NCAA Rankings 2/19/13

[admiral0417]

Re: NCAA Rankings 2/19/13

What if Williams beats Bowdoin?? Doesn't Bowdoin beat out Eau Claire for last Pool C ?? That would mean an 8-3 split with SNC, SJ and Adrian out west. Oswego, Babson, Mass Dartmouth, Wentworth/salve regina , Utica, Bowdoin, Norwich and Williams. That would mean the end of the line for Hobart, Eau Claire right?

 

[NUProf]

Re: NCAA Rankings 2/19/13

What if Williams beats Bowdoin?? Doesn't Bowdoin beat out Eau Claire for last Pool C ?? That would mean an 8-3 split with SNC, SJ and Adrian out west. Oswego, Babson, Mass Dartmouth, Wentworth/salve regina , Utica, Bowdoin, Norwich and Williams. That would mean the end of the line for Hobart, Eau Claire right?

Bowdoin has no shot at a Pool C bid. Their SOS is among the worst in the country.

 

[PrezdeJohnson09]

Re: NCAA Rankings 2/19/13

First off, these are my thoughts and purely my thoughts on how it will shake out.

I see two scenarios here.

FIRST SCENARIO

Two QF's out West and all three play-in games in the East. With Utica and Norwich both failing to win their conference tournaments, the West has an awful lot of ammunition to fight for no play-in games.

1E. Norwich
2E. Utica
3E. Oswego
4E. Bowdoin
5E. Babson
6E. Wentworth
7E. UMass-Dartmouth

1W. Eau Claire
2W. St. Norbert
3W. Adrian
4W. St. Johns

Play-In Round
UMass-Dartmouth @ Utica
Wentworth @ Oswego
Babson @ Bowdoin

Quarterfinals
Oswego/Wentworth vs. Utica/UMD
Bowdoin/Babson @ Norwich
St. Johns @ Eau Claire
Adrian @ SNC

SECOND SCENARIO

SNC jumps Eau Claire for the No. 1 west seed and instead of forcing SNC to play Adrian in a play-in game, committee sends Adrian to Oswego for a straight quarterfinal. This scenario makes less sense and is a royal screw job to Eau Claire if you ask me.

Play-In Round:
St. John's @ Eau-Claire
Wentworth @ Babson
UMass-Dartmouth @ Bowdoin

Quarterfinals:
St. Johns/Eau Claire @ SNC
Babson/Wentworth @ Norwich
Umass-Dartmouth/Bowdoin @ Utica
Adrian @ Oswego

Had Norwich won, I think this scenario would have been a bit more likely.

If Williams wins tomorrow, simply swap Bowdoin out with Williams and move Babson up to 4E while Williams slots in at 5E.

 

[NUProf]

Re: NCAA Rankings 2/19/13

First off, these are my thoughts and purely my thoughts on how it will shake out.

I see two scenarios here.

FIRST SCENARIO

Two QF's out West and all three play-in games in the East. With Utica and Norwich both failing to win their conference tournaments, the West has an awful lot of ammunition to fight for no play-in games.

1E. Norwich
2E. Utica
3E. Oswego
4E. Bowdoin
5E. Babson
6E. Wentworth
7E. UMass-Dartmouth

1W. Eau Claire
2W. St. Norbert
3W. Adrian
4W. St. Johns

Play-In Round
UMass-Dartmouth @ Utica
Wentworth @ Oswego
Babson @ Bowdoin

Quarterfinals
Oswego/Wentworth vs. Utica/UMD
Bowdoin/Babson @ Norwich
St. Johns @ Eau Claire
Adrian @ SNC

SECOND SCENARIO

SNC jumps Eau Claire for the No. 1 west seed and instead of forcing SNC to play Adrian in a play-in game, committee sends Adrian to Oswego for a straight quarterfinal. This scenario makes less sense and is a royal screw job to Eau Claire if you ask me.

Play-In Round:
St. John's @ Eau-Claire
Wentworth @ Babson
UMass-Dartmouth @ Bowdoin

Quarterfinals:
St. Johns/Eau Claire @ SNC
Babson/Wentworth @ Norwich
Umass-Dartmouth/Bowdoin @ Utica
Adrian @ Oswego

Had Norwich won, I think this scenario would have been a bit more likely.

If Williams wins tomorrow, simply swap Bowdoin out with Williams and move Babson up to 4E while Williams slots in at 5E.

You really think they are going to line it so Norwich gets Babson again? Both your projections have that feature as a likely outcome.

 

[PrezdeJohnson09]

Re: NCAA Rankings 2/19/13

You really think they are going to line it so Norwich gets Babson again? Both your projections have that feature as a likely outcome.

Why not? If we're keeping bracket integrity that's how it should play out. Elmira has played Neumann and Manhattanville in the NCAA Quarterfinals in the last seven years.

 

[UCBadger]

Re: NCAA Rankings 2/19/13

How much longer till "the show?"

 

[CARDS_rule_the_Burgh]

First off, these are my thoughts and purely my thoughts on how it will shake out.

I see two scenarios here.

FIRST SCENARIO

Two QF's out West and all three play-in games in the East. With Utica and Norwich both failing to win their conference tournaments, the West has an awful lot of ammunition to fight for no play-in games.

1E. Norwich
2E. Utica
3E. Oswego
4E. Bowdoin
5E. Babson
6E. Wentworth
7E. UMass-Dartmouth

1W. Eau Claire
2W. St. Norbert
3W. Adrian
4W. St. Johns

Play-In Round
UMass-Dartmouth @ Utica
Wentworth @ Oswego
Babson @ Bowdoin

Quarterfinals
Oswego/Wentworth vs. Utica/UMD
Bowdoin/Babson @ Norwich
St. Johns @ Eau Claire
Adrian @ SNC

SECOND SCENARIO

SNC jumps Eau Claire for the No. 1 west seed and instead of forcing SNC to play Adrian in a play-in game, committee sends Adrian to Oswego for a straight quarterfinal. This scenario makes less sense and is a royal screw job to Eau Claire if you ask me.

Play-In Round:
St. John's @ Eau-Claire
Wentworth @ Babson
UMass-Dartmouth @ Bowdoin

Quarterfinals:
St. Johns/Eau Claire @ SNC
Babson/Wentworth @ Norwich
Umass-Dartmouth/Bowdoin @ Utica
Adrian @ Oswego

Had Norwich won, I think this scenario would have been a bit more likely.

If Williams wins tomorrow, simply swap Bowdoin out with Williams and move Babson up to 4E while Williams slots in at 5E.

Just went through and did this myself, and I came up with something very similar, but not exactly the same...

Babson likely doesn't pass Bowdoin, but UMass-Dartmouth looks like they would slot in ahead of either Wentworth or Salve (against WIT, UMD wins SOS, COP, and RNK, and is only slightly behind in WIN), so the ECAC-NE winner is the 7E seed no matter what.

So, basically, I came to an identical conclusion as you, but swapping UMD and WIT in my brackets.


To clarify, if Williams beats Bowdoin, they would play the games you and I have slotted for Babson, and Babson would play the game we have slotted for Bowdoin, it wouldn't just be a straight swap Williams for Bowdoin.

 

[PrezdeJohnson09]

Re: NCAA Rankings 2/19/13

To clarify, if Williams beats Bowdoin, they would play the games you and I have slotted for Babson, and Babson would play the game we have slotted for Bowdoin, it wouldn't just be a straight swap Williams for Bowdoin.

Correct. Sorry if I was unclear with that.

Alright....I'm going on my 15th hour at the office. Time to go home. Good night USCHO.

 

[CARDS_rule_the_Burgh]

Re: NCAA Rankings 2/19/13

Also, I personally think the "Adrian East" scenario is more likely, although d3hockey and Derek seem to disagree with me. Here's why:

Looking at the comparisons, if I had to make the unofficial national seeds, I would come up with some thing along these lines:
1. Norwich
2. St. Norbert (interchangeable with Norwich)
3. Utica
4. Eau Claire
5. Oswego
6. Adrian
7. Bowdoin
8. Babson
9. St. john's
10. Dartmouth
11. Wentworth

In the Adrian East scenario, we see my #6 Adrian and #5 Oswego get byes over my #4 UWEC, but that is the only part that runs against seeding (and at #5, Oswego would ideally get a bye anyway, so Adrian over UWEC is really the only issue).

In the Adrian West model, we see the #6 Adrian get a bye over #3 Utica and #5 Oswego, as well as the #9 St. john's getting a bye over #3, #5, #7, and #8 team in the list.

So, if we assume these rankings that I just formed (by moving down each regional ranking, making the comparison of criteria and slotting in the winner on the rankings and comparing the loser with the next team of the other region) is even remotely close to what the committee comes up with, in terms of bracket integrity the Adrian East model has far fewer - and less egregious - unwarranted byes. That's why I think that model is the more likely to be seen tomorrow night, even despite what Derek (rightly) pointed out in terms of the West now having a basis to claim 4 byes based on the results in the East

 

[Fishman'81]

Re: NCAA Rankings 2/19/13

Let me supply you with an example to show why it works that way. It is an illogical side effect of the selection criteria. Lets take two mythical teams that somehow have played identical schedules, played each other 4 times during the season and split with each other. They each have records of 20-4-0. They will have identical WIN, OWP, OOWP, COP and H2H records. The only place where they would differ is in RNK. Suppose that team A's other two losses were against ranked teams and team B's other two losses were against unranked teams. If 4 of their games in their identical schedules were against ranked teams then Team A's RNK would turn out to be 0.500 while team B's RNK would be 1.000. Team B wins that criterion and according to the NCAA scheme is a better candidate for post season play than Team A. The moral is that if you are going to lose games, lose them to unranked teams. I agree it is stupid, but that is how it works.

There are lots of things about this process that don't make sense, but what is even worse is the fact that it is the same process that the NCAA uses in all Division 3 sports, and individual sports do not have the option to use different criteria.

The other thing is that according to the NCAA calculation formula Babson is ahead of Hobart in SOS and record vs. ranked teams. The difference in Win percentage is less than you think because the losses against St. Anselm don't count in the primary criteria.

Thanks for taking the time to explain the situation.

One reason (among others) that this process is so frustrating/not-worth-the-time-to-study is that such a situation is even in place... That defies all logic, as you've already stated, and it makes for bad stats.

 

[peddler]

Also, I personally think the "Adrian East" scenario is more likely, although d3hockey and Derek seem to disagree with me. Here's why:

Looking at the comparisons, if I had to make the unofficial national seeds, I would come up with some thing along these lines:
1. Norwich
2. St. Norbert (interchangeable with Norwich)
3. Utica
4. Eau Claire
5. Oswego
6. Adrian
7. Bowdoin
8. Babson
9. St. john's
10. Dartmouth
11. Wentworth

In the Adrian East scenario, we see my #6 Adrian and #5 Oswego get byes over my #4 UWEC, but that is the only part that runs against seeding (and at #5, Oswego would ideally get a bye anyway, so Adrian over UWEC is really the only issue).

In the Adrian West model, we see the #6 Adrian get a bye over #3 Utica and #5 Oswego, as well as the #9 St. john's getting a bye over #3, #5, #7, and #8 team in the list.

So, if we assume these rankings that I just formed (by moving down each regional ranking, making the comparison of criteria and slotting in the winner on the rankings and comparing the loser with the next team of the other region) is even remotely close to what the committee comes up with, in terms of bracket integrity the Adrian East model has far fewer - and less egregious - unwarranted byes. That's why I think that model is the more likely to be seen tomorrow night, even despite what Derek (rightly) pointed out in terms of the West now having a basis to claim 4 byes based on the results in the East

Why do you have Norwich over Utica, when UC seems to now have a 3-2 edge?

 

[93GreatLaker]

Re: NCAA Rankings 2/19/13

I think the West will get screwed just like last year. If last year's formula is used then the West gets two play ins, St Johns at WEC, and Adrian at St Norbert. One play in in the East, Umass at Wentworth (or vice versa).

Quarterfinals would then be West play in winners, East play in winner vs Norwich, Oswego hosting Bowdoin, and Utica hosting Babson. (If Williams wins, Babson goes to Oswego and Williams goes to Utica).

 

[Fishman'81]

Re: NCAA Rankings 2/19/13

Actually, no.

WinPct:
Hobart 0.769
Babson 0.705

SoS -- Not exactly the NCAA figures but very close.
Hobart: 0.5336
Babson: 0.5432

RvR
Hobart: 5-5-1
Babson: 6-2-3

Depends on how you calculate SOS, but yes, the two teams are close in that regard by any reasonable measure, with Hobart coming out on top of the comparison in any logical way you care to slice it. (Excepting the very curious NCAA-way of doing things, of course.)

That's all irrelevant now, with Babson getting an AQ. Hobart's very impressive season is unofficially over, and UMass-Dartmouth will play on...

That fact alone is evidence enough that the process needs some very serious tweaking.

 

[Fishman'81]

Re: NCAA Rankings 2/19/13

I will only agree with Wich. That is just my jut feeling. The men in the smoking room will leave us scratching our heads as usual.
Oswego 88

Norwich and Utica are obvious locks, IMO... (Babson/Hobart would've been something to argue about, maybe, if the E didn't have an AQ.)

 

[jerrynu26]

Re: NCAA Rankings 2/19/13

Let me supply you with an example to show why it works that way. It is an illogical side effect of the selection criteria. Lets take two mythical teams that somehow have played identical schedules, played each other 4 times during the season and split with each other. They each have records of 20-4-0. They will have identical WIN, OWP, OOWP, COP and H2H records. The only place where they would differ is in RNK. Suppose that team A's other two losses were against ranked teams and team B's other two losses were against unranked teams. If 4 of their games in their identical schedules were against ranked teams then Team A's RNK would turn out to be 0.500 while team B's RNK would be 1.000. Team B wins that criterion and according to the NCAA scheme is a better candidate for post season play than Team A. The moral is that if you are going to lose games, lose them to unranked teams. I agree it is stupid, but that is how it works.

There are lots of things about this process that don't make sense, but what is even worse is the fact that it is the same process that the NCAA uses in all Division 3 sports, and individual sports do not have the option to use different criteria.

The other thing is that according to the NCAA calculation formula Babson is ahead of Hobart in SOS and record vs. ranked teams. The difference in Win percentage is less than you think because the losses against St. Anselm don't count in the primary criteria.

I am sure I will regret questioning your results, but how could team b be 1.000? If thet are being compared, they have one loss to a ranked team A, or did I miss the boat?

 

[CARDS_rule_the_Burgh]

That's all irrelevant now, with Babson getting an AQ. Hobart's very impressive season is unofficially over, and UMass-Dartmouth will play on...

That fact alone is evidence enough that the process needs some very serious tweaking.

And yet, there are some who truly feel that if you can't prove you're the champion of your own conference, you shold have no business calling yourself the the champion of all DIII Men's Ice Hockey, and therefore they want to get rid of the Pool B/C bids... So you want to get rid of the AQs, and others want to get rid of the At-Large bids. I guess the only way to make everyone happy is to just not have a tournament and just declare a national champion in early October!

But in all seriousness, there was a lot of frumbling when the ECAC-NE first got ita AQ, most have just accepted that it's reality at this point. But, as has been said here a number of time, the NCAA can do nothing to change the fact that the AQs are based in the conference postseason tournaments. They didn't decide that, the conferences did. The NCAA makes the rules for which conferences get an AQ, the conference makes the rules for how a team earns it... It just happens to be the case that every single conference in every team sport with a conference tournament uses that tournament to decide it. And that's very much a money issue, since in some cases the conference tournament would attract significantly less attention if there were no AQ on the line or if teams had already earned their AQ in the RS, and that same phenomenon would also apply to every RS game a team plays after locking up the RS title and the autobid. That's not a universal phenomenon, bur all the conferences use the AQ to prevent it from even being a possibility.

Ultimately, I agree there are potential flaws in the system, but not everybody can agree on what those flaws are specifically, and so there really is no way to find a perfect way to name a national champion. As stupid as some of the aspects of the system are, I think overall this system does a decent job at trying to balance the 2 viewpoints...

Just my opinion, of course.

 

[CARDS_rule_the_Burgh]

Why do you have Norwich over Utica, when UC seems to now have a 3-2 edge?

Mostly because I was trying to take a stab at predicting how the committee would do it, and a part of me feels the committee would still put the cadets in front. It's actually a 3-1 win for UC, but it was when the committee put Norwich #1 last week, too!

I mean, I suppose you could put those first 3 in any order. The relative positioning of NRW and UTC would affect the bracketing (swap the two in my bracket predictions if you choose to predict that UTC will be the 1E, not the 2E.)

 

[bakdraft21]

Re: NCAA Rankings 2/19/13

I think the West will get screwed just like last year. If last year's formula is used then the West gets two play ins, St Johns at WEC, and Adrian at St Norbert. One play in in the East, Umass at Wentworth (or vice versa).

Quarterfinals would then be West play in winners, East play in winner vs Norwich, Oswego hosting Bowdoin, and Utica hosting Babson. (If Williams wins, Babson goes to Oswego and Williams goes to Utica).

oswego vs bowdin,lets hope we have the same results in the past few yrs if that happens

 

[polarbearfan]

Re: NCAA Rankings 2/19/13

oswego vs bowdin,lets hope we have the same results in the past few yrs if that happens

Lets not!! Wouldn't mind seeing Oswego though, if only because it would mean the Polar Bears were dancing.

 

[NUProf]

Re: NCAA Rankings 2/19/13

I am sure I will regret questioning your results, but how could team b be 1.000? If thet are being compared, they have one loss to a ranked team A, or did I miss the boat?

No I missed a detail that you picked up

I did this in a bit of a hurry, because I forgot that Team A and Team B in my scenario would be ranked. You caught me, however the basic principle still applies. If all your loses are to Ranked teams, you are worse off than if those loses had been to unranked teams.