The Bubble

[forgetful]

The Bubble is strange this year.  Pitt is considered reasonably safe.  If they lose to Notre Dame today they will be 21-8.  With 1 bad loss.

Best win against Stanford (RPI 41).

There better record and better RPI is built on not playing anyone that is any good in the pre-conference.

Even winning this game, they are 21-7 with no good wins besides Stanford.

[LAZER]

The Bubble is strange this year.  Pitt is considered reasonably safe.  If they lose to Notre Dame today they will be 21-8.  With 1 bad loss.

Best win against Stanford (RPI 41).

There better record and better RPI is built on not playing anyone that is any good in the pre-conference.

Even winning this game, they are 21-7 with no good wins besides Stanford.

Agreed, I also don't understand the love SEC teams are getting.

[ChicosBailBonds]

Well the only way we are on the bubble and in the discussion is if we win 2 out of last 3 and 1 tournament game.  That means we either beat Villanova or St. Johns or both.  Plus another win against a St. Johns/Providence in tournament.  That would mean beating all the Big East bubble teams (sweeping or splitting against them) or getting a win on the road against a top 10 team.  Either way, it helps our resume a lot.  And if that means our RPI is up in the 50s.....its enough to get the committee to take a closer look at us in which case I like our chances over Georgetown. 

And I am not talking about 25 vs 85.  I am talking about Xavier being 46 or 51.  Or any team for that matter having an RPI in the high 40s or low 50s.  Frankly it doesn't matter as its still a good win.  There is no magic cut off that has 25 as a great win and 26 as not.  Or a win against RPI 50 having more merit than 51.  They will look at all of our wins and loses very closely and compare to others.

I am not arguing about if we are in or out at the moment.  Everyone knows we are out and have a lot of work to do but winning 2 of 3 plus 1 tourney game gets us on the bubble and in the conversation and I will take our resume over Georgetown at that point in time unless they win out.  Better conference record by at least 2 games and a season sweep speaks volumes over them having 2 more top 50 wins. 

I'm just telling you how things are categorized based on two chairs of the committee that I've had the pleasure of talking to. If you look at the mock bracket sheets, they are also categorized in this fashion.

As indicated a few weeks ago, the various tools they use do categorize teams into buckets or categories.  So though you may be entirely right that the difference between 50 and 51 is by a hair, the way it is categorized is differently.  Couldn't we say the same thing about the AP polls?  The difference between 25 and 26 is almost none, yet one is considered "ranked" and one is getting votes but doesn't end up with a number next to them.

Categories are drawn for reasons.  If I remember correctly it is broken down by top 25 wins, top 50, 51-100, etc.  It may have changed.

[4th and State]

Doing so would increase their tourney credits. Why not?

I forgot how literal everyone takes things here....

[Eldon]

I forgot how literal everyone takes things here....

Lol.  Oh the irony...

[MarsupialMadness]

Georgetown will finish the Big East season with a sub .500 record. It will be a joke if they make the tournament.

[forgetful]

Another interesting bubble item and strangeness of the RPI.  VCU is currently 22 in the RPI.

They are 21-7.  The biggest difference between them and us as far as resume is that they are 11-0 against the bottom 150.  Whereas we are 7-0.  Swap 4 of our top 50 losses for bottom 150 wins and we would have the same record.

They have 4 more games against the bottom 150 yet still have a higher SOS than us.  Truly absurd.

VCU does have a big win at Virginia, which goes a long way, but not 49 spots in the RPI different.

[JamilJaeJamailJrJuan]

Another interesting bubble item and strangeness of the RPI.  VCU is currently 22 in the RPI.

They are 21-7.  The biggest difference between them and us as far as resume is that they are 11-0 against the bottom 150.  Whereas we are 7-0.  Swap 4 of our top 50 losses for bottom 150 wins and we would have the same record.

They have 4 more games against the bottom 150 yet still have a higher SOS than us.  Truly absurd.

VCU does have a big win at Virginia, which goes a long way, but not 49 spots in the RPI different.

Yep, the RPI is very flawed. Despite what some here will tell you, the committee uses RPI less and less every year. There are much better metrics out there.

[Lennys Tap]

Yep, the RPI is very flawed. Despite what some here will tell you, the committee uses RPI less and less every year. There are much better metrics out there.

100% correct.

[Jay Bee]

Another interesting bubble item and strangeness of the RPI.  VCU is currently 22 in the RPI.

They are 21-7.  The biggest difference between them and us as far as resume is that they are 11-0 against the bottom 150.  Whereas we are 7-0.  Swap 4 of our top 50 losses for bottom 150 wins and we would have the same record.

They have 4 more games against the bottom 150 yet still have a higher SOS than us.  Truly absurd.

VCU does have a big win at Virginia, which goes a long way, but not 49 spots in the RPI different.

Yes - "SOS" doesn't measure difficulty of schedule.

As for the "big win at Virginia", it's important to their RPI in that it's a road win.. but it's the same as winning anywhere, not just Virginia. In fact, excluding that impact (i.e., i.e., winning helping VCU's adjusted winning percentage), the fact that they won at Virginia instead of losing at Virginia actually lowers their RPI.

RPI concerns itself more so with who you play, not who you beat.

RPI bloooows. Just one thing of many to glance at.

[forgetful]

Yes - "SOS" doesn't measure difficulty of schedule.

As for the "big win at Virginia", it's important to their RPI in that it's a road win.. but it's the same as winning anywhere, not just Virginia. In fact, excluding that impact (i.e., i.e., winning helping VCU's adjusted winning percentage), the fact that they won at Virginia instead of losing at Virginia actually lowers their RPI.

RPI concerns itself more so with who you play, not who you beat.

RPI bloooows. Just one thing of many to glance at.

I believe you, but the fact that people use any metric that results in that is absurd.

[Eldon]

I believe you, but the fact that people use any metric that results in that is absurd.

+1.  If that's true, it should be abandoned ASAP.

[g0lden3agle]

Jay bee care to explain the math, or at least post references where I could figure out the math myself?

[TAMU, Knower of Ball]

Plenty of bubble action tonight!

Ole Miss falls to Texas A&M (Whoop!)
UMass gets upset by Dayton
Miami probably pops the bubble of NC State
Vandy gets literally doubled up by Tennessee
Northern Iowa officially ends Indiana State's bubble chances
TCU is slaughtered by West Virginia
Texas Tech goes down to Baylor
Mississippi State gets destroyed by Missouri
Colorado gets tripped up by Utah
Richmond gets EMBARRASSED by 23 at Rhode Island
St. Joe's flies past St. Bonaventure
LSU gets chomped by Florida
Georgia goes down to Arkansas
UWGB beats Detroit
BYU gets past San Diego
Oregon stomps USC
Creighton is shocked by Xavier
Northwestern bows down to Nebraska
Stephen F. Austin stays undefeated in the Southland by beating SE Louisiana
Minnesota can't handle Michigan
California is destroyed by Arizona State
Toledo survives an OT scare against Western Michigan
Boise State bests Wyoming
Rice is beaten by Middle Tennessee
Columbia can't stop Harvard
Kansas blows it against Oklahoma State

Bubble teams go 18-8

Key Wins:
Xavier over Creighton
Oklahoma State over Kansas

Bubbles Popped:
Ole Miss
NC State
Indiana State
Richmond

[mattyv1908]

Plenty of bubble action tonight!

Ole Miss falls to Texas A&M (Whoop!)
UMass gets upset by Dayton
Miami probably pops the bubble of NC State
Vandy gets literally doubled up by Tennessee
Northern Iowa officially ends Indiana State's bubble chances
TCU is slaughtered by West Virginia
Texas Tech goes down to Baylor
Mississippi State gets destroyed by Missouri
Colorado gets tripped up by Utah
Richmond gets EMBARRASSED by 23 at Rhode Island
St. Joe's flies past St. Bonaventure
LSU gets chomped by Florida
Georgia goes down to Arkansas
UWGB beats Detroit
BYU gets past San Diego
Oregon stomps USC
Creighton is shocked by Xavier
Northwestern bows down to Nebraska
Stephen F. Austin stays undefeated in the Southland by beating SE Louisiana
Minnesota can't handle Michigan
California is destroyed by Arizona State
Toledo survives an OT scare against Western Michigan
Boise State bests Wyoming
Rice is beaten by Middle Tennessee
Columbia can't stop Harvard
Kansas blows it against Oklahoma State

Bubble teams go 18-8

Key Wins:
Xavier over Creighton
Oklahoma State over Kansas

Bubbles Popped:
Ole Miss
NC State
Indiana State
Richmond

Xavier hasn't really been on the bubble at all.  They've been squarly in for two months.

[TAMU, Knower of Ball]

Some bubble action tomorrow!

Depaul at St. John's (11 AM, CBSSN)
Maryland at Clemson (12 PM, ESPN3)
George Mason at George Washington (12 PM, NBCSN)
Marquette at Villanova (1 PM, CBS)
Southern Miss at Florida Atlantic (1 PM, Regional)
Ohio State at Indiana (3 PM, CBS)
Louisiana Tech at UAB (3 PM, Regional)
Georgia Tech at Florida State (5 PM, ESPN3)

We will be the only bubble team projected to lose tomorrow. Let's prove the haters wrong.

RING OUT AHOYA!!!

[TAMU, Knower of Ball]

Xavier hasn't really been on the bubble at all.  They've been squarly in for two months.

Lunardi had them as a 10 seed less than a week ago. That's bubble to me

[Jay Bee]

Jay bee care to explain the math, or at least post references where I could figure out the math myself?

Sounds like great fun on a Saturday night... let me dream something up quickly..

OK, consider: Basketball RPI can be thought of as having 3 factors - (1) Adjusted winning percentage; (2) Opp's winning percentage; and, (3) Opp's Opp's winning percentage... these are weighted 25%, 50%, 25%.

(2) and (3) are what people like to call "SOS" or strength of schedule. This is a misnomer.. the example I'm going over actually makes some logical sense in some regards.. but, a goofy SOS example - let's say MU is scheduling for next year. The last game is going to be against Duke.. Buzz calls you and ax's, "we want a good SOS... which will give us a stronger SOS - playing AT Duke or playing at Bradley Center?"..

Any logical purpose knows that your schedule is... harder... or more challenging.. if you play AT Duke rather than host them... but SOS doesn't. The answer for Buzz is, "it doesn't matter... SOS doesn't care if you play at home, away or on Mars.. it's all the same to your 'strength of schedule'"

...but, back to the numbers.

So.. what I said is taking out the fact that VCU won a road game..(doesn't matter who they beat, factor (1) / adjusted winning percentage doesn't care)... VCU's RPI - in this example, their "SOS" actually is lower because they won.. it would be better if they had lost.

Factor (2) and Factor (3) are calculated differently - there are a couple of significant differences. In this case, the one to be aware of is that the Virginia's unadjusted win-loss record included in VCU's factor (2) is excluding their game(s) with VCU. So, win or lose,... no impact.

Factor (3) includes Virginia IF any of VCU's opponents have played them.. in the case of this season, I believe there are six common opponents.. so in the vast pool of VCU's factor 3 opponents of VCU opponents... there are several weightings of Virginia included. Virginia's contribution to VCU's factor 3 includes the VCU/Virginia game... so, because Virginia lost, it actually hurts VCU's factor 3.

So.. you have a bad guy on Factor 3... and no impact on factor 2.. for a net bad guy to VCU's SOS because they beat Virginia instead of losing to Virginia.

[mattyv1908]

Lunardi had them as a 10 seed less than a week ago. That's bubble to me

A ten seed would put them safely in by 10-12 spots.  Besides, if the 3rd team in the NCAAT is squarely on the bubble as you say, the real bubble teams including Marquette are much further out of the picture than anyone here would want to think.

Xavier's in at this point and probably a 7-9 seed after their win against Creighton.

[willie warrior]

A ten seed would put them safely in by 10-12 spots.  Besides, if the 3rd team in the NCAAT is squarely on the bubble as you say, the real bubble teams including Marquette are much further out of the picture than anyone here would want to think.

Xavier's in at this point and probably a 7-9 seed after their win against Creighton.

Definitely agree that Xavier is in. Think it is likely that BEast only gets those three in dance, with n ouside chance of St. John's being the fourth if they finish strong. MU would have to move big mountains to get in now. Is Ox up to that challenge? He will need plenty of help.

[T.V. Diener 34]

Lunardi was just talking about Marquette on ESPN Bracketology

[willie warrior]

Lunardi was just talking about Marquette on ESPN Bracketology

Didn't hear it all, but I thought it was in the context of them playing both St. John's and Providence who he has on the bubble. Did Lunardi give us any chance?

[ChicosBailBonds]

Yep, the RPI is very flawed. Despite what some here will tell you, the committee uses RPI less and less every year. There are much better metrics out there.

Uses it less and less every year?  Can you back this up?   

Go back each year and do a search on SCOOP around March and how many people say it is used less and less, or has no impact at all.  Then the selections come out and each year people are dumbfounded with the comments by the chair, the actual selections, etc. I don't know why people put themselves through this time and time again.  Its like Groundhog Day.

Is it flawed....absolutely.  Is every system flawed...absolutely.  Is simply using the eyeball test flawed....perhaps even more so.  No one is saying the RPI is great, but people that want to pretend it isn't used or plays little to no role are annually upset at the results.

Just look  at what happened with the mock committee only 2 weeks ago.  Ignore it if you wish, or listen to what they say, what the chairs of past committees say. To each their own.  Not the end all be all, but to say it is used less and less is based on what?

[ChicosBailBonds]

100% correct.

Yes, 100% that it is flawed.   0% correct that it is used less and less.

[ChicosBailBonds]

Here's a pretty good article with last year's chair about the RPI.  Similar to what Doug and Dan have told me in the past, who were both committee chairs.  It is used as an organizing tool.

And here's the big one, and Bobinsky mentions it in the article.  The other systems (almost all) use margin of victory in their analysis and the NCAA wants nothing to do with that as an element.  They pulled it out of the BCS long ago and it has no bearing in the RPI.  Now, all of us may bitch and moan that its not in there, but it isn't.  As a result, that props up the RPI over other tools.  Dan said the same thing to me.  Other tools are used, but because the margin of victory is in there in the other tools, it becomes problematic.  They do not want a system that rewards teams for pounding another team into submission to get style points and running up the score.

As Bobinsky says, it isn't the end all be all.  Tons of stuff goes into selecting a team, including how teams can game their RPI, etc.  But to deny its usage and even its importance to them as an organizing vehicle is crazy talk.  Whether we hear it publicly from a chairperson in the news, or more candid conversations with former chairs one on one.


http://blogs.wsj.com/dailyfix/2013/02/20/the-ncaa-tournament-and-the-rpi/