RIT to go D-I

[ithacat]

A few interesting paragraphs from the article on USCHO...

[Atlantic Hockey commissioner] DeGregorio and [Canisius AD] Dillon also met with management of the Blue Cross Arena in downtown Rochester about making it a permanent home for the annual Atlantic Hockey championships. The D-I committee has awarded a 2007 regional to the recently renovated 11,215-seat venue that is home to the AHL Rochester Americans. "It's another nice piece for RIT if things were able to be worked out," Dillon said. The Blue Cross Arena could also be home to an in-season tournament, he suggested.

The move to D-I will give Rochester, N.Y. its first Division I program. The nearest D-I sport in the metro area of more than 1 million people is lacrosse at Hobart College.

As part of its review of other applicants interested in joining the league, Atlantic Hockey plans to visit Air Force in January.

I love the idea of Rochester getting an annual tournament and the league championship. It could be the beginning of getting Rochester into a regular rotation for the NCAA tournament. Here's hoping Cornell's in that regional in '07.

[calgARI '07]

Rochester would be a horrible location for the league championship because it is far from almost every school except Cornell and Colgate.  That is the reason they left Lake Placid in the first place.  Geographically, Albany is probably the best location.

[Tub(a)]

[Q]calgARI '07 Wrote:

 Rochester would be a horrible location for the league championship because it is far from almost every school except Cornell and Colgate.  That is the reason they left Lake Placid in the first place.  Geographically, Albany is probably the best location.[/q]

The Atlantic Hockey league, not the ECACHL.

[Robb]

More to the point, an *NCAA* regional - I hope Cornell NEVER participates in the AHA league tournament!  

It would be nice to have a regional in OUR backyard for a change, rather than traveling to eastern New England or out west.

[ithacat]

This is all in regard to the AH tournament and the NCAAs...it'd be great for Rochester, which is a good hockey town, and hopefully would allow Cornell fans to travel a couple of hours once in awhile for NCAA games.

[billhoward]

Indeed, Rochester / Blue Cross Arena would be a good place for an NCAA regional. The capacity is up from 7250 (when it was the War Memorial) to something like 11,200 circa 1996, so that should be enough to keep the NCAA happy for regionals. They probably want more capacity (>15,000?) for an NCAA Frozen Four now.

But there has to be a host school so Cornell would have to step up to the plate. That has the advantage of getting you into your own regional section.

Don't know how happy the NCAA was with Cornell's handling of the 2004 NCAA lax quarterfinals with the scoreboard woes and under-supplied refreshment stands. Who was at the games - was it really that bad?

RIT could offer to host it and might want to for the prestige. RIT might also want to play a couple of its big D1 games there, say against Cornell.

[billhoward]

Maybe Cornell and RIT could play a home over Thanksgiving weekend, say Friday in Rochester and Saturday in Ithaca. Bored-from-being-at-home Cornellians in the Rochester region would have something to do Friday night. And if it's downtown, there's that incredible club scene afterwards.

[jeh25]

[Q]calgARI '07 Wrote:
Geographically, Albany is probably the best location.[/q]

You sure about that? Both Hartford and Springfield seem more centrally located than Albany to me.  

I have an ECAC distance matrix on a cd somewhere at home that I created while learning Multidimensional scaling (MDS). As far as I recall, of the 12 ECAC cities, New Haven was the most centrally located. I didn't explicitly test whether Hartford or Springfield is even more centrally located, but I could, assuming I can find the cd. Or you could  trust me.

On the pro side, Hartford/Springfield have a strong hockey history and fanbase (formerly the Whalers, now the Wolfpack and the Falcons ).

On the con side, the closest host schools are Quinnipiac (51± 10 miles), Yale (55± 10 miles), Brown (94± 10 miles), Harvard (97± 10 miles), RPI (100± 10 miles), Union (112± 10 miles) and Dartmouth (135± 10 miles). (Man, with the straight shot down I91, I thought DC would be the 2nd closest; Hanover is way up there!)

You could argue that Albany is closer to teams that are historically more likely to actually appear (UVM, CCT, SLU, Cornell) in the championship semis and finals, but frankly, even I'm not pedantic enough to quantify that.

Of course, by the time I get tenure somewhere, UConn and Quinnipiac will have bolted to HE, Sacred Heart and American International will have folded and Union will have returned to D3, leaving us with something like this:

HE
Boston University - Northeastern
Maine - New Hampshire
Boston College - Providence
Massachusetts - UVM
Quinnipiac - UConn
Mass.-Lowell - Merrimack   

The league formerly known as the ECAC
Princeton-Army
Cornell-Colgate
SLU-Clarkson
Yale-Brown
Dartmouth-Harvard
RPI-Holy Cross

You could also add Canisus-RIT and Mercyhurst-Robert Morris, and move to a 16 team league with 8 teams each in eastern and western divisions.

Yes, I'm bored.

[billhoward]

"Multidimensional scaling"? That had to be for course credit because a normal person would use existing toolsets like Streets & Trips or going lowest common denominator, MapQuest.

It would be good to also map out geographic centers where you give extra credit to schools likely to show up for the tournament based on say attendance per game plus percentage of seats filled. And to map it out based on chances of winning the tournament.

I'm thinking Syracuse would make a central location in that case.

[ithacat]

[Q]Don't know how happy the NCAA was with Cornell's handling of the 2004 NCAA lax quarterfinals with the scoreboard woes and under-supplied refreshment stands. Who was at the games - was it really that bad?[/Q]

I was at those games -- the lacrosse was great. The experience was terrible. The facility needs a major upgrade. The concessions are too small and under supplied (they ran out of water at one point & then later they were selling warm water and cold pretzels). The restrooms are a nightmare. The scoreboard snafu was very unfortunate and there seemed (to a casual observer) to be no urgency towards repairing it.  It was a sad display that seemed rather amateurish. I'm sure people worked very hard on it but it appeared Cornell was a bit out of its league.

A couple of Navy fans sitting next to us were pretty critical of the operation...Navy fans? Another couple next to us also drove up from Maryland (SU fans) and since they had a baby they were pretty much prepared for anything, so they just laughed it off.

For Cornell's sake I hope they never attempt to host such an event again until a new crescent stands.

[ben03]

From USCHO front page:
[Q]"Rochester Institute of Technology will join Atlantic Hockey for the 2006-07 season. Atlantic Hockey commisioner Bob DeGregorio and Canisius AD Tim Dillon joined RIT president Al Simone at a press conference this morning to make the news official. RIT will play an independent D-I schedule in 2005-06. RIT athletic director Lou Spiotti and head coach Wayne Wilson outlined RIT's commitment to competing at the D-I level, including the hiring of two full-time assistant coaches, facility improvements, marketing efforts, and additional support staff."[/Q]
Jumping right into the mix ... should be interesting to see how they make out.

[jeh25]

[Q]billhoward Wrote:

 "Multidimensional scaling"? That had to be for course credit because a normal person would use existing toolsets like Streets & Trips or going lowest common denominator, MapQuest.
 [/q]

Nope. For me, the only way to learn a statistical technique is to play with a real dataset. I'm slow like that.

I wanted to learn MDS so I needed a dataset to play with. You are correct, however, that I used MapQuest to determine the distances between all pairs of ECAC schools so I could generate the distance matrix in Excel prior to feeding it into Statistica. Sadly, it was far more productive than some afternoons I spent in lab as a grad student at Cornell...

Just ask Age about the time he stopped by the lab, only to find and the technician and me  weighting pocket change on the electronic balance. (We were trying to figure out for a given weight, which has a higher monetary value, dimes or quarters. We were bored.)

[billhoward]

[Q]jeh25 Wrote:Just ask Age about the time he stopped by the lab, only to find and the technician and me  weighting pocket change on the electronic balance. (We were trying to figure out for a given weight, which has a higher monetary value, dimes or quarters. We were bored.)[/q]

It's a dead heat, isn't it? At least to one decimal place. Quarters are worth 2-1/2 times as much and weigh ~2-1/2 times as much. Doesn't a dollar work out to 3/4 of an ounce or 8/10 of an ounce either way, dimes or quarters? Nickels lose by a factor or 2 or 4 I think.

[KeithK]

[Q]jeh25 Wrote:
I have an ECAC distance matrix on a cd somewhere at home that I created while learning Multidimensional scaling (MDS). As far as I recall, of the 12 ECAC cities, New Haven was the most centrally located. I didn't explicitly test whether Hartford or Springfield is even more centrally located, but I could, assuming I can find the cd. Or you could  trust me.

You could argue that Albany is closer to teams that are historically more likely to actually appear (UVM, CCT, SLU, Cornell) in the championship semis and finals, but frankly, even I'm not pedantic enough to quantify that. [/q]Fortunately I am (or at least bored enough) to quantify.

I used Mapquest to pull distances and driving times from each school (well, city actually) to Albany, Hartford, Springfield and Lake Placid.  Doing a straight average yields the following:

City             with UVM       with QC
-----------    Dist   Time    Dist   Time
Albany        143.6   2:37   143.0   2:34
Hartford      183.3   3:08   166.6   2:52
Springfield   168.5   2:54   156.0   2:42
Lake Placid   195.3   3:33   214.2   3:47

Albany is the winner here with both UVM and QC.  Hartford is actually not much better than Lake Placid with the Cats in the league.

If we weight on number of appearances in Placid/Albany over the last 10 years we get the following:

City             with UVM       with QC
-----------    Dist   Time    Dist   Time
Albany        156.8   2:53   157.0   2:53
Hartford      212.2   3:40   211.1   3:39
Springfield   193.9   3:23   193.2   3:23
Lake Placid   192.9   3:35   199.1   3:39

Albany still wins.  Hartford is actually worse than Lake Placid because of all of the North Country/UVM appearances.

Weighting over the last 20 years (all the years since the split) gives essentially the same results.

City             with UVM       with QC
-----------    Dist   Time    Dist   Time
Albany        157.1   2:53   157.2   2:52
Hartford      211.0   3:39   209.7   3:39
Springfield   192.0   3:22   191.1   3:22
Lake Placid   188.8   3:28   195.2   3:32

None of the above considers weather, which obviously would make Placid look worse.  Also with Mapquest driving times, YMMV.

[jeh25]

[Q]KeithK Wrote:

Fortunately I am (or at least bored enough) to quantify.

I used Mapquest to pull distances and driving times from each school (well, city actually) to Albany, Hartford, Springfield and Lake Placid.  Doing a straight average yields the following...
 [/q]
Nice analysis. I can't point my finger on it, but something feels odd about doing a straight mean.

You'd understand the math far better than I ever could or will, but couldn't you calculate a geometric centroid, treating the 12 ECAC schools as point masses? Then you could just compare the distance from the centroid to Albany and Springfield.

EDIT: Ah. I think I figured my problem with a straight mean. Union and RPI drag down the mean to Albany a lot right?. The appropriateness of the model depends on the league's goal.

If the league is trying to minimize mean travel time, then Albany is the right choice.  

However, if the goal is the most central location, to distribute travel time more evenly, then Albany is a worse choice, because it rewards RPI and Union while penalizing the other schools.  Put another way, it is amazing that Springfield is only 15 miles worse than Albany given that the closest school is 50 miles away.

In other words, you didn't fidn the most central location. You minimized mean travel time, which isn't the same thing.

Assuming  the goal is to evenly distribute travel time to maximize fan attendance (big assumption), could you a) use scary calculus to calculate the centroid, or b) recalculate your tables using the median instead of the mean? Or c) you could use a trimmed mean and throw out the top 2 and bottom 2 values.