Bubble teams

[UCFKnights]

Blind test, how do these three match up? (not much of a blind test but one can go look for themselves, just comparing the three)

View attachment 51759


Cinci, UCF, Tenn

Not sure how much it matters but team 1's only Q1 wins are over team 2 where team 2's win over team 1 only counts as a Q2.

 

[Plebe]

I'd say the middle team (Team 2) has the worst overall profile - two bad losses and only 1 vaguely notable win (without knowing where in the RPI 1-50 that win came). OOC SOS is high, though, so I could see why the committee might want to reward their gumption.

Team 3 has the fewest Q1 and Q2 wins, and the worst OOC SOS, but also the only team without a Q3 loss. Team 1 is kind of the reverse - best overall set of Q1 and Q2 wins, but also a Q3 loss.

Kind of a tossup between Team 1 and Team 3, though Team 1 has the better OOC SOS and I tend to prefer teams that have higher upside, so I'd go with the team with multiple Q1 wins.

**Asterisk being that I don't know how good those Q1 wins are - big difference between beating Baylor and beating Texas, even though both are "Q1"

Okay, how 'bout if I told you this: Team 1's only Q1 wins are over Team 2.

 

[sivad40]

Western Kentucky, Old Dominion, and James Madison being in seems odd to me reviewing their Massey resumes.

Must look better in the RPI.

RPI: Western Kentucky-29, Old Dominion- 33, and James Madison - 30.

 

[TheFarmFan]

Okay, how 'bout if I told you this: Team 1's only Q1 wins are over Team 2.

Haha, well the first rule of bubble watch is that when your only "good" wins are over other teams on the bubble watch, you're in trouble, so...

(Also, I realized after that the team names were posted in tiny font at the bottom.)

 

[Plebe]

The one element in Tennessee's resume that the committee may or may not be bothered by is the fact that they are only 2-0 vs. Q2 teams. You definitely improve your chances of going undefeated to sub-50 teams if you limit your exposure to teams in the 51-100 range. All other bubble teams have sub-50 losses, but their exposure was also greater in that Q2 range.

But even if that issue concerns the committee, the fact remains that almost all the other bubble teams have lost to teams ranked outside the top 100. One exception is Creighton, whose 7-6 record vs Q2 includes two losses to #98 Butler. Creighton does have two good wins over DePaul and Marquette and multiple wins over other bubble teams. A couple of near-exceptions are Old Dominion (lost twice to #101 UAB) and Bradley (lost to #107 Southern Illinois).

 

[DefenseBB]

All of these teams listed have major, major problems with their resume. Tennessee's, for all its weaknesses, is actually least problematic of the bunch.

You recall what happened to the Roman who delivered the bad news back in 100 AD right?

 

[Blueberries]

Blind test, how do these three match up? (not much of a blind test but one can go look for themselves, just comparing the three)

View attachment 51759


Cinci, UCF, Tenn

without looking at the teams, I'd take the middle, then the left, then the right..... the right one dropped off real fast in my numbers game. the other two were close, but the middle was the best of the lot.

 

[Plebe]

You recall what happened to the Roman who delivered the bad news back in 100 AD right?

I wasn't around at that time

 

[Plebe]

Central Michigan's loss to Toledo in the MAC tourney will now ignite speculation as to their potential as an at-large team.

My opinion is no. They have a high RPI (currently #22, after today's loss), but their best wins are over Dayton and Marist. They've also lost 3 of their last 4 games, including two losses to #148 Toledo.

if they are included in the tournament, it will be throwing them a huge bone just for their RPI.

 

[Plebe]

Updating the "bubble math" report:


The following 31 teams are absolute locks (regardless of any potential loss in conference tournament, for those still in action, marked by *):

• Pac-12: Oregon, Stanford, UCLA, Arizona, Oregon St, Arizona St
• Big Ten: Maryland, Northwestern, Iowa, Indiana, Ohio St, Michigan
• SEC: South Carolina, Mississippi St, Texas A&M, Kentucky, Arkansas
• ACC: Louisville, NC State, Florida St, Duke
• Big 12: Baylor*, TCU*, Iowa State*
• Others: UConn, Gonzaga, DePaul, Missouri St*, Princeton, South Dakota, Drake*

The following 7 teams are highly probable NCAA tournament teams. Call them "bubble adjacent."

• Big Ten: Purdue, Rutgers
• Big 12: Texas
• SEC: LSU
• ACC: Virginia Tech
• Others: Marquette, FGCU*

Barring any "bid steals" in the remaining conference tournaments, these 38 teams will include 12 automatic qualifiers and 26 at-large teams. That would leave 6 remaining at-large bids available.

Edit (3/9): Gonzaga lost in the WCC semifinals, and so now they drop to the at-large pool. That means the above 38 teams include 11 auto bids and 27 at-large teams, with 5 remaining at-large bids for bubble teams.

And now ... THE BUBBLE. And it's a gnarly mess this year. The following is a "long list" of all teams with a remote possibility of consideration as an at-large team.

The following 11 bubble teams have already lost in their conference tournament:

• SEC: Tennessee, Georgia, Alabama
• ACC: Boston College, Georgia Tech
• Big East: Creighton, St. John's, Seton Hall
• AAC: UCF, Cincinnati
• MAC: Central Michigan

The following 13 bubble teams are still in contention for a conference tournament title and automatic bid:

• Big 12: West Virginia, Oklahoma St, Kansas St
• Missouri Valley: Bradley, Northern Iowa, Illinois St
• Conference USA: Old Dominion, Western Kentucky, MTSU, Rice, Charlotte
• Colonial: James Madison, Drexel

Of these 24 teams, I believe the 12 most likely to receive an at-large bid are:

Tennessee
Creighton
James Madison*
Boston College
Old Dominion*
Western Kentucky*
Bradley*
Northern Iowa*
MTSU*
Rice*
West Virginia*
Georgia

Notes:

• The eight teams marked with * still have to play their conference tournaments. Their position on the bubble may improve or diminish accordingly. As many as four of them may earn an automatic bid.
• Barring any further "bid steals," 5 at-large teams will emerge from this list. Each bid steal will reduce this number by one.
• Edit (3/11): Central Michigan appeared to be relatively safe as recently as two week ago, buoyed by an extremely high RPI (in the 16-18 range). But the ended the season losing 3 out their last 4 games, including two losses to sub-.500 Toledo.

 

[Plebe]

Indeed, Charlie Creme thinks CMU is in as an at-large team. I have grave doubts about that.
Again, their *only* case for inclusion is a high RPI (#22 after today's loss). Their resume is markedly worse than UCF's was last year. UCF had a higher RPI (#15) and only one bad loss, and barely made it in as a 12 seed. CMU now has 3 bad losses, and has lost 3 of its last 4.

 

[nwhoopfan]

Indeed, Charlie Creme thinks CMU is in as an at-large team. I have grave doubts about that.
Again, their *only* case for inclusion is a high RPI (#22 after today's loss). Their resume is markedly worse than UCF's was last year. UCF had a higher RPI (#15) and only one bad loss, and barely made it in as a 12 seed. CMU now has 3 bad losses, and has lost 3 of its last 4.

I'm so confused about how RPI works. W/ several bad losses, and no particularly good wins, how is their RPI so high? Overall record of 23-7 is nothing exceptional. I just don't get it.

The MAC is continuing to shoot itself in the foot, Eastern Michigan just beat Ball St.

 

[Lorcan]

Explain to me why Purdue is a lock with a 6-13 record against the Top 100, with a sub 100 loss as well is considered a lock for you?

EDIT: or highly probable.

 

[Plebe]

I'm so confused about how RPI works. W/ several bad losses, and no particularly good wins, how is their RPI so high? Overall record of 23-7 is nothing exceptional. I just don't get it.

The MAC is continuing to shoot itself in the foot, Eastern Michigan just beat Ball St.

How To Game the RPI 101, with Professor Plebe.

First and foremost, schedule OOC games against the best teams from bad conferences -- i.e., teams that aren't as great as their W-L records will make them look.

Opponents' W-L record is by far the weightiest component (50%) of RPI. Maximize that component without suffering too many losses, and you're well on your way to maximizing your RPI.

CMU's OOC schedule includes:

• Green Bay (19-13) (they turned out worse than expected)
• Louisville (28-4) (not from a bad conference, but still an RPI booster)
• Western KY (22-7)
• Dayton (25-8)
• Marist (26-4)
• South Dakota St (22-10) (also worse than expected, but still an RPI helper)
• Bethune-Cookman (21-6)
• UCF (20-10)

If CMU hadn't lost 3 of their last 4 games, their RPI would've been in the mid-teens, exactly where UCF was last year.

 

[Plebe]

I should hasten to add that I'm not bothered by teams -- especially mid-majors -- that find the way to game the RPI. The deck is already stacked high enough against the decent mid-majors, especially when you consider that many of the so-called "power" conference teams refuse to schedule home-and-aways with them, or refuse to schedule them at all.

 

[Plebe]

Explain to me why Purdue is a lock with a 6-13 record against the Top 100, with a sub 100 loss as well is considered a lock for you?

EDIT: or highly probable.

Remember that it's a comparison game. To leave Purdue out, you have to be prepared to show that there are 32 at-large resumes better than theirs.

Raw W-L record per se doesn't tell us a whole lot. Gotta look under the hood to see who all they beat and who all they lost to. In general, losses to good teams (top 50, or at-large caliber teams) don't hurt a resume on their own.

Purdue has 4 top-50 wins. I believe that is more than any of the bubble teams I've listed. Two of those wins are over RPI top-25 teams.

Purdue has only 2 losses outside the top 50, and neither of those is a horrible loss, in the grand scheme of things (#92 Minnesota and #102 Michigan State). Again, almost all of the bubble teams have losses worse than these, with the notable exception of Tennessee and possibly Creighton. (But both of these two teams lack the quality wins that Purdue has.)

Again, comparison game. Which teams should be moved ahead of Purdue, based on body of work?

 

[nwhoopfan]

Thanks @Plebe for the tutorial. USC needs to get better at gaming the RPI, because they are better than most of these bubble teams I firmly believe. Their wins over UCLA and Oregon St. are also a better pair than just about any of the bubble teams. A bad loss in OOC against UCSB, I suppose their losses to Washington and Utah are considered bad as well.

I looked at Purdue's schedule. Something that jumped out at me, other than 1 win against Ohio St. they beat NONE of the good teams from the Big 10. OOC wins over Drake and Arizona St. help (not sure I'm really gonna give them credit for a good win against Western Kentucky).

 

[nwhoopfan]

I seem to recall before Neighbors left UW, he had kind of informally given all the conference coaches helpful hints on maximizing scheduling. I think the conference benefited from that for a season or two, but appears some of those lessons have been forgotten.

 

[Plebe]

Thanks @Plebe for the tutorial. USC needs to get better at gaming the RPI, because they are better than most of these bubble teams I firmly believe. Their wins over UCLA and Oregon St. are also a better pair than just about any of the bubble teams. A bad loss in OOC against UCSB, I suppose their losses to Washington and Utah are considered bad as well.

I looked at Purdue's schedule. Something that jumped out at me, other than 1 win against Ohio St. they beat NONE of the good teams from the Big 10. OOC wins over Drake and Arizona St. help (not sure I'm really gonna give them credit for a good win against Western Kentucky).

Make no mistake -- Purdue is in largely because of their quality OOC wins over Drake and Arizona State. Beating Western Kentucky is a top-50 win as well, albeit over a bubble team. Those wins matter.

I don't think Versyp is an outstanding coach, but I give her credit for scheduling up -- #20 NC SOS -- which is way more than I can say for many of her colleagues. And this year it really saved her bacon, because only a win over Ohio State would've likely been insufficient. Michigan State did better in B10 play against a slightly tougher in-conference schedule, but they had an awful OOC where the beat a bunch of horrible RPI teams and lost to every decent team they played. Those games matter.

The loss to #245 UCSB is indeed awful. So is the fact that 7 of USC's 11 OOC games were against teams outside the top 200.

Having two good wins is nice, but not always enough. Ask Georgia Tech about that -- they have 3 good wins, which altogether are better than USC's wins -- and they don't have any losses as bad as UCSB. Yet they're almost certainly out (although their minuscule chances are higher than USC's).

 

[Plebe]

I seem to recall before Neighbors left UW, he had kind of informally given all the conference coaches helpful hints on maximizing scheduling. I think the conference benefited from that for a season or two, but appears some of those lessons have been forgotten.

Oh yes, the Mike Neighbors-led RPI-boosting campaign openly reported on as a major reason the Pac-12 started getting more teams into the tournament:

2015 Pac-12 Women's Basketball Media Day: Mike Neighbors' RPI advice leads to more NCAA berths for Pac-12

Washington head coach Mike Neighbors is being credited with helping boost the RPI of Pac-12 women's basketball.

pac-12.com

Not all his lessons have been forgotten, not by all of the coaches anyway.