Pairwise Analysts, Please Come Help Us 2013

[Ralph Baer]

Re: Pairwise Analysts, Please Come Help Us 2013

You also got blown out by the (former) dead last team in the EZAC.

If RPI gets to be a TUC, it will probably cost Yale some comparisons.

[purpleinnebraska]

Re: Pairwise Analysts, Please Come Help Us 2013

If RPI gets to be a TUC, it will probably cost Yale some comparisons.

And SCSU, as well. Good for the Purple Mavs, though!

[burgie12]

Re: Pairwise Analysts, Please Come Help Us 2013

- Yale has the 5 spot presently, but a 3-2-3 TUC may drop them down a few slots. Anyone have a hunch as to how far?

Yale is in the 5 spot because of their high RPI. As mentioned earlier, they are going to need to keep winning to keep their RPI high and a large percentage of their remaining games (7 of 11) are against teams that are (currently) TUCs. If Yale is going to stay at the 5 spot or higher, their TUC Record is going to improve.

7 of Yale's 28 comparisons are ties broken by RPI. They are losing to UNH and will continue to be losing once TUC comes into play. Niagara and Robert Morris either won't come into play or won't come into play until the very end of the conference tournaments. They are handily winning the TUC comparison against Ferris State. LSSU, SCSU, and NoDak all have records around 0.500, just like Yale, but at this time, if the 10 game requirement was dropped, Yale would be winning all of them.

If Yale starts falling down the ranks, it would be easy to put the blame on their TUC Record, but that would largely be false. If any comparisons are flipped by TUC Record, it will be because Yale performs poorly down the stretch, lowering their RPI, which is a much bigger factor.

[burgie12]

Re: Pairwise Analysts, Please Come Help Us 2013

And SCSU, as well. Good for the Purple Mavs, though!

St. Cloud State is currently 7-7-0 in their TUC Record with none of those games coming against a team with an RPI lower than 0.52. How is adding a team against whom they have a 1-1-0 record going to hurt their TUC?

[goblue78]

Re: Pairwise Analysts, Please Come Help Us 2013

If RPI gets to be a TUC, it will probably cost Yale some comparisons.

Possibly, but the more I look at the PWR, the more interesting it gets. As RPI edges toward TUC status, the Yale loss costs them in TUC, but improves them in RPI (the rating, not the school). Since RPI status is way more important than TUC status, the TUC effect is muted. In addition, while RPI getting in would hurt them, CC getting in would help (though they need to pick things up a bit), and together they'd be a wash. In adddition, HC could drop out which would help. In fact, for Yale, the two upcoming QU games will, I suspect, swamp any TUC effects from teams on the bubble.

[Numbers]

Re: Pairwise Analysts, Please Come Help Us 2013

Thanks for all this discussion, people. It is what I am looking for....

Right now, my thoughts are like this:

In the current PWR, only a few schools are more than one place of from where their RPI rank would place them. The major ones seem to be:

BU - And there has been lots of discussion here about that. TUC record is a hindrance, and OOC record hurts their compares with WCHA teams especially.

UNO - Here, their record against CCHA and ECAC teams hurts the ComOpp criterion a lot. So, they can lose several compares on the bubble by way of TUC (even if only a few % points) and ComOpp even if their RPI would be higher.

WMU - A very strong TUC record helps them. Plus they benefit from BU being weaker in PWR than in RPI.

Also, it doesn't show right now, but SCSU went 0-2 vs CCHA and 0-2 vs HE. That could affect the ComOpp part of the equation for them if they end up on the bubble, and so do, let's say, Fairbanks, WMU and NMU.

Everyone else, it really is a matter of "Just win" At least for awhile, until we have more clarity toward the end of the season.

[Numbers]

Re: Pairwise Analysts, Please Come Help Us 2013

On this week's schedule, every game is always important. However, I think that especially:

For BU - they play Providence. PC is a TUC. This is a crucial weekend. It would be very helpful to BU to get 3 points at least.

The St Cloud - North Dakota series is interesting, especially if someone sweeps.

Notre Dame has been in a bit of a funk. Ferris is a 20s level TUC. They are important games, too.

And, Yale has 2 TUCs on the schedule.

[Patman]

Re: Pairwise Analysts, Please Come Help Us 2013

Re: Lowell

Right now I'm seeing it as RPI uber alles. The RPI is the singularly most important factor and the rest fall in later.

We should also note that Lowell maintains an OK 1-1 record vs WCHA and a 4-0 record vs ECAC.

[Numbers]

Re: Pairwise Analysts, Please Come Help Us 2013

Re: Lowell

Right now I'm seeing it as RPI uber alles. The RPI is the singularly most important factor and the rest fall in later.

We should also note that Lowell maintains an OK 1-1 record vs WCHA and a 4-0 record vs ECAC.

Quickly, Patman, I see 3 comparisons that Lowell would lose if the season played out exactly as it has so far (ie - everyone ends with exactly their current RPI): Notre Dame, Denver, and St Cloud. Denver and St Cloud would depend on the particulars of some COp matchups. Is that what you get also?

[burgie12]

Re: Pairwise Analysts, Please Come Help Us 2013

Re: Lowell

Right now I'm seeing it as RPI uber alles. The RPI is the singularly most important factor and the rest fall in later.

We should also note that Lowell maintains an OK 1-1 record vs WCHA and a 4-0 record vs ECAC.

Let's say that UML sweeps Northeastern, neither UMass nor Providence falls off of the TUC cliff, and some conglomeration of results moves CC up to an RPI of 0.5000.

UML's RPI climbs to 0.5543 (approximately). Their TUC Record climbs to 4-6-0. Hooray!

Let's look at the individual comparisons. They could flip the NoDak comparison in their favor from 1-1 (losing RPI) to 2-1 (winning RPI, losing TUC). The Notre Dame and SCSU comparisons become instant losses and the DU comparison is still up for grabs because UML's loss to DU is not double-counted in their TUC Record. Overall? A net -1 comparison win by sweeping this weekend.

UML has the same issue facing them as BU did last week. If they win, win, and win some more, especially against TUCs to raise their TUC Record to 0.500 or greater, they have a chance to hold their ground. But otherwise? Expect a harsh bump down a couple spots or more. And the worst part is that UML performed so poorly against UNH and BC and those records aren't going anywhere. They need the UMasses and Providences of the league to pick up their performance and stay well above the cliff so that when UML completes their season sweep, they'll still have an RPI above 0.500.

[burgie12]

Re: Pairwise Analysts, Please Come Help Us 2013

What I don't understand, and maybe someone could explain this to me, is why the TUC # changes. Look at LSSU for example. Sometimes they are 7-6-1, 7-5-0, 6-5-1, 7-4-1. Why does that change based on the team your comparing to?

Games against fellow TUCs are not double-counted in their TUC Record. So, they have a TUC Record of 7-6-1 against every TUC that they haven't played, 7-5-0 against TUCs not named Western Michigan, a 6-5-1 TUC Record for the comparisons against Miami, Alaska, and Ferris State (against whom they split), and use a 7-4-1 TUC Record for their comparison against Notre Dame because they were swept in that series.

[Patman]

Quickly, Patman, I see 3 comparisons that Lowell would lose if the season played out exactly as it has so far (ie - everyone ends with exactly their current RPI): Notre Dame, Denver, and St Cloud. Denver and St Cloud would depend on the particulars of some COp matchups. Is that what you get also?

I haven't looked at a single comparison. Of course, I have forgotten that for TUC to matter you need at least 10 games

[Umileated]

Re: Pairwise Analysts, Please Come Help Us 2013

I haven't looked at a single comparison. Of course, I have forgotten that for TUC to matter you need at least 10 games

I should have been clearer - that's what I was getting at earlier this morning when citing the 3-2-3 record. I was speculating that even a 5-2-3 will drop them, as they're ranking in the company of stronger TUC records. Again, I haven't been able to run those comparisons explicitly yet today.

[Numbers]

Re: Pairwise Analysts, Please Come Help Us 2013

Sunday Morning 1-27. A new kid on the block.

Basically, every thing is running according to the RPI right now, with the exception of: Alaska and No Michigan being up, and Nebraska-Omaha being down.

In the case of BU, Providence and Merrimack are now among the TUCs, and that has helped their TUC record. In the case of Lowell, their TUC record is not in play yet.

And, interestingly, Fairbanks is winning several compares by the strength of TUC/COp points, in spite of a lower RPI.

At the top, Miami and Yale are surging, and BC is suffering. Who would have guessed that?

And, No Dakota is sliding, although not hard. And, they usually are power in the second half.

It's a strange year...

[Numbers]

Re: Pairwise Analysts, Please Come Help Us 2013

Seems the big game tonight might actually be the Niagara game. Niagara is sitting a lofty #9. Their TUC record will never come into play, so their ranking in the PWR will almost assuredly follow their RPI. Weaker conference means they need to win, win, win.

Alaska, in spite of bubble-worthy RPI, is actually in a strong position. Their TUC record is stable. And, the schools just beneath them are struggling right now - BU, Lowell, NoDak etc.

At the top, big game for Qpac tonight. Their strong TUC record is great. They simply need to not falter badly in the TUC games left, and they will be fine.