has - doesn't really have to do with Syer. Ivy League rules changed in recent years, where Cornell/Brown/Dartmouth are now able to keep up with the recruiting of Harvard/Yale/Princeton, which itself had changed a couple years before that.

[Trotsky]

Ivy League rules changed in recent years, where Cornell/Brown/Dartmouth are now able to keep up with the recruiting of Harvard/Yale/Princeton, which itself had changed a couple years before that.

Are you talking about financial aid?

[Beeeej]

I'm pretty sure Cornell was #1 in PWR at the start of the 2003 NCAA tournament. Not sure if there's a reliable week by week history anywhere, but you could probably dig up end of season ratings from USCHO or something.

If the links are relatively consistent you might be able to use archive.org to get some of that data.

We have most of this historical stuff on CHN. Follow the variety of links from the Pairwise page.

...except you don't really have historical charts - you have the ability to generate them. In order to get each week's PWR, you have to enter the date manually in the "Customize" tab. Essentially, to get the answer to the question we're looking for, we'd have to enter a few hundred different dates and compare manually. That's correct, isn't it?

(EDIT to add) And you also only go back to 2005-06, which is not helpful in terms of finding out how high we got in the PWR late in, say, the 2002 season or where we were the week before we hit #1 in 2003.

[Trotsky]

All I want from life is a csv file of every D-1* (and before the era, every NC$$, and before that era, define it how you will) game that has ever been played since 1895, so we can generate PWR for any interval by any set of PWR rules.

Why is that so hard?  I mean, I did the one for Cornell.  Only about 100 schools (counting defuncts) to go.

[Beeeej]

All I want from life is a csv file of every D-1* (and before the era, every NC$$, and before that era, define it how you will) game that has ever been played since 1895, so we can generate PWR for any interval by any set of PWR rules.

Why is that so hard?  I mean, I did the one for Cornell.  Only about 100 schools (counting defuncts) to go.

:-) I'm not saying that what I want isn't at least a little complicated, I'm just saying that the suggestion CHN already has what I want is a little bit facile. I'm actually wondering, and I'm asking because I genuinely don't know - how difficult would it be to add code to any given historical PWR page on CHN so that there's a left arrow pointing to a page that generates the previous week's PWR from the same customization code, and similarly a right arrow pointing to the subsequent week? Obviously even parsing it by week is a little imperfect, but it's a start.

[Trotsky]

All I want from life is a csv file of every D-1* (and before the era, every NC$$, and before that era, define it how you will) game that has ever been played since 1895, so we can generate PWR for any interval by any set of PWR rules.

Why is that so hard?  I mean, I did the one for Cornell.  Only about 100 schools (counting defuncts) to go.

:-) I'm not saying that what I want isn't at least a little complicated, I'm just saying that the suggestion CHN already has what I want is a little bit facile. I'm actually wondering, and I'm asking because I genuinely don't know - how difficult would it be to add code to any given historical PWR page on CHN so that there's a left arrow pointing to a page that generates the previous week's PWR from the same customization code, and similarly a right arrow pointing to the subsequent week? Obviously even parsing it by week is a little imperfect, but it's a start.

No, I'm not being sarcastic.  I honestly think that if the SIDs at each of the 60 schools put an unpaid intern on it at .1 FTE we'd have a 99.9% correct dataset of every D-1-"qualifying" game in history by this time next year.

[adamw]

All I want from life is a csv file of every D-1* (and before the era, every NC$$, and before that era, define it how you will) game that has ever been played since 1895, so we can generate PWR for any interval by any set of PWR rules.

Why is that so hard?  I mean, I did the one for Cornell.  Only about 100 schools (counting defuncts) to go.

That is all I want from life as well ... since I spent last summer doing this for about 10 schools, including Cornell. Those datasets are already included on CHN, and searchable as well, using various criteria, from our schedule page. (See the dropdown menus folks)

Along the way, I discovered numerous errors and omissions in media guides, where I filled in the blanks. So I'm not even sure I'd trust just some intern at each SID office to turn their media guide stuff into CSV, but it couldn't hurt I guess.

The goal is to do this for every "D-I" game for all-time - but that might be a lifelong project. Feel free to volunteer to help out.

I'm not being a wise-ass --- or facile (great word) --- but there's a treasure trove of data on CHN right now that is there. I know it's not always the easiest to find - but I do my best to make it as easy as possible. It's a challenge coming up with the best ways to make it obvious it's there, while not intruding on the basic function of what the site is (current news/data/etc...)

[adamw]

I'm pretty sure Cornell was #1 in PWR at the start of the 2003 NCAA tournament. Not sure if there's a reliable week by week history anywhere, but you could probably dig up end of season ratings from USCHO or something.

If the links are relatively consistent you might be able to use archive.org to get some of that data.

We have most of this historical stuff on CHN. Follow the variety of links from the Pairwise page.

...except you don't really have historical charts - you have the ability to generate them. In order to get each week's PWR, you have to enter the date manually in the "Customize" tab. Essentially, to get the answer to the question we're looking for, we'd have to enter a few hundred different dates and compare manually. That's correct, isn't it?

(EDIT to add) And you also only go back to 2005-06, which is not helpful in terms of finding out how high we got in the PWR late in, say, the 2002 season or where we were the week before we hit #1 in 2003.

We have back to 2003-04

https://www.collegehockeynews.com/ratings/ncaapwcr.php?custom=1&pwseason=20032004

Where are you seeing only back to 2005-06? Is there a page somewhere which only links to that? If so, let me know, so I can fix it.  If you go to the main Pairwise page and hover over the "Pairwise Historical" tab, you'll see a link there that takes you back to 2003-04.

2002-03, Cornell was most definitely No. 1 overall in Pairwise prior to the tournament. There is a complication with generating 2002-03 Pairwise - which I can't remember at the moment - but I have to go back and remember the reason why. I know I've come across this issue a bunch of times in trying to do so. And I've tried exactly for this reason - to show Cornell No. 1 - which I know they were. You could find my Pairwise Analysis article from that time and verify that.

Prior to 2002-03, things get tricky because of the numerous changes in the Pairwise criteria. It could be generated based on current criteria, but that wouldn't give you an accurate assessment of what was what.  Also, prior to a fuzzy/hazy time before 2002-03, the Committee didn't use the Pairwise like it does now - so it's almost irrelevant, actually.  Again, you could see it for academic purposes, but it wouldn't give you a true snapshot of time.

By the way, these pages - one for each season - have a summation of all of this kind of stuff.  For example, 2003 here:

https://www.collegehockeynews.com/reports/ncaa-tournament/2003

And it has the left/right arrow paradigm that you refer to. I haven't done that kind of thing for a week-by-week of the Pairwise - yet - not sure how used that would ever be. Or even if a week-by-week concept even makes sense in the context of the Pairwise, since it literally can change at any time. I get that "one week" is a good time frame in college hockey terms, but it's also kinda arbitrary when it comes to the Pairwise.

[Beeeej]

We have back to 2003-04

https://www.collegehockeynews.com/ratings/ncaapwcr.php?custom=1&pwseason=20032004

Where are you seeing only back to 2005-06?

On this very page to which you are directing me. When I click on the drop-down menu for Modify Date Parameters, the earliest season it gives me is 2005-06. So even if what you've directed me to is a page that, when I click on the "Main chart tab" gives me the end of the 2003-04 season, it doesn't give me access to any other part of the 2003-04 season or any other PWR snapshot prior to 2005-06.

[jkahn]

Here's the Pairwise the week before we hit #1 in 2003.
I had cut and pasted it on to a spreadsheet.


Rk Team                GP  W- L- T  Win%  Rk     RPI  Rk  PWR
 1 Colorado College    38 28- 5- 5 0.8026  2 | 0.5931  1 | 29
 2 Cornell             31 26- 4- 1 0.8548  1 | 0.5899  2 | 28
 3 New Hampshire       38 25- 7- 6 0.7368  4 | 0.5894  3 | 27
 4 Boston University   40 24-13- 3 0.6375 13 | 0.5854  4 | 26
 5 Maine               38 24- 9- 5 0.6974  7 | 0.5806  6 | 24
   Ferris State        38 29- 8- 1 0.7763  3 | 0.5664  8 | 24
 7 Minnesota           39 22- 8- 9 0.6795  9 | 0.5840  5 | 23
 8 Boston College      37 23-10- 4 0.6757 11 | 0.5795  7 | 22
 9 North Dakota        41 26-10- 5 0.6951  8 | 0.5637  9 | 20
   Michigan            38 26- 9- 3 0.7237  5 | 0.5606 10 | 20
11 MSU-Mankato         38 20- 8-10 0.6579 12 | 0.5596 11 | 19
12 Ohio State          39 24-10- 5 0.6795  9 | 0.5528 13 | 18
13 Harvard             31 21- 8- 2 0.7097  6 | 0.5543 12 | 16
14 Michigan State      38 23-13- 2 0.6316 15 | 0.5462 16 | 15
15 St. Cloud State     37 17-15- 5 0.5270 27 | 0.5524 14 | 14
   Providence          36 19-14- 3 0.5694 20 | 0.5461 17 | 14
17 Dartmouth           32 19-12- 1 0.6094 17 | 0.5248 19 | 13
   Northern Michigan   38 20-16- 2 0.5526 25 | 0.5244 20 | 13
19 Denver              41 21-14- 6 0.5854 18 | 0.5508 15 | 12
20 Massachusetts       37 19-17- 1 0.5270 27 | 0.5313 18 | 11
21 Notre Dame          39 17-16- 6 0.5128 30 | 0.5216 21 | 10
22 Minnesota-Duluth    39 20-14- 5 0.5769 19 | 0.5194 22 |  9
23 Yale                32 18-14- 0 0.5625 22 | 0.5145 24 |  6
24 Merrimack           36 12-18- 6 0.4167 40 | 0.5154 23 |  5
   Miami               41 21-17- 3 0.5488 26 | 0.5083 26 |  5
26 Western Michigan    38 15-21- 2 0.4211 39 | 0.5118 25 |  4
27 Mass.-Lowell        36 11-20- 5 0.3750 45 | 0.5057 27 |  3
   Alaska-Fairbanks    36 15-14- 7 0.5139 29 | 0.5011 28 |  3
29 Brown               33 16-12- 5 0.5606 23 | 0.5010 29 |  2
30 Wayne State         39 21-16- 2 0.5641 21 | 0.4798 38 |  0

[adamw]

We have back to 2003-04

https://www.collegehockeynews.com/ratings/ncaapwcr.php?custom=1&pwseason=20032004

Where are you seeing only back to 2005-06?

On this very page to which you are directing me. When I click on the drop-down menu for Modify Date Parameters, the earliest season it gives me is 2005-06. So even if what you've directed me to is a page that, when I click on the "Main chart tab" gives me the end of the 2003-04 season, it doesn't give me access to any other part of the 2003-04 season or any other PWR snapshot prior to 2005-06.

It was available under the "Pairwise Historical" tab ... but I've now also fixed the dropdown to allow back to 2003-04 ... You could also replace the URL, although again, it won't work prior to that season.