Polls

[French Rage]

CHN's pairwise probability matrix has been posted. As usual, it assigns higher-ranked teams too high a chance of winning and lower-ranked teams too low a chance. And as usual, I will waste everyone's time complaining about this. This year I will keep it short and instead highlight a few examples to illustrate the aforementioned issue: the matrix gives Q a 54% chance of winning the ECAC tournament. That is to say, they are more likely to win it than the entire rest of the field combined. The matrix also gives Brown, Yale, Dartmouth, and Princeton a combined 0.0% chance of winning the ECAC tournament. Not even a 0.1% chance—they quite literally have a 0.0% chance of winning the tournament. Collectively, the bottom 8 seeds have a 2.1% chance of winning the ECAC tournament, according to the matrix.

EDIT: BTW, the matrix gives Cornell a 10% chance to win the ECAC tourney and a 0% chance at an at-large bid.

Importantly*, they give us a 0.0% chance at an at-large bid. The teams below us all have this field blank. I think this means our chances at an at-large bid are greater than zero but round down to 0.0%.

Challenge accepted!

[upprdeck]

running numbers on just the known games over the next 7days or so in the matrix its hard to get Cornell much above the 17-19 range..
the issue is so many teams to jump this late.

We can beat Quin in the Tourney and flip 1 PWR, had we not screwed up the first Clark game would could have flipped that one as well.

but you have Conn/Merri/Prov/BU/Mass Low/NE all ahead of us hard to jump and entire league since they will play each other..  You need mass to beat one of them and then have BC/UNH/maine knock off a couple too. But jumping them all only gets you to 14-16 range..

its why the PWR hurts we beat Quin/Clark/BU all h2h but we dont own the PWR over any of them.

[cth95]

So you're telling me there's a chance!?

[Trotsky]

Knocking on the door at 21.

Give Princeton a vote, cowards!

[upprdeck]

9-10 Ariz st is getting votes?  

a couple more weeks off and we can make it in.

[dbilmes]

For what it's worth, the ECAC not doing well in KRATCH rankings. Q is ranked No. 2, while Harvard is 19 and Cornell 22. Dartmouth and Yale are two worst teams in the country.

[Trotsky]

Sisyphus is rolling, again.

18 in USCHO, 17 is USA.

[osorojo]

What criteria are used to construct college hockey rankings, and which ranking system has most successfully predicted the winners of contests between ranked teams?

[Trotsky]

What criteria are used to construct college hockey rankings, and which ranking system has most successfully predicted the winners of contests between ranked teams?

Google PWR, RPICH, KRACH, and PEBCAK.

[jtwcornell91]

What criteria are used to construct college hockey rankings, and which ranking system has most successfully predicted the winners of contests between ranked teams?

Google PWR, RPICH, KRACH, and PEBCAK.

There's some (slightly old, like pre-weird-OT-and-shootout-system) discussion at http://elynah.com/tbrw/2020/rankings.shtml

Is PEBCAK a new rating system, or did you just tell him to google GOATSE?

[ugarte]

grrrrrrrrrrrrrrahhhhhhhhhhggggggggggggghhhhhhhhhhhhhhhhhhhhh

[upprdeck]

if we get a system better at predicting outcomes let me know.. I have some money to spend.

[Trotsky]

Climb every mountain, ford every stream.  15.

Q 1, Harvard 9.
BU 7.

[osorojo]

Which rating system has most accurately predicted D-1 college hockey game winners so far this season?

[BearLover]

Which rating system has most accurately predicted D-1 college hockey game winners so far this season?

Rankings/ratings/polls are not meant to predict future outcomes, ergo no one tracks how predictive they are.