Bracketology Starts

[BearLover]

The best matchup we could get per the pairwise was Michigan Tech. There's nothing in the pairwise that says the 2nd-best team getting the 14th-best team is any more unjust than the 3rd-best team getting the 13th-best team.

[Beeeej]

The best matchup we could get per the pairwise was Michigan Tech. There's nothing in the pairwise that says the 2nd-best team getting the 14th-best team is any more unjust than the 3rd-best team getting the 13th-best team.

Sure there is. It's called finishing in 2nd, which is slightly higher than 3rd, and gets slightly better treatment. The only way you can justify giving us MTU instead of BU is that you think the #3 deserves better treatment than the #2. You're also forgetting that MTU just won their tournament, something we (badly) failed to do ourselves. You think if we were playing MTU, they'd be easier to beat because they'd only have a couple hundred fans in the building?

How many different ways are you guys gonna find to say exactly the same thing?

In order to win a national championship, Cornell has to beat four teams who did what they needed to do in order to be in the tournament. We didn't get nearly as badly screwed this week as we did in 2003, during which regional we went ahead and beat the team with which we felt the NCAA had tried to screw us, and then we went ahead and beat the next team, too. If we have what it takes to win two games this weekend, it's not going to matter a good goddamn whether Jack Parker and Chris Drury themselves drive the caravan down from Agganis and park a hundred thousand rabid Terrier fans on the glass. Let's focus on getting Cornellians in the building and hoping Coach and his players have their shit together.

[adamw]

FWIW - I personally believe swapping the opponents of the 2-3 seeds (Notre Dame/Cornell) is "more just" than swapping 3-4 -- solely because the difference between BU/Princeton is much greater than the difference between Princeton/Tech.  But that is solely an opinion, since their respective final Pairwise ranks are 15-18-21.  I actually asked the Committee Chair this directly, and I basically got a lot of mumbo jumbo back that led me to believe he didn't really understand the nuance of my question. So I'm guessing they "protected" Notre Dame, and then just swapped OSU/Cornell because they had to. And didn't put much further thought into it.

I will also say this - because I had this discussion with certain Cornell people who were annoyed at having to play "at" BU .... The idea that BU is going to pack throngs of people into Worcester is somewhat silly these days. It won't happen. It doesn't happen. There should be plenty of Cornell fans there. If there aren't - shame on Cornell fans.  This isn't 1972. OR even 1992.  BU is not packing tons of people into there.  The only concern I would have is that BU is hot, and they certainly have the inner talent.  So, bad matchup from that standpoint.  But the crowd and the "getting screwed" factor, I wouldn't worry about.

[Trotsky]

Comparing the 1st band teams and projecting chalk except for them advancing:

Code Select Expand

By straight PWR:

1 St. Cloud   27  6  4  2   39
2 Notre Dame  21  7  3  1   32
3 Cornell     15  8  2  1   26
4 Ohio State  18  5  1  2   26

By offset from 1st band team seed:

1 St. Cloud   26  5  3  1    35
2 Notre Dame  19  5  1 -1    24
3 Cornell     12  5 -1 -2    14
4 Ohio State  14  1 -3 -2    10


[adamw]

Has Cornell ever gotten a positive outcome from a subjective selection committee decision? They get the 3rd overall seed and get sent to a Massachusetts regional against a Boston team in the first round and possibly another Boston team in the second round? I would predict Cornell is about 15% to win this regional.

Well, their then-AD, who was then-Committee Chair, Laing Kennedy - did somehow get Cornell in the tournament in 1991 when it probably had blown its chance.  So there's that.  Thus was born the Pairwise.

[Dafatone]

FWIW - I personally believe swapping the opponents of the 2-3 seeds (Notre Dame/Cornell) is "more just" than swapping 3-4 -- solely because the difference between BU/Princeton is much greater than the difference between Princeton/Tech.  But that is solely an opinion, since their respective final Pairwise ranks are 15-18-21.  I actually asked the Committee Chair this directly, and I basically got a lot of mumbo jumbo back that led me to believe he didn't really understand the nuance of my question. So I'm guessing they "protected" Notre Dame, and then just swapped OSU/Cornell because they had to. And didn't put much further thought into it.

I will also say this - because I had this discussion with certain Cornell people who were annoyed at having to play "at" BU .... The idea that BU is going to pack throngs of people into Worcester is somewhat silly these days. It won't happen. It doesn't happen. There should be plenty of Cornell fans there. If there aren't - shame on Cornell fans.  This isn't 1972. OR even 1992.  BU is not packing tons of people into there.  The only concern I would have is that BU is hot, and they certainly have the inner talent.  So, bad matchup from that standpoint.  But the crowd and the "getting screwed" factor, I wouldn't worry about.

I'm not thrilled at playing BU in their home state, but as long as it's not a situation where they're gonna snap up every available ticket (NoDak in Sioux Falls, for instance), it's okay. I don't like that they're hot and talented, but it happens. Personally, I think this is more "fair" than giving us Michigan Tech. You give the #1 team the easiest matchup if possible. Then you give the #2 team the second easiest matchup if possible. Then you look at the #3 and adjust accordingly.

If I ruled the world, I'd ditch the host team concept entirely, and then I'd let teams play conference opponents in the first round. But I'm not jumping up and down to play Princeton right away.

It's the tournament. Everyone we face is gonna be great. Hopefully we'll be greater.

[KenP]

For what it's worth don't forget we still receive the structural advantage of being #1 seed: we are the home team with regard to bench location and line changes.

[Trotsky]

If I ruled the world, I'd ditch the host team concept entirely, and then I'd let teams play conference opponents in the first round.

Seconded, and if that means ditch the regionals and go to campuses for the first two rounds, so be it.

[Jim Hyla]

Has Cornell ever gotten a positive outcome from a subjective selection committee decision? They get the 3rd overall seed and get sent to a Massachusetts regional against a Boston team in the first round and possibly another Boston team in the second round? I would predict Cornell is about 15% to win this regional.

Well, their then-AD, who was then-Committee Chair, Laing Kennedy - did somehow get Cornell in the tournament in 1991 when it probably had blown its chance.  So there's that.  Thus was born the Pairwise.

So I think the exception proves the point. You had to go back 27 years, not a strong case.

We can certainly come up with a lot more recent screwings than that.

[Trotsky]

Comparing the 1st band teams and projecting chalk except for them advancing:

Code Select Expand

By straight PWR:

1 St. Cloud   27  6  4  2   39
2 Notre Dame  21  7  3  1   32
3 Cornell     15  8  2  1   26
4 Ohio State  18  5  1  2   26

By offset from 1st band team seed:

1 St. Cloud   26  5  3  1   35
2 Notre Dame  19  5  1 -1   24
3 Cornell     12  5 -1 -2   14
4 Ohio State  14  1 -3 -2   10


I was asked to explain the tables.

The four teams are the 1st bands, in order by PWR.

In the first table, the columns are the rounds and the value is the PWR of the team to be played.  The team is assumed to be from a chalk bracket.  The final column is the total of the PWR -- the higher the number the easier the path.

In the second table all I did was subtract the row seed from the value in each cell.  This offset gives the relative difficulty of the path and kinda sorta normalizes between the rows.

Things that jump out: Cornell gets hurt in the First Round because we play the hardest 4th-band opponent.  However, if we get past them then the best possible team we can play in the QF (the 2nd band team) is actually the easiest that any of the 1st-bands have to play.  So, the bracket taketh away in the 1R but giveth in the QF.

The second thing that jumps out is that by absolute strength our hardest path is no easier than Ohio State's -- this gives some support to the theory (to which I don't subscribe) that we got screwed compared with tOSU.  Also note that in terms of relative strength, we are again closer to tOSU than to the midpoint between tOSU and Notre Dame.  

However, the third thing: note that in the QF column of the relative chart we retain the same distance from our best possible QF opponent as Notre Dame and even St. Cloud.  This shows tOSU was screwed compared to the rest of the 1st band teams as they ought being the worst of them.

But leaving those tables, I really think anybody who envies Ohio State being sent to Allentown should have their head examined.  That's the Group of Death: the defending champion who just won their conference title and is the best 2nd-band team; the de facto home team in Penn State, and Princeton who we just saw is the real deal.

[jkahn]

The best matchup we could get per the pairwise was Michigan Tech. There's nothing in the pairwise that says the 2nd-best team getting the 14th-best team is any more unjust than the 3rd-best team getting the 13th-best team.

Sure there is. It's called finishing in 2nd, which is slightly higher than 3rd, and gets slightly better treatment. The only way you can justify giving us MTU instead of BU is that you think the #3 deserves better treatment than the #2. You're also forgetting that MTU just won their tournament, something we (badly) failed to do ourselves. You think if we were playing MTU, they'd be easier to beat because they'd only have a couple hundred fans in the building?

How many different ways are you guys gonna find to say exactly the same thing?

Here's one different way.  Notre  Dame and Cornell are actually tied in PWR comparisons, with ND winning on the RPI tiebreaker.  So the committee would be swapping things for two teams who had the same # of favorable comparisons.  Also, the committee didn't seem to have any problem with putting #6 in the same region with #1, so attendance and geography seem factors over "protection of a team."   I had done my own guess at the final brackets, and had everything as it is except I had Cornell in Bridgeport, ND in Allentown and OSU in Worcester. Perhaps also, I'm bitter as I need to be in NYC for family reasons this weekend and could have swung side trips to Bridgeport but don't have the time to do Worcester.

[BearLover]

The best matchup we could get per the pairwise was Michigan Tech. There's nothing in the pairwise that says the 2nd-best team getting the 14th-best team is any more unjust than the 3rd-best team getting the 13th-best team.

Sure there is. It's called finishing in 2nd, which is slightly higher than 3rd, and gets slightly better treatment. The only way you can justify giving us MTU instead of BU is that you think the #3 deserves better treatment than the #2.

No, that is wrong. What you're effectively saying is that the gap between 2 and 3 matters more than the gap between 3 and 4. ND finished higher than Cornell, who finished higher than Ohio State. Due to the no-intraconference-matchups rule, either #2 or #3 is getting a harder matchup, and either #3 or #4 is getting an easier matchup. Here, #4 benefited at the expense of #3, but #3 could just have easily benefited at the expense of #2, and the Pairwise says nothing about which outcome it prefers.

You're also forgetting that MTU just won their tournament, something we (badly) failed to do ourselves.

We outshot and, overall, outplayed Princeton. It wasn't a horrible showing. We lost, but that's going to happen in an individual hockey game even when you're the better team. The entire point of the Pairwise is to not let the result of one game crowd out all the others in evaluating a team.

You think if we were playing MTU, they'd be easier to beat because they'd only have a couple hundred fans in the building?

No...I think they'd be easier to beat because they're clearly and by all objective measures not as good as BU. You're also conflating the pairing issue with the home-ice-for-the four-seed issue. The former is less unjust but more impactful; the latter is more unjust but less impactful.

[Beeeej]

The best matchup we could get per the pairwise was Michigan Tech. There's nothing in the pairwise that says the 2nd-best team getting the 14th-best team is any more unjust than the 3rd-best team getting the 13th-best team.

Sure there is. It's called finishing in 2nd, which is slightly higher than 3rd, and gets slightly better treatment. The only way you can justify giving us MTU instead of BU is that you think the #3 deserves better treatment than the #2.

No, that is wrong. What you're effectively saying is that the gap between 2 and 3 matters more than the gap between 3 and 4. ND finished higher than Cornell, who finished higher than Ohio State. Due to the no-intraconference-matchups rule, either #2 or #3 is getting a harder matchup, and either #3 or #4 is getting an easier matchup. Here, #4 benefited at the expense of #3, but #3 could just have easily benefited at the expense of #2, and the Pairwise says nothing about which outcome it prefers.

You're also forgetting that MTU just won their tournament, something we (badly) failed to do ourselves.

We outshot and, overall, outplayed Princeton. It wasn't a horrible showing. We lost, but that's going to happen in an individual hockey game even when you're the better team. The entire point of the Pairwise is to not let the result of one game crowd out all the others in evaluating a team.

You think if we were playing MTU, they'd be easier to beat because they'd only have a couple hundred fans in the building?

No...I think they'd be easier to beat because they're clearly and by all objective measures not as good as BU. You're also conflating the pairing issue with the home-ice-for-the four-seed issue. The former is less unjust but more impactful; the latter is more unjust but less impactful.

What you're claiming I mean isn't at all what I've actually said, and that leads me to believe you still genuinely don't understand how the committee does this. Yet we've already spent way more energy and time on this than it was worth, and I have no desire to waste more trying to clarify. I hope to see you in Worcester, and I hope to enjoy that interaction considerably more than I've enjoyed this one.

[BearLover]

The best matchup we could get per the pairwise was Michigan Tech. There's nothing in the pairwise that says the 2nd-best team getting the 14th-best team is any more unjust than the 3rd-best team getting the 13th-best team.

Sure there is. It's called finishing in 2nd, which is slightly higher than 3rd, and gets slightly better treatment. The only way you can justify giving us MTU instead of BU is that you think the #3 deserves better treatment than the #2.

No, that is wrong. What you're effectively saying is that the gap between 2 and 3 matters more than the gap between 3 and 4. ND finished higher than Cornell, who finished higher than Ohio State. Due to the no-intraconference-matchups rule, either #2 or #3 is getting a harder matchup, and either #3 or #4 is getting an easier matchup. Here, #4 benefited at the expense of #3, but #3 could just have easily benefited at the expense of #2, and the Pairwise says nothing about which outcome it prefers.

You're also forgetting that MTU just won their tournament, something we (badly) failed to do ourselves.

We outshot and, overall, outplayed Princeton. It wasn't a horrible showing. We lost, but that's going to happen in an individual hockey game even when you're the better team. The entire point of the Pairwise is to not let the result of one game crowd out all the others in evaluating a team.

You think if we were playing MTU, they'd be easier to beat because they'd only have a couple hundred fans in the building?

No...I think they'd be easier to beat because they're clearly and by all objective measures not as good as BU. You're also conflating the pairing issue with the home-ice-for-the four-seed issue. The former is less unjust but more impactful; the latter is more unjust but less impactful.

What you're claiming I mean isn't at all what I've actually said, and that leads me to believe you still genuinely don't understand how the committee does this. Yet we've already spent way more energy and time on this than it was worth, and I have no desire to waste more trying to clarify. I hope to see you in Worcester, and I hope to enjoy that interaction considerably more than I've enjoyed this one.

This quote

It's called finishing in 2nd, which is slightly higher than 3rd, and gets slightly better treatment. The only way you can justify giving us MTU instead of BU is that you think the #3 deserves better treatment than the #2.

is not correct. There is equal justification, according to the PWR, to giving Cornell MTU as there is giving Cornell BU. I do not mean to convey any ill will in saying this. Did the CHN and USCHO writers who predicted Cornell would play Michigan Tech also "not understand how the committee does this"?

Not worth arguing about, not worth complaining about, LGR

[abmarks]

Yet we've already spent way more energy and time on this than it was worth, and I have no desire to waste more trying to clarify. I hope to see you in Worcester, and I hope to enjoy that interaction considerably more than I've enjoyed this one.

Which is to say that there is still a strong likelihood that you won't enjoy that interaction, either.

"my kingdom for a user-mute button"