PWR volatile, Cornell -> #8t

[Al DeFlorio]

I think Cornell, Clarkson, and Vermont made it in 96 or thereabouts.

[DeltaOne81]

And Yale, Princeton, and Clarkson in 98. Man, that sounds really weird. Now, these were 12 team tournaments, but there were also fewer teams around.

[jy3]

well cornell was up to 6th but then we lost the UNO comparison. that one will be interesting down the road. UNO has 4 games left against TUCs and cornell has plenty - that is where the comparison will likely be decided is TUC record unless one team or the other has a largely changed RPI. No more COP left for either team. should be interesting

[jkahn]

As of Sunday, 1/23 we're up to #6.  We were still #8 until Colgate lost, which flipped that comparison, moving us up to a tie for sixth wth Harvard, and also causing 'gate to lose some other comparisons.  Later, we won back the UNO comparison when Ferris St. ceased being a TUC by losing at Fairbanks, thus moving us one comparison ahead of Harvard.  Earlier in the evening, the UNO comparison had flipped the other way, when UMass had dropped out of being a TUC - and since that eliminated a loss from UNO's TUC record, it had given them the comparison for awhile.  Just another example of how poor PWR is - if a team you lost to (i.e. UMass in UNO's case) winds up being considered as a weaker team (non-TUC), then the loss doesn't count as much.
I shouldn't complain too much about PWR though when we're #6.

[Greg Berge]

We're 6 in PWR and 9 in KRACH.  The objective conculsion is that PWR is a better measure. :-)

[DeltaOne81]

[Q]Greg Berge Wrote:

 We're 6 in PWR and 9 in KRACH.  The objective conculsion is that PWR is a better measure.  [/q]
This brings up an interesting topic. In the past, didn't KRACH tend to reflect better on the ECAC than PWR? Our 'excuse' was just that PWR, or at least RPI, overweights S.O.S.

But this year, PWR is doing us better, even though the WCHA and HEA still definitey have the higher S.O.S, PWR still has us above Wisc, UND, and UNH, while KRACH does not.  So now it seems that argument is out. Anyone so inclined care to take a shot at 'what gives'?

[KeithK]

That's not how I remember it.  Over the last few years it seems to me that KRACH rates WCHA and HE teams very highly, more so than PWR>  The last place team in the WCHA usually was ahead of the middle of the middle of the pack of the ECAC.

Here's an interesting game if someone wants to waste large amounts of computer time and has access to PWR generating code.  Take the season scores list from USCHO.  Calculate a new PWR after each game completes (using start times plus game time from the box score, or efor simplicity assuming that all games take an equal amount of time) and stpre the results for each team with date/time tags.  Then plot the PWR as a function of time, showing us the beauty of the crazy fluctuations.

[KenP]

Next week we play Clarkson and St Lawrence for the first time.  My thoughts as to how that will impact PWR:

RPI will be hurt slightly
Clarkson lost to Ferris, UNH and UML.  They beat Ohio State (1-0-1).  Anthing less than a series sweep will hurt the COP comparison for the first 3 comparisons.
SLU(t) has a win against MSU but losses to UNH and UML.  Again, this is an important game for these COP comparisons.
SLU(t) is also a TUC (and will likely remain one), despite a very tough remaining season.

What's the morale other than "wins are good / losses are bad"?  I'm not sure.

[kaelistus]

As far as I know, the ECAC has always (NOTE: History for me began on 1995) done better at PWR than KRACH.

ECAC bias has never been an argument for or against those two.