C'mon man, your entire argument hinges on winning the CT and one extra win (which you're basically counting twice because you're counting the CT and the win, which is the same win).

Cburton lost in the semis, yes, to the number 1 RPI.  It's not like he choked.  

So that's fine, whatever, maybe it's a one spot difference on the s-curve.  But you're arguing that it's fine that they should be separated by at least 8 teams.  There's no way that GC is 9-15 spots better.

Of course, the reason it happened is because GC won the CT, and that is weighted too heavily.

Posted by tianyi7886 on 10/23/2011 8:53:00 PM (view original):

Posted by cburton23 on 10/23/2011 8:39:00 PM (view original):
So should Boise have been in the NC game in CFB the past few years? They won more games? Every year there are teams that win 25 games in NCAAB that get left out of the dance. If the difference is 3-4 wins, ok. But 1 win? And CT wins are soley dependant on how good the conference is so using that as a factor is just silly.

You are making it sound like GC is in all an sim conf, which it's not.

You have a better record against rpi 1-50 with 6-6, but GC has better record against rpi 50-100. On top of that, you lost a game against a team with rpi 100-200. How often have we seen a team get dropped in seeding for March Madness due to "bad losses?"

You win the rpi 1-50 battle, GC win the rpi 50-100 battle, and the rpi 100-200 battle, so this is pretty much a wash. GC has 1 more win and a CT in a #8 conf, pushing him over the edge. 

Do you think there is a general problem with seeding?

I already said there are some problems with seeding, but that doesn't mean this is one of them.

To isack: when did I say that GC is 10-15 spots better than New Mexico is justfieid? All I said was that I can see why GC got a higher seed. And did you even read why cburton started this thread? He claims his team's resume is better than GC, and the only reason that GC got a better seed was because GC was ranked in the T25. This is completely false because T25 ranking is not used in seeding. I proceed to show how GC's resume is slightly better than his (1 more win, going 2 rounds further in the CT, resulting in a CT championship). 

And there's a reason for why I count the +1 win and the CT championship separately; the CT championship and the +1 win are not the same thing. GC went 2 rounds further than New Mexico in the Conf Tournament. The reason it only resulted in 1 win is because GC got a 1st round bye while New Mexico got to face a weak opponent in round 1 to get an additional win. GC went on 2 rounds farther (making it past semis into the championship, and winning the championship) in the CT and should technically be +2 wins when compared to New Mexico.

For all those that claim New Mexico should be rated above GC, you have to explain why GC CT shouldn't matter, why New Mexico's record against rpi 1-50 outweighs his worse record against rpi 50-100 and 100-200, and why GC should not get credit for his additional win (which technically, is 2 additional wins). 

Tianyi: "when did I say that GC is 10-15 spots better than New Mexico is justfieid?"

Tianyi: "There are some problems with seeding but this isn't one of them."

GC is a 4 seed.  Cburton's squad is a 7.  By saying there is no seeding problem, you are saying that GC is at least 9 teams better.

And yes, I read what he wrote.  No, I don't think rankings play a role outside of the factors used to determine them being used for seeding.  But I also think there are major issues with seeding in general.

Although I do find it ironic that you argue cburton faced weak competition but refuse to acknowledge that he lost to a team significantly better than any team GC played.

I find it amusing that the argument now is GC has a better resume because they fared better against teams 50-100 than I did. I would think that the faring better against 1-50 would be of greater importance, but I guess not. Also to take it a step further I fared better if we combine it and go 1-100. But oh well.

Also tianyi I think it's a little unfair to hold against me the fact I didn't have a first round by as 2 of the teams in my division are in the S16 and the other has been top 10 all year.

Also, I did a little look at real life, and I know it's not comparable but I still found it interesting because it helps my side. If we look at similar teams from last year, meaning the # 4 team from the #1 conference and the #1 team from the #8 conference we get UConn and Memphis. Now if I go on to win it all I will be shocked, but UConn was a 3 seed. But they won the CT. So instead let's look at Georgetown. Lost in round 2 of CT only 21 wins. 6 seed. Or we could look at ND. 26 wins. Lost in Semis. 2 seed. Memphis meanwhile, the top team out of the number 8 conference with 26 wins gets a 12 seed.

Posted by cburton23 on 10/23/2011 11:19:00 PM (view original):
I find it amusing that the argument now is GC has a better resume because they fared better against teams 50-100 than I did. I would think that the faring better against 1-50 would be of greater importance, but I guess not. Also to take it a step further I fared better if we combine it and go 1-100. But oh well.

Also tianyi I think it's a little unfair to hold against me the fact I didn't have a first round by as 2 of the teams in my division are in the S16 and the other has been top 10 all year.

You can combine it to 1-100 all you want, but CS specifically told you the system treats 1-50 rpi as one category, 51-100 as another category, and so on. And why do you think losing 2x to 51-100 rpi, and 1x to team outside of rpi 100 isn't relevant? If someone like Georgetown or Syracuse lost to BU, then to Hofstra, then to Morehead State, it absolutely will get dropped to lower seed.  

Posted by isack24 on 10/23/2011 11:04:00 PM (view original):
And yes, I read what he wrote.  No, I don't think rankings play a role outside of the factors used to determine them being used for seeding.  But I also think there are major issues with seeding in general.

Although I do find it ironic that you argue cburton faced weak competition but refuse to acknowledge that he lost to a team significantly better than any team GC played.

When did I argue that cburton faced weak competition? I acknowledged his #1 SOS and better stats against rpi 1-50. He lost to the #1 rpi team in semis, ok, so what? Unless this means New Mexico can run the tables and take the CT in GC's conf, it doesn't devalue GC's CT championship being used in the seeding equation. 

So you are saying that 50-100 should count the same as 1-50 record? Yep I had one bad loss. When did Michigan State loose to SDSu or somebody like that? It happens.

But who had a better resume last year, Lousville or Memphis? Or Georgetown? Or ND?

Posted by cburton23 on 10/24/2011 12:29:00 AM (view original):
So you are saying that 50-100 should count the same as 1-50 record? Yep I had one bad loss. When did Michigan State loose to SDSu or somebody like that? It happens.

But who had a better resume last year, Lousville or Memphis? Or Georgetown? Or ND?

I'm not saying that rpi 50-100 games should count the same as 1-50, but it should count, just like 100-200 games, and 200-300 games. You are ahead in the rpi 1-50 games, you are behind in 50-100, 100-200, wins, Conf Tournament. The question comes down to, does your rpi 1-50 edge outweigh the other 4 factors. Obviously you think so and the seeding system doesn't. With you losing to a team outside of rpi 100 and being 2 rounds behind in CT, I agree with the seeding system. 

And you didn't just have 1 bad loss, you lost to 3 teams outside of rpi 50. Which 4 seed in last years big dance had 2 losses to rpi 50-100, and 1 outside of that?

The bottom line is that we're splitting hairs here ... and imho you're splitting the wrong hairs.

Some of the points are just non-points (like calling losses to teams in the 50-100 rpi range "bad losses", or asking for an exact 4 seed from RL last year who had the exact same set of losses), and some of them are overemphasized points (like having one more win, which basically doesn't matter in the grand scheme of things).

The committee puts a huge emphasis on how strong your schedule was and how you did against NT-caliber opponents. GC fared very poorly against NT-caliber competition. So yes, I think having the clear edge in that area trumps some small differences elsewhere. (And again, any time it's close enough to be a reasonable debate, the fact that one team beat the other head-to-head is a pretty obvious tiebreaker.)

Posted by tianyi7886 on 10/24/2011 12:53:00 AM (view original):

Posted by cburton23 on 10/24/2011 12:29:00 AM (view original):
So you are saying that 50-100 should count the same as 1-50 record? Yep I had one bad loss. When did Michigan State loose to SDSu or somebody like that? It happens.

But who had a better resume last year, Lousville or Memphis? Or Georgetown? Or ND?

I'm not saying that rpi 50-100 games should count the same as 1-50, but it should count, just like 100-200 games, and 200-300 games. You are ahead in the rpi 1-50 games, you are behind in 50-100, 100-200, wins, Conf Tournament. The question comes down to, does your rpi 1-50 edge outweigh the other 4 factors. Obviously you think so and the seeding system doesn't. With you losing to a team outside of rpi 100 and being 2 rounds behind in CT, I agree with the seeding system. 

And you didn't just have 1 bad loss, you lost to 3 teams outside of rpi 50. Which 4 seed in last years big dance had 2 losses to rpi 50-100, and 1 outside of that?

Kentucky last year, a four seed, lost to:

Alabama, Mississippi, Arkansas.


  http://web1.ncaa.org/app_data/weeklyrpi/2010MBBrpi1.html

I cant keep track here - is this a discussion of what SHOULD be the seeding methodology or what IS the HD seeding methodology?

It seems to be both mets.