Too early for the PWR? Princeton and Brown say no!

[Priceless]

Re: Too early for the PWR? Princeton and Brown say no!

It's probably because Bemidji's actual RPI, when not rounding to the nearest 10,000th, is actually below .5000.

.4999637 Now, is that considered close enough for the Committee, or do they not round up?

[FlagDUDE08]

Re: Too early for the PWR? Princeton and Brown say no!

.4999637 Now, is that considered close enough for the Committee, or do they not round up?

Personally I would truncate, but that's just me.

[ScoobyFanClub]

Re: Too early for the PWR? Princeton and Brown say no!

TBRW breaks ties using the winner of the comparison, SiouxSports breaks ties using RPI. The NCAA has never said how they break ties between two teams and only once has the RPI and comparisons conflicted; that was in 2005 and the teams had to be switched to avoid a conference meeting (UNH-BU) anyway. I have asked the NCAA for clarification of the tie-breaking procedure, which I expect them to answer around the time the Sun burns out. For now, I will break ties between two teams using the comparison.

It's happened more than once, and it's been written about extensively elsewhere. It's clearly broken with RPI. Not sure why this is even disputed anymore.

[Priceless]

Re: Too early for the PWR? Princeton and Brown say no!

It's happened more than once, and it's been written about extensively elsewhere. It's clearly broken with RPI. Not sure why this is even disputed anymore.

It has? When? Where has it been written about? I emailed Jayson Moy to ask him because last year he broke ties using comparisons, not RPI. In that bracketology, Maine and CC were tied for 7th. Maine won the comparison but CC had an RPI .0101 higher than Maine. He broke the tie in Maine's favor. This year he breaks all ties using RPI.

I have looked at the final PWR of every season since 2003 and I have found only one instance of head to head ties where the RPI and comparison conflicted. If it's been addressed by the NCAA I'd love to see it, because they have stonewalled me when I've asked. It seems to me that if they had already ruled on it they could just cut and paste the answer.

[ScoobyFanClub]

Re: Too early for the PWR? Princeton and Brown say no!

It has? When? Where has it been written about?

I use this URL to find Pairwise articles at CHN - which can be helpful when trying to remember all this stuff
http://www.collegehockeynews.com/new...hp?kw=pairwise

So - going back through - I believe 2007, there was a conflict:
http://www.collegehockeynews.com/new..._final2007.php

There was a 4-way tie in comparison wins at No. 11 last year ... and the tiebreaker was discussed then too:
http://www.collegehockeynews.com/new...20_bracket.php

[Priceless]

Re: Too early for the PWR? Princeton and Brown say no!

I use this URL to find Pairwise articles at CHN - which can be helpful when trying to remember all this stuff
http://www.collegehockeynews.com/new...hp?kw=pairwise

So - going back through - I believe 2007, there was a conflict:
http://www.collegehockeynews.com/new..._final2007.php

There was a 4-way tie in comparison wins at No. 11 last year ... and the tiebreaker was discussed then too:
http://www.collegehockeynews.com/new...20_bracket.php

When three or more teams tie they use RPI to break the tie because each team could win a comparison against another. However, when there is a tie between two teams then only one can win the comparison. In that scenario the tie-breaking procedure hasn't been explained. There is still a question whether head to head ties are broken via comparisons or RPI. That situation has only happened once and it was resolved by switching seeds to avoid an intra-conference game between UNH and BU. In the most recent discussion on CHN, Adam lists Merrimack fourth. As I linked below, Jayson used comparisons to break ties last year:

We break ties in the PWR by looking at the individual comparisons among the tied teams, and add in any current league leaders that are not currently in the top 16. The only team that is not is RIT.

So the question remains: when two teams are tied and there is a disagreement between RPI and the comparison how is the tie broken? I doubt the NCAA is going to tell us unless it has to.

[ScoobyFanClub]

Re: Too early for the PWR? Princeton and Brown say no!

When three or more teams tie they use RPI to break the tie because each team could win a comparison against another. However, when there is a tie between two teams then only one can win the comparison. In that scenario the tie-breaking procedure hasn't been explained.

Yes, but in 2007, I believe that in the 3-way scenario, that one team won both comparisons against the other, but was ultimately rated below the others because of RPI.... and in the discussion that year, many including USCHO had it wrong, while CHN had it right -- so there must be something to it. You have to conclude, I think, that a two-way tie would be the same way. The committee may never come right out and say so - but then again, it doesn't have to. My guess is, they don't even have a clue as to what we're talking about. They just take what's given to them.

[Priceless]

Re: Too early for the PWR? Princeton and Brown say no!

1	Yale (EC)
2	North Dakota (WC)
3	Boston Coll (HE)
4	Merrimack (HE)
5	Denver U (WC)
6	Union (EC)
7	NE-Omaha (WC)
8	Minn-Duluth (WC)
9	Michigan (CC)
10	Miami (CC)
11	Notre Dame (CC)
12	New Hampshire (HE)
13	RPI (EC)
14	Dartmouth (EC)
15	Wisconsin (WC)

Bridgeport	Manchester	St Louis	Green Bay
Yale		BC		North Dakota	Merrimack
UNO		Union		UMD		Denver
Miami		UNH		Michigan	Notre Dame
AHA Champ	Dartmouth	RPI	        Wisconsin

[mcfarljd]

Re: Too early for the PWR? Princeton and Brown say no!

Yale UNO and AH champ. Ummm, dibs!

1	Yale (EC)
2	North Dakota (WC)
3	Boston Coll (HE)
4	Merrimack (HE)
5	Denver U (WC)
6	Union (EC)
7	NE-Omaha (WC)
8	Minn-Duluth (WC)
9	Michigan (CC)
10	Miami (CC)
11	Notre Dame (CC)
12	New Hampshire (HE)
13	RPI (EC)
14	Dartmouth (EC)
15	Wisconsin (WC)

Bridgeport	Manchester	St Louis	Green Bay
Yale		BC		North Dakota	Merrimack
UNO		Union		UMD		Denver
Miami		UNH		Michigan	Notre Dame
AHA Champ	Dartmouth	RPI	        Wisconsin

[Priceless]

Re: Too early for the PWR? Princeton and Brown say no!

MSU leads 5-2 late and UNO leads 3-1. Assuming those scores hold...

1	Yale (EC)
2	North Dakota (WC)
3	Boston Coll (HE)
4	Merrimack (HE)
5	Denver U (WC)
6	Michigan (CC)
7	Minn-Duluth (WC)
8	Union (EC)
9	NE-Omaha (WC)
10	Notre Dame (CC)
11	New Hampshire (HE)
12	Miami (CC)
13	RPI (EC)
14	Wisconsin (WC)
15	Western Mich (CC)

Bridgeport	Manchester	St Louis	Green Bay
Yale		BC		North Dakota	Merrimack
Union		Michigan	UMD		Denver
UNO		UNH		Notre Dame	Miami
AHA Champ	RPI		W Mich		Wisconsin

[redwings8831]

Re: Too early for the PWR? Princeton and Brown say no!

I would put Denver/Miami in St. Louis and Minnesota-Duluth/Notre Dame in Green Bay. Agree on everything else.

[Priceless]

Re: Too early for the PWR? Princeton and Brown say no!

The TUC line

Bemidji State		0.5035
Ohio State		0.5029
AK-Anchorage 		0.5029
Northern Mich		0.5004
--
Quinnipiac		0.4981
Mich State		0.4946

[J.D.]

Re: Too early for the PWR? Princeton and Brown say no!

I would put Denver/Miami in St. Louis and Minnesota-Duluth/Notre Dame in Green Bay. Agree on everything else.

Agree

[Priceless]

Re: Too early for the PWR? Princeton and Brown say no!

Today's games have no impact on the pairwise.

Some preliminary calculations show Yale and North Dakota have clinched NCAA bids. In fact, they finish no worse than 8th in worst-case scenarios I've run. I have managed to get Denver and BC down to 13th, which means they can be caught. Merrimack I have gotten down to 14th. Union and Michigan can still miss the tournament completely.

Conversely, I have been able to get every current TUC into the tournament as an at-large team. I have not tried to calculate AHA but it seems safe to say RIT and RMU will not both be getting berths.

[UnionHockeyManiac]

Re: Too early for the PWR? Princeton and Brown say no!

Union can still miss the tournament completely.

Who do I need to root for/against for that not to happen?