PWR Pedagogery

Rosey · February 22, 2010, 01:37:30 PM

Quote from: Jim Hyla
Quote from: Kyle Rose
Quote from: billhowardWould be a sad way to mark the 40th anniversary of the 29-0 season.
Especially with this lineup. What a waste of a graduating (and likely early-departure) class, followed next year by what is sure to be an ugly rebuilding season.
But they said that about SU's BB from last year till this, and look at what happened there.

I never give up all hope, but I just look at who's (definitely+likely) leaving and who's staying, and an awful lot of underclassmen will have to step up big-time to fill the void. It's just really too bad that this team couldn't get it together: on paper, they're at least as good as '03, but they have been unable to get over their inconsistency.

Jim Hyla · February 22, 2010, 02:01:02 PM

Quote from: Kyle Rose
Quote from: Jim Hyla
Quote from: Kyle Rose
Quote from: billhowardWould be a sad way to mark the 40th anniversary of the 29-0 season.
Especially with this lineup. What a waste of a graduating (and likely early-departure) class, followed next year by what is sure to be an ugly rebuilding season.
But they said that about SU's BB from last year till this, and look at what happened there.
I never give up all hope, but I just look at who's (definitely+likely) leaving and who's staying, and an awful lot of underclassmen will have to step up big-time to fill the void. It's just really too bad that this team couldn't get it together: on paper, they're at least as good as '03, but they have been unable to get over their inconsistency.

I don't disagree, I was responding to the "what is sure to be an ugly rebuilding season.". I'm just saying it doesn't have to be.

mikek · February 22, 2010, 02:15:43 PM

Quote from: Jim Hyla
Quote from: ebilmesAm I reading this correctly that Cornell only has a 22% chance of making the NCAA tournament, according to their statistical projections?

http://playoffstatus.com/ncaahockey/ncaahockeytournpartprob.html
Now we are down to 20%, but the sig. stat is no (zero) chance if we lose the conf. tourney. But that assumes all teams win and lose games as expected.

That 0% seems a little fishy. We have a better chance of making the tourney if we win based on expected performance than if we win all games and then lose the con. final (to Yale presumably)?

ugarte · February 22, 2010, 03:18:22 PM

Quote from: mikek
Quote from: Jim Hyla
Quote from: ebilmesAm I reading this correctly that Cornell only has a 22% chance of making the NCAA tournament, according to their statistical projections?

http://playoffstatus.com/ncaahockey/ncaahockeytournpartprob.html
Now we are down to 20%, but the sig. stat is no (zero) chance if we lose the conf. tourney. But that assumes all teams win and lose games as expected.

That 0% seems a little fishy. We have a better chance of making the tourney if we win based on expected performance than if we win all games and then lose the con. final (to Yale presumably)?

I think it means that we have 0% chance if we fail to win out if every non-Cornell game goes according to expectations. It is possible that a combination of non-Cornell results could provide a cushion for us to get into the tournament without winning ECACs.

Right?

mikek · February 22, 2010, 03:54:56 PM

Quote from: ugarte
Quote from: mikek
Quote from: Jim Hyla
Quote from: ebilmesAm I reading this correctly that Cornell only has a 22% chance of making the NCAA tournament, according to their statistical projections?

http://playoffstatus.com/ncaahockey/ncaahockeytournpartprob.html
Now we are down to 20%, but the sig. stat is no (zero) chance if we lose the conf. tourney. But that assumes all teams win and lose games as expected.

That 0% seems a little fishy. We have a better chance of making the tourney if we win based on expected performance than if we win all games and then lose the con. final (to Yale presumably)?
I think it means that we have 0% chance if we fail to win out if every non-Cornell game goes according to expectations. It is possible that a combination of non-Cornell results could provide a cushion for us to get into the tournament without winning ECACs.

Right?

The way I read it is we have a 20% chance if every game including Cornell's goes to expectation. If we win out and then lose in the final and all other games go to expectation then we have a 0% chance. Seems to suggest we'd be better off losing 2 games than 1... Could it be a PWR quirk that beating Union twice (I imagine that would be the expected semi matchup) actually hurts their value so much (maybe eliminates them as a TUC) that it drags us down?

ursusminor · February 22, 2010, 03:58:12 PM

There has got to be something wrong with the algorithms used on that site. On this page http://playoffstatus.com/ecachockey/ecacmig.html the ECAC games being played next Friday are ranked as to importance in the standings on a 0 to 100 basis. Somehow the RPI-Colgate game received a score of 114. ::rolleyes::

Jim Hyla · February 22, 2010, 05:20:12 PM

I think it means that we are expected to beat all the teams we will play except Yale and maybe toss-up with Union. Therefore we have good probability of winning all the games up to Yale in the ECAC finals. If we win all of those, but lose to Yale we can't go. However, we also have a chance of losing one of those games up to the finals, therefore with our expected probabilities we have a 20% chance of going. For example, we lose to Union but then win the tourney, we go. But if we lose to Union, the chance of winning the tourney lessens, thus the 20%.

Yale is in a better shape because they are more likely to beat the teams they are playing than we are. Also their current position is such that even if the lose the finals they go. However if they lose a game before the finals it drops their chance to 66%.

Maybe?::dribble::

marty · February 23, 2010, 08:56:54 AM

Quote from: ursusminorThere has got to be something wrong with the algorithms used on that site. On this page http://playoffstatus.com/ecachockey/ecacmig.html the ECAC games being played next Friday are ranked as to importance in the standings on a 0 to 100 basis. Somehow the RPI-Colgate game received a score of 114. ::rolleyes::

Maybe someone at Colgate divided by zero? An engineer would never make such an error.

kaelistus · February 23, 2010, 01:30:36 PM

All those numbers look very iffy to me. I don't trust the math that determines the Cornell vs Union game is the least important game of next week.

Trotsky · February 23, 2010, 01:56:18 PM

Quote from: kaelistusAll those numbers look very iffy to me. I don't trust the math that determines the Cornell vs Union game is the least important game of next week.

That could be because it's only important for 2 or 3 teams, while games in the middle of the pack can affect 8 or 9.

ursusminor · February 24, 2010, 06:10:12 AM

Quote from: Trotsky
Quote from: kaelistusAll those numbers look very iffy to me. I don't trust the math that determines the Cornell vs Union game is the least important game of next week.
That could be because it's only important for 2 or 3 teams, while games in the middle of the pack can affect 8 or 9.

I suspect that is correct, and also the teams can't move far in the standings based upon the result.

Jim Hyla · February 24, 2010, 07:35:46 AM

Quote from: ebilmesAm I reading this correctly that Cornell only has a 22% chance of making the NCAA tournament, according to their statistical projections?

http://playoffstatus.com/ncaahockey/ncaahockeytournpartprob.html

Hey, we're up to 24% and <1% if we lose conf tourney. At least it's not 0 like it was.

mikek · February 27, 2010, 01:09:58 PM

Up to #10 after beating Union Friday. Let's hope we can avoid the Saturday let down this weekend. It being senior night tonight I don't see intensity or motivation being a problem though.

Not many comparisons out there left to win though. BC, UNH, and CC are the only ones we have a shot at.

7. BC - we need to get to 10 TUC and beat their record (currently .500, our record would be .500 with one more win).

8. ND - Probably can't win this baring a collapse and at least 2 losses to CC.

9. UNH - Lots of possibilities are in play with UNH. Their rpi is reachable but we need them to lose some games. TUCs could come into play if we get 2 more (Union in semis and Yale in final is the only way). We would have to win both (making this moot as we would have won the ECAC, but I'll go on anyway) and have them lose at least 2 TUC games. We can even up COp with a win vs. RPI tonight and take the comparison with one positive COp result in either the HE or ECAC playoffs. We have H2H so we just need one of those comparisons to go the other way.

BC and UNH play 2 next week so our rooting interest in that should be interesting.

16. CC - To beat CC we need them to lose to ND tonight which would give us the COp comparison and probably allow us to pass them in rpi if we beat RPI.

Things below us get pretty complicated if we get to 10 TUC games. Lots of teams are sitting at .500 or just 1 game above in TUC which is probably just out of our range. We could get to 5-4-2 but that would mean we got the autobid. For non title scenarios we could only get to .500 if we beat Union in the semis and lost to someone other than Yale in the championship. We hold the COp in a lot of those close TUC matchups right now so we could survive losing the TUC comparisons, but depending on playoff matchups things could go the other way. RPI seems to be a COp in many of our close matchups so a win tonight is crucial for both COp and maintaining our rpi.

If we lost to RPI tonight we would likely lose the Mich St. and Alaska comparisons if TUC becomes a factor.

We also stand to lose the Vermont comparison if TUC becomes a factor. We could win back the COp comparsion with the right playoff matchup and results. We either need them to lose to BU on Sunday or if they beat BU then have BU fall apart and no longer be a TUC.

Still lots of different ways for things to shake out depending on what side of 10 TUC games we fall on. As usual just winning is the simplest solution.

Jim Hyla · February 27, 2010, 02:03:30 PM

Thanks again. Please lets us know which teams we should cheer for next weekend.::cheer::

Robb · February 27, 2010, 07:25:26 PM

Good - I nominate mikek to do the honors in the morning. I'm staying up to watch the game tonight (dammit!) after having worked approximately 75 hours this week, and at least another 12 to go tomorrow. Big work deadline on Monday...

In the meantime, here's all the wisdom I can muster: winning = good. losing = bad.

LGR!