TonyTheTiger20
#SOAR
Re: 2015-2016 D-I Polls and Rankings
Lots of respect for NU this week
Lots of respect for NU this week
USCHO Fan Forum
This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!
-
The USCHO Fan Forum has migrated to a new plaform, xenForo. Most of the function of the forum should work in familiar ways. Please note that you can switch between light and dark modes by clicking on the gear icon in the upper right of the main menu bar. We are hoping that this new platform will prove to be faster and more reliable. Please feel free to explore its features.
2015-2016 D-I Polls and Rankings
- Thread starter ARM
- Start date
Re: 2015-2016 D-I Polls and Rankings
Then....
And now....
February 22, 2016
1 Boston College (15) 34- 0-0 150 1
2 Minnesota 29- 3-1 134 3
3 Wisconsin 30- 3-1 121 2
4 Quinnipiac 26- 2-5 104 4
5 Clarkson 26- 3-5 91 5
6 Northeastern 26- 7-1 70 6
7 Bemidji State 22- 9-3 60 7
8 Princeton 21- 6-2 48 8
9 Colgate 20- 7-7 25 10
10 North Dakota 16-11-5 16 9
Others receiving votes: Harvard 6.
No question that Harvard was the most over-rated team, having dropped from #4 in the preseason poll to the "Others receiving votes" category in the final regular season poll.
But for them the good news is: the season's not over yet!
Then....
And now....
February 22, 2016
1 Boston College (15) 34- 0-0 150 1
2 Minnesota 29- 3-1 134 3
3 Wisconsin 30- 3-1 121 2
4 Quinnipiac 26- 2-5 104 4
5 Clarkson 26- 3-5 91 5
6 Northeastern 26- 7-1 70 6
7 Bemidji State 22- 9-3 60 7
8 Princeton 21- 6-2 48 8
9 Colgate 20- 7-7 25 10
10 North Dakota 16-11-5 16 9
Others receiving votes: Harvard 6.
No question that Harvard was the most over-rated team, having dropped from #4 in the preseason poll to the "Others receiving votes" category in the final regular season poll.
But for them the good news is: the season's not over yet!
HockeyEast33
New member
Re: 2015-2016 D-I Polls and Rankings
I honestly don't understand how Harvard can still be receiving top ten votes. They lost 10 games including a key last game to Cornell, a team that is barely .500......the voters are lost in potential vs. performance...
Then....
And now....
February 22, 2016
1 Boston College (15) 34- 0-0 150 1
2 Minnesota 29- 3-1 134 3
3 Wisconsin 30- 3-1 121 2
4 Quinnipiac 26- 2-5 104 4
5 Clarkson 26- 3-5 91 5
6 Northeastern 26- 7-1 70 6
7 Bemidji State 22- 9-3 60 7
8 Princeton 21- 6-2 48 8
9 Colgate 20- 7-7 25 10
10 North Dakota 16-11-5 16 9
Others receiving votes: Harvard 6.
No question that Harvard was the most over-rated team, having dropped from #4 in the preseason poll to the "Others receiving votes" category in the final regular season poll.
But for them the good news is: the season's not over yet!
I honestly don't understand how Harvard can still be receiving top ten votes. They lost 10 games including a key last game to Cornell, a team that is barely .500......the voters are lost in potential vs. performance...
Re: 2015-2016 D-I Polls and Rankings
Considering they are 2-0 against #8 and 1-1 against #9 I highly doubt it. As for #7, western posters would make them an underdog and I would hope eastern posters make them a favorite. I know I would make them a favorite
.
Considering they are 2-0 against #8 and 1-1 against #9 I highly doubt it. As for #7, western posters would make them an underdog and I would hope eastern posters make them a favorite. I know I would make them a favorite
.
ARM
Fan of chipmunk-like mascots.
Re: 2015-2016 D-I Polls and Rankings
I think it would be a coin flip. When I've watched Clarkson, it will dominate territorially, but not always have much in the way of goals to show for it. Clarkson would have the majority of the puck possession against Bemidji, but the Beavers are used to that. Where the Golden Knights would have to be careful is that they like to get their D involved around the offensive net, and BSU has some speed that can burn you going the other way. If Tiley is playing well, you'd probably be okay. One way or another, I'd expect a 2-1, 2-0, or 1-0 type game.
Re: 2015-2016 D-I Polls and Rankings
vicb... I wrote that about Clarkson in NOVEMBER. Their first half of the year was unimpressive. No one has made a bigger turnaround from H1 to H2 than Clarkson. I've got them at #5 and I'm confident they can beat anyone from #7 on in a 3 game series.
Considering they are 2-0 against #8 and 1-1 against #9 I highly doubt it. As for #7, western posters would make them an underdog and I would hope eastern posters make them a favorite. I know I would make them a favorite
vicb... I wrote that about Clarkson in NOVEMBER. Their first half of the year was unimpressive. No one has made a bigger turnaround from H1 to H2 than Clarkson. I've got them at #5 and I'm confident they can beat anyone from #7 on in a 3 game series.
TonyTheTiger20
#SOAR
Who do you replace them with? Colgate is the only one in the discussion and Colgate is behind Harvard in KRACH and is 0-2 against Harvard.
Re: 2015-2016 D-I Polls and Rankings
Oh... and my #6 is Bemidji. I'd love to see Clarkson vs BSU.
Oh... and my #6 is Bemidji. I'd love to see Clarkson vs BSU.
ARM
Fan of chipmunk-like mascots.
Re: 2015-2016 D-I Polls and Rankings
Anyway, the Colgate/Harvard question gets answered definitively on the ice this weekend. Until it is, let Colgate represent the placeholder.
Are you remembering to impose the non-quality result decrement on Harvard?
Anyway, the Colgate/Harvard question gets answered definitively on the ice this weekend. Until it is, let Colgate represent the placeholder.
TonyTheTiger20
#SOAR
I think it was answered pretty definitively by Colgate's non-quality 0-2 results against Harvard this season, one of which was literally this weekend.
If Colgate manages to suddenly sweep Harvard this weekend, then they deserve the spot, but as it is we have two head to head results to use here. Given that every ranking system we have has the two teams close, a worst, and Harvard ahead, at best, then I think H2H is pretty useful here, no?
0-2 AGAINST HARVARD THO
Last edited:
ARM
Fan of chipmunk-like mascots.
Re: 2015-2016 D-I Polls and Rankings
If Colgate wins two of three over Harvard this weekend, do you mean to tell me you're going to keep voting for Harvard because H2H 3>2? I'm saying for this one week, it is pointless to argue about it, because next weekend in Hamilton trumps everything that went before.
pokechecker
Banned
Re: 2015-2016 D-I Polls and Rankings
OK folks, we have the perfect opportunity to explain how the RPI is valid (for those that think it is) or invalid (for those who think it isn’t). WI & MN play the exact same conference schedule and finish with the same record. And they both lost & tied against UND, while splitting with each other. One would think based upon conference play that both should be ranked exactly the same, regardless of the ranking system used.
Now compare the non conference schedule:
Both teams beat all of their opponents. But it appears MN played the tougher schedule based on the opponents records. The SOS seems to agree with that. So how then, is WI ranked higher (even if only slightly so) than MN?
FYI, the opponents, note that they both played two games against an ECAC team, and two against a CHA team, conference record first, all games record second:
WI=
Dartmouth 6-13-3, 6-19-3
Lindenwood 5-11-1, 8-22-4
Providence 6-16-2, 10-22-2
MN=
Yale 9-11-2, 10-17-2
Penn State 6-8-6, 10-18-6
St. Cloud 9-15-4, 13-16-4
Yale appears to be a tougher opponent than Dartmouth, and Penn State appears to be better than Lindenwood. This is confirmed by SOS numbers as well.
And St. Cloud appears to be better than Providence, also confirmed by SOS.
So you would think MN would have the higher RPI & Pairwise.
How can any rating system that places WI above MN be valid given the above?
The argument here seems to be:
how should Princeton, Harvard, Colgate, or UND be rated with respect to one nother.
When you look at RPI/Pairwise, clearly, teams that play a tougher schedule are penalized in relation to teams that play a weaker schedule.
Of course this isn't news to anybody, least of all college coaches.
Which explains their reluctance to play anybody not a cupcake.
Why would they not want to fix a system that is obviously broken?
They complain nobody comes to watch women's hockey,
how about giving people a reson to?
OK folks, we have the perfect opportunity to explain how the RPI is valid (for those that think it is) or invalid (for those who think it isn’t). WI & MN play the exact same conference schedule and finish with the same record. And they both lost & tied against UND, while splitting with each other. One would think based upon conference play that both should be ranked exactly the same, regardless of the ranking system used.
Now compare the non conference schedule:
Both teams beat all of their opponents. But it appears MN played the tougher schedule based on the opponents records. The SOS seems to agree with that. So how then, is WI ranked higher (even if only slightly so) than MN?
FYI, the opponents, note that they both played two games against an ECAC team, and two against a CHA team, conference record first, all games record second:
WI=
Dartmouth 6-13-3, 6-19-3
Lindenwood 5-11-1, 8-22-4
Providence 6-16-2, 10-22-2
MN=
Yale 9-11-2, 10-17-2
Penn State 6-8-6, 10-18-6
St. Cloud 9-15-4, 13-16-4
Yale appears to be a tougher opponent than Dartmouth, and Penn State appears to be better than Lindenwood. This is confirmed by SOS numbers as well.
And St. Cloud appears to be better than Providence, also confirmed by SOS.
So you would think MN would have the higher RPI & Pairwise.
How can any rating system that places WI above MN be valid given the above?
The argument here seems to be:
how should Princeton, Harvard, Colgate, or UND be rated with respect to one nother.
When you look at RPI/Pairwise, clearly, teams that play a tougher schedule are penalized in relation to teams that play a weaker schedule.
Of course this isn't news to anybody, least of all college coaches.
Which explains their reluctance to play anybody not a cupcake.
Why would they not want to fix a system that is obviously broken?
They complain nobody comes to watch women's hockey,
how about giving people a reson to?
Last edited:
Re: 2015-2016 D-I Polls and Rankings
I can try to explain just how RPI arrives at this result, though with the caveat that I haven't played around with the numbers at all. This explanation doesn't improve my opinion of the validity of RPI and PWR. Start with what I wrote in the Scores & Results thread over the weekend:
What happens is that once the process discards those games that would lower each team's RPI, Minnesota's strength of schedule advantage is eliminated. There's no way to know exactly which games are being eliminated from the calculations without recalculating it for each possibility, stopping when all of the remaining wins raise the RPI, but my guess is that both teams' games against Minnesota State and maybe Ohio State are removed, along with all of Wisconsin's nonconference games and maybe Minnesota's games against Penn State.
When that's done, the two team's RPI strength of schedule components are almost identical. Wisconsin might even have an edge, depending upon exactly where the cutoff is. My assumptions above would leave either Penn State or Yale as the weakest nonconference team left in either's calculation, giving the Badgers a higher value. On top of that, because Wisconsin played one more nonconference game than Minnesota did, it has a higher winning percentage. Add it all together, and Wisconsin has an RPI for PWR purposes that is greater by such a small margin that it would be a rounding error with one fewer decimal places.
I can try to explain just how RPI arrives at this result, though with the caveat that I haven't played around with the numbers at all. This explanation doesn't improve my opinion of the validity of RPI and PWR. Start with what I wrote in the Scores & Results thread over the weekend:
What happens is that once the process discards those games that would lower each team's RPI, Minnesota's strength of schedule advantage is eliminated. There's no way to know exactly which games are being eliminated from the calculations without recalculating it for each possibility, stopping when all of the remaining wins raise the RPI, but my guess is that both teams' games against Minnesota State and maybe Ohio State are removed, along with all of Wisconsin's nonconference games and maybe Minnesota's games against Penn State.
When that's done, the two team's RPI strength of schedule components are almost identical. Wisconsin might even have an edge, depending upon exactly where the cutoff is. My assumptions above would leave either Penn State or Yale as the weakest nonconference team left in either's calculation, giving the Badgers a higher value. On top of that, because Wisconsin played one more nonconference game than Minnesota did, it has a higher winning percentage. Add it all together, and Wisconsin has an RPI for PWR purposes that is greater by such a small margin that it would be a rounding error with one fewer decimal places.
pokechecker
Banned
Re: 2015-2016 D-I Polls and Rankings
I said SOS when I meant RPI/Pairwise
comparing the non conference opponents, each of MN opponents are ranked higher than the comparable WI opponent using RPI/Pairwise
so The ranking system itself says that Yale is better than Dartmouth, Penn State is better than Lindenwood, and St. Cloud is better than Providence
common sense would say that if the ranking system ranks the opponents of team A higher than team B, if the ranking system is valid, then team A in that same ranking system would be ranked higher than team B
if it isn't there is something wrong with the ranking system.
normally if A> B
and C>D
and E>F
then it can be concluded A+C+E> B+D+F
but the RPI/Pairwise says, with some voodoo math apparently:
If A> B
and C> D
and E>F
then A+C+E is slightly less than B+C+F
the math doesn't add up
I said SOS when I meant RPI/Pairwise
comparing the non conference opponents, each of MN opponents are ranked higher than the comparable WI opponent using RPI/Pairwise
so The ranking system itself says that Yale is better than Dartmouth, Penn State is better than Lindenwood, and St. Cloud is better than Providence
common sense would say that if the ranking system ranks the opponents of team A higher than team B, if the ranking system is valid, then team A in that same ranking system would be ranked higher than team B
if it isn't there is something wrong with the ranking system.
normally if A> B
and C>D
and E>F
then it can be concluded A+C+E> B+D+F
but the RPI/Pairwise says, with some voodoo math apparently:
If A> B
and C> D
and E>F
then A+C+E is slightly less than B+C+F
the math doesn't add up
Last edited:
Re: 2015-2016 D-I Polls and Rankings
That's because you missed my point. I'll try again. If you look at the raw RPI, you'll see that everything you say is correct. The issue is that once you remove the games in which a win for a team lowers its RPI, none of your comparison of nonconference strength of schedule is in any way relevant to the question. As far as PWR is concerned, Wisconsin never played Lindenwood, Dartmouth, or Providence. What happens is that Minnesota's strength of schedule is lowered below that of Wisconsin, because the nonconference games that remain in its calculation have a weaker SoS component than its average games do, so while their RPI is overall helped by the additional win more than it's lowered by the SoS factor, it still leaves Minnesota with a lower SoS relative to Wisconsin than your analysis shows.
That's because you missed my point. I'll try again. If you look at the raw RPI, you'll see that everything you say is correct. The issue is that once you remove the games in which a win for a team lowers its RPI, none of your comparison of nonconference strength of schedule is in any way relevant to the question. As far as PWR is concerned, Wisconsin never played Lindenwood, Dartmouth, or Providence. What happens is that Minnesota's strength of schedule is lowered below that of Wisconsin, because the nonconference games that remain in its calculation have a weaker SoS component than its average games do, so while their RPI is overall helped by the additional win more than it's lowered by the SoS factor, it still leaves Minnesota with a lower SoS relative to Wisconsin than your analysis shows.
Re: 2015-2016 D-I Polls and Rankings
Maybe. It all depends upon exactly which games are removed. It's possible that there are more wins removed for Wisconsin than there are for Minnesota, and so Minnesota ends up with a higher winning percentage but that this is not enough to compensate for Wisconsin's improved SoS.
Maybe. It all depends upon exactly which games are removed. It's possible that there are more wins removed for Wisconsin than there are for Minnesota, and so Minnesota ends up with a higher winning percentage but that this is not enough to compensate for Wisconsin's improved SoS.