Bracketology 2012

[Shocktoberfest]

Stepping back and taking a big breath. It's awesome that we are debating our seeding rather than wondering if we will sneak in. 3, 4, 5. Bring it on! I like our chances at any of those spots.

[Jamar Howard 4 President]

Winning % vs. top 50:
Georgetown - 67%
Baylor - 58%
Marquette - 60%
Michigan - 60%
Wisconsin - 50%
Louisville - 55%
Indiana - 60%
Florida - 50%
Florida State - 40%
WSU - 40%

Shox1989, % takes out the "they had more chances for quality wins" argument. Is this still just "BCS bias"?

As for UNI, they were way overrated in the RPI early on this year. For a team that went 9-9 in the MVC, the fact that they are still top 100 should be a pleasant surprise. They have no business being anywhere near the top 50.

[SHOXAAC]

No, Who's on secound.

What.

[The Mad Hatter]

Here is the problem with using the "Top 50 Wins" criteria:

Do you believe that RPI is an accurate assessment of the quality of a team?

If the answer is yes, then why not seed teams based directly off of their RPI because it does reflect how they stack up to everyone else?

If the anser is no, then why do you care that you beat a team ranked in the top 50 of a measurement tool that doesn't actually accurately tell you the quality of the teams you faced?

[Dan]

If you're going to count good wins, you also have to look at the losses. Show us what those rankings look like.

[Ixiah]

Davidson is currently predicted to be a top 50 RPI (barely) if they win out.

[Jamar Howard 4 President]

Here is the problem with using the "Top 50 Wins" criteria:

Do you believe that RPI is an accurate assessment of the quality of a team?

If the answer is yes, then why not seed teams based directly off of their RPI because it does reflect how they stack up to everyone else?

If the anser is no, then why do you care that you beat a team ranked in the top 50 of a measurement tool that doesn't actually accurately tell you the quality of the teams you faced?

The point you bring up is my favorite point to discuss. It is also hard to explain briefly. I will get to it when I get a chance, hopefully later today. Without any explanation, all I can say is that it would be a terrible idea to seed directly off RPI, but you can seed very well off RPI when you group opponents (top 25, top 50, 51-100, etc.)

[dTm]

Winning % vs. top 50:
Georgetown - 67%
Baylor - 58%
Marquette - 60%
Michigan - 60%
Wisconsin - 50%
Louisville - 55%
Indiana - 60%
Florida - 50%
Florida State - 40%
WSU - 40%

Shox1989, % takes out the "they had more chances for quality wins" argument. Is this still just "BCS bias"?

As for UNI, they were way overrated in the RPI early on this year. For a team that went 9-9 in the MVC, the fact that they are still top 100 should be a pleasant surprise. They have no business being anywhere near the top 50.

Yes this is still BCS bias because your sample size is wayyyyy too small to say that % is a factor, nice way of skewing your numbers in your favor though. Factor in bad losses with your top 50 wins....you cant have one without the other.

[Jamar Howard 4 President]

If you're going to count good wins, you also have to look at the losses. Show us what those rankings look like.

Dan, a very fair request. I didn't mean for my top 50 thing to be the whole picture. Mostly, I was just responding to those who think we should easily be a 3 seed if the committee knew what they were doing. WSU's resume has some nice strengths, but there are some weaknesses as well.

[im4wsu]

I really dislike that RPI gets talked about so much when it is such a flawed ranking tool compared to the other statistical models out there. Thankfully, it looks like more sophisticated models are getting more play of late, including ESPN introducing their own BPI. Just to show how the polls and RPI compare to some of the major ranking systems, see the following graph:



Just to give some reference on where WSU stacks up on some of the major rankings listed:

Bayesian LMRC: 5
Ken Pomeroy: 8
Massey Consensus (averages 40 rankings): 10
Sagarin: overall - 10 predictor - 9
ESPN's BPI: 9

If the committee decides to use these kinds of rankings heavily, and I have seen many in the media urging committee members to do so, WSU will benefit in seeding. Traditionally, the committee uses a blend of these kind of rankings along with the polls and RPI. Time will tell exactly what the committee values more.

What does the scale on the y-axis represent? 440 what?

[Rlh04d]

Dan, a very fair request. I didn't mean for my top 50 thing to be the whole picture. Mostly, I was just responding to those who think we should easily be a 3 seed if the committee knew what they were doing. WSU's resume has some nice strengths, but there are some weaknesses as well.

That was going to be my question. I don't see how you can look at wins without looking at losses as well.

Just sticking with FSU, since I think it's fair to beat up on them when I've seen nearly every game they played this year ... Their worst loss was to BC, a 234 RPI team, in regulation. That is head and shoulders above WSU's worst loss, at Drake in 3OT with their 130 or so RPI. Even Clemson is a worse loss than Drake. FSU has six losses to teams with an RPI higher than 25 ... WSU has one. Against Top 100 RPI, FSU is 7-7, while WSU is 7-3. Obviously those numbers can be skewed any way to make any argument, really, but I think it's fair to say that FSU has better wins, but much worse losses.

Also, if you're basing so many numbers on RPI, how would WSU with a 10 RPI not look better than FSU at 25?

[The Mad Hatter]

The chart was of predicting wins in NCAA tournament matchups.

[GoShockers89]

Georgetown is a curious team.

They sport an outstanding 8-3 record against the RPI top 50.

And a putrid 0-3 record against RPI 51-100.

[shocktheheart]

Palm's new bracket is out. Hopefully winning 3 this weekend can get the Shox to the 3 seed and games closer to home.

http://www.cbssports.com/collegebasketball/bracketology

[Rlh04d]

Winning % vs. top 50:
Georgetown - 67%
Baylor - 58%
Marquette - 60%
Michigan - 60%
Wisconsin - 50%
Louisville - 55%
Indiana - 60%
Florida - 50%
Florida State - 40%
WSU - 40%

Shox1989, % takes out the "they had more chances for quality wins" argument. Is this still just "BCS bias"?

As for UNI, they were way overrated in the RPI early on this year. For a team that went 9-9 in the MVC, the fact that they are still top 100 should be a pleasant surprise. They have no business being anywhere near the top 50.

Also, not to be overly picky, but if you remove the first week of the season from those numbers, WSU and Baylor are probably the only two teams whose numbers change. And WSU becomes a 67% team against the Top 50 ... which would put them in a tie for the top spot.

I know you can't really pick and choose things like that when you're arguing body of work, but at the same time ... for a team that gets as few chances as WSU to perform against the top level talent, how is it fair for their entire body of work to be decided by how they performed in their first four games, while replacing three starters? There is probably no other team in the country that is going to have their entire season judged so heavily on their first four games of the year. I would think late November through early March might be a slightly more fair judge of "body of work" than November 13-20.

The opening tournaments are interesting. They're fun at the time. But how much do they really say about a team come December, let alone come March?