Ohio State - 4
Kansas - 1
Duke - 2
Pitt - 10

UNC - 6
ND - 8
SD - 3
FL - 7

Cuse - 17
Purdue - 12
UConn - 14
BYU - 5

WI - 15
KY - 9
LOU - 16
TX - 13

WVU - 19
VAN - 26
AZ - 20
KSU - 23

XAV - 22
St. John's - 24
CIN - 34
GTown - 11

WASH - 33
A&M - 28
Temple - 29
UCLA - 43

GMason - 27
UNLV - 25
MICH - 49
Butler - 38

NOVA - 35
IL - 47
TN - 32
ODU - 21

UGA - 46
FSU - 52
Penn St - 39
Mich St - 45

Marq - 63
USC - 70
VCU - 51
Mizzou - 35
Zaga - 54

UAB - 31
Clemson - 55
Richmond - 42
Memphis - 30
Utah State - 18

What's clear is that (for the most part), the committee rewards teams that have played more difficult schedules and have more wins against quality teams on their resumes (i.e. Marquette, Cuse, etc.)

And the teams that they penalize are the ones that haven't played tough schedules and/or don't really have quality wins ( (i.e. Utah State, Memphis, ODU).

I think WIS should work on taking a page from this book.

Mike Bobinski, A.D. at Xavier is on the selection committee this year. In a newpaper interview today, he said, "The RPI is simply a way to organize the teams- I don't pay a lot of attention...it's a good guideline, it is a tool, but we never pick a team and say, "Well, heck, their RPI is this. They either need to be in or they need to get a high seed because of their RPI." We look a lot more deeply than that

Interestingly, every committee member is assigned leagues to watch during the year. He was the primary contact for the Big East, Horizon, Mid-American and Ohio Valley. He was the secondary contact for the Big Ten, Metro Atlantic, and Patriot. They're expected to be the expert on their assigned conferences and be able to answer questions from the other committee members. They also have periodic conference calls throughout the season with representatives from these conferences.

Good post, seems the responsibility is on the non-power conference teams to schedule hard.  Easier to do in HD than in real life, since it seems that the power teams are reluctant to play on the road against the non-power teams, which is not necessarily the case in HD.  The power conference teams are almost automatically going to end up with a good schedule just from conference play (I've learned this with my Iowa State team in Knight).  

One thing I expected to find is hardly any teams just over .500 in the field.  I saw a couple with records just a little better than .500, like Illinois and Tennessee maybe, but those teams had great schedules I think.  Having a hard schedule is one thing, but winning some of those games is important too.  Maybe for WIS, we could put a little more emphasis on wins against Top 50-RPI?  I know that it's already considered, according to the FAQ at least, but maybe amp up the emphasis a little?

It was strange to hear Alabama listed as a bubble team, with an RPI in the 80s, that seems highly unlikely in WIS, but I know real life tends to not match RPI as closely as WIS, in terms of selection at least.

Not totally related... Wouldn't mind seeing an extra five games added to the HD non-conference schedule to more closely match the real life total number of games, but I don't expect to see this.

I think seble has said that he will work on selection and seeding soon.  My big thing would be putting emphasis on top 50 RPI wins and maybe even take into account top 10 or top 15 RPI wins.  We could have quality wins (top 100), high quality wins (top 50), and signature wins (top 10).

I think you guys are missing a major issue here: in real life, you can watch the teams.

I think RPI is a bad metric generally, at least compared to being able to watch teams and see what they are capable of.  But that's not the case in HD.  In HD, RPI is the best metric we have, so we use it.

Posted by isack24 on 3/13/2011 10:46:00 PM (view original):
I think you guys are missing a major issue here: in real life, you can watch the teams.

I think RPI is a bad metric generally, at least compared to being able to watch teams and see what they are capable of.  But that's not the case in HD.  In HD, RPI is the best metric we have, so we use it.

I think most people do appreciate that rpi can be a flawed metric, and that in real life you have the luxury of a human committee, etc. 

But I'm not even saying they should or shouldn't rely on our rpi. I'm simply saying they need two tweak the system so that quality wins matter more.

Yeah, and that makes a lot of sense.  But I fear that will cause more problems than it will solve, at least at the lower levels.  Here, you often have good-but-not-great teams with great RPIs because of scheduling, lots of wins over OK sims, etc.

I agree that the system is a little too reliant on RPI, but I would worry that the other options are a little subjective (e.g. how much weight to place on vs. top-50).

Isn't it a little counter productive to weight "quality wins" on rpi which most of us agree is somewhat flawed in itself?

Posted by nbstowman on 3/14/2011 1:10:00 AM (view original):
Isn't it a little counter productive to weight "quality wins" on rpi which most of us agree is somewhat flawed in itself?

Really no other way to weigh them. Even the committe which doesn't look at RPI per se to decide who is in or out uses RPI to judge a top 50  or top 100 win. Yes it's somewhat easy to manipulate RPI and get a higher RPI then you deserve but if you have 6 wins against the RPI top 50 a few of those are going to be against legit top 50 teams, not just 6 games against teams that manipulated the RPI system. The way RPI works is teams can actually get a solid RPI, say top 40 and only have 1 or 2 wins against top 50 teams, simply by playing a lot of road games and beating teams that will end up with good records because they play in crappy conferences and are the best in a weak bunch.

I think if he wanted to, Seble would pretty easily be able to put a system in place that fairly closely matches the types of things the real-life committee stresses.  Make a formula - 3 points for a road win vs a top 25-rpi team, 2 pts for a home win over a top-25 rpi team, 2 pts for a road win vs a 26-50 rpi team, et cetera. 

countdown to colonel 19

Posted by girt25 on 3/13/2011 7:42:00 PM (view original):
What's clear is that (for the most part), the committee rewards teams that have played more difficult schedules and have more wins against quality teams on their resumes (i.e. Marquette, Cuse, etc.)

And the teams that they penalize are the ones that haven't played tough schedules and/or don't really have quality wins ( (i.e. Utah State, Memphis, ODU).

I think WIS should work on taking a page from this book.

I disagree.  I think WIS (and real life) already reward teams far too much for losing to good teams.  Anybody can lose to Kansas, so what does that losing to them prove?  Nothing other than they're better than you (and a few hundred other teams).

I do agree that quality wins should mean a lot.  But quality losses are counter intuitive.

I think quality losses are ok IF and ONLY IF margin is taken into consideration.

Losing to Kansas at Kansas by 2 is one thing.  Losing at Kansas by 30 is another.