Bracketology - week of Jan 27

[BRS24]

Women's Bracketology: How a new metric will help (or hurt) bubble teams

Understanding wins above bubble helps explain why Virginia, which has a 37 NET ranking, is on the outside looking in.

www.espn.com

The NCAA added another metric-based criterion for the selection process this season: wins above bubble (WAB), which calculates how many more or fewer wins a team has compared to what an average bubble team would average against the same schedule. The NET is the basis for determining opponent strength and the bubble team is based on a team ranked 45th in the NET. It’s a way for the committee to empirically see if a team is better or worse than a standard bubble team. The men’s selection committee began using WAB last season; how much it will factor into the women’s selections is something we will start to be able to gauge this March. But WAB has already become a dilemma for one team. Virginia is 14-6 with a 37 NET ranking, but the Cavaliers are winless against Quad 1 opponents. Even more telling: Virginia’s WAB has been negative most of the season. That means the Cavaliers would be expected to have fewer wins (-0.97 is the current number) than the average bubble team. That’s the main reason Virginia is on the outside looking in this week.​

Oh goody, another metric to try and understand, debate, and wonder if the seeding committee actually uses it as intended.

Maryland in and Kentucky out of the top 16. Villanova in Last Four Byes, Seton Hall still in the Last Four Out.

12 -Big Ten
11 -SEC
9 - ACC
8 - Big 12
2 - Big East

 

[ConnMotors]

USC is not a tournament team this year. Next season with Juju and company that a yes for sure.

 

[packwrap]

USC is not a tournament team this year. Next season with Juju and company that a yes for sure.

If they get their act together, they can still sneak into the field this year. That assumes Jazzy Davidson keeps improving, they start winning some close games, and they play well in Big10 tourney.

Next year they will be back to a 4-8 ranked team.

 

[fomalhaut]

Despite playing in the Big East, vs. UCLA in the Big Ten and others in the SEC, Connecticut is still 2nd in the country in Wins above Bubble, only 0.3 behind UCLA.

Wins Above Bubble

 

[Phil]

I know I'm being hypercritical but I'm troubled by these three sentences.

But WAB has already become a dilemma for one team. Virginia is 14-6 with a 37 NET ranking, but the Cavaliers are winless against Quad 1 opponents. Even more telling: Virginia’s WAB has been negative most of the season.

The first asserts that the WAB is a dilemma for one team. The next sentence notes that Virginia as a decent net rating but doesn't have a win against Quad 1 opponents. Unless I'm missing something, that has absolutely nothing to do with WAB. The third sentence starts with "even more telling" which suggests that the previous sentence was relevant to WAB. I do have some criticisms of Creme but he's normally easy to follow. I am new to WAB and I have no doubt that a high WAP is correlated to number of wins against Quad 1 teams but I don't see any necessary relationship between the two. It would all make a lot more sense if you switch the first and second sentences. Is that all it is or am I missing something?

 

[Phil]

USC is not a tournament team this year. Next season with Juju and company that a yes for sure.

USC is almost certainly a tournament team, assuming they don't self-destruct. They are currently ranked number 20 in Massey. I don't have the stats at my fingertips, but I think there's only been one team in history ranked that high the didn't make it in and they weren't from one of the top four conferences.

 

[Phil]

Yes, isn't it interesting after weeks of being number one in the NET rankings they come up with a new metric in which UConn's not number one. Nope, I'm not a conspiracy theorist but it is interesting.

 

[Vowelguy]

Yes, isn't it interesting after weeks of being number one in the NET rankings they come up with a new metric in which UConn's not number one. Nope, I'm not a conspiracy theorist but it is interesting.

WAB did not just originate this week, and kinda lame that you would imply otherwise.

Was used for the men last season and added for the women before this season.

 

[NycUcWbbFan]

I know I'm being hypercritical but I'm troubled by these three sentences.



The first asserts that the WAB is a dilemma for one team. The next sentence notes that Virginia as a decent net rating but doesn't have a win against Quad 1 opponents. Unless I'm missing something, that has absolutely nothing to do with WAB. The third sentence starts with "even more telling" which suggests that the previous sentence was relevant to WAB. I do have some criticisms of Creme but he's normally easy to follow. I am new to WAB and I have no doubt that a high WAP is correlated to number of wins against Quad 1 teams but I don't see any necessary relationship between the two. It would all make a lot more sense if you switch the first and second sentences. Is that all it is or am I missing something?

Crème is sensationalizing the reason why NET-37 Virginia (The Selection Committee (12 members) will only be selecting 37 at-large teams) is “outside looking in”.

• The black mark on VA’s resume is that it had a “bad loss” (one of the specifically enumerated other resume factors). The loss is to NET-204 UMBC.

VA’s case illustrates that the NET’s use as probabilities can contradict its construction:

• It is constructed using “resume” records of PPP and Opponent PPP (PPP = Points Per Possession).
• So VA’s cupcake schedule can boost its NET, despite its bad loss. It has a negative WAB because it should have won its bad loss (since WAB is calculated in reference to NET-45).

 

[visitingcock]

VA’s case illustrates that the NET’s use as probabilities can contradict its construction:

• It is constructed using “resume” records of PPP and Opponent PPP (PPP = Points Per Possession).
• So VA’s cupcake schedule can boost its NET, despite its bad loss. It has a negative WAB because it should have won its bad loss (since WAB is calculated in reference to NET-45).

I believe Charlie’s mathematical reasoning is faulty.

Yes, there are 68 slots for the tournament and 31 of those slots will go to AQ conference champs. That leaves 37 at large slots.

However, you ant just look at the top 37 ranked teams because at least 5 (and possibly more) of the at ranked teams will be filled by the P/5 champs who have already qualified as AQ’s. They can’t be both a AQ and an at large! Thus, a ranking of 37 is not the cutoff to determine at large bids. It must be expanded by least five ranking places to 42..

If you add in a Princeton as an AQ who could also qualify as an at large, it expands the ranking to 43.

I’m not sure I have stated by reasoning clearly. Please feel free to correct me if I am wrong.

 

[YKCornelius]

Yes, isn't it interesting after weeks of being number one in the NET rankings they come up with a new metric in which UConn's not number one. Nope, I'm not a conspiracy theorist but it is interesting.

Phil, the NCAA seeding committee decided over the summer to include WAB as part of this year's enhanced selection criteria. Most fans of WCBB don't keep up to speed on these types of decisions, so I understand your questioning of Charlie's "announcement" at the end of January.

Perhaps the fact that he just now factored it into his Bracketology projections for Virginia suggests that maybe he didn't get the memo earlier? Better late than never.....

See the below link.

DI Women’s Basketball Committee announces selection enhancements - NCAA.org

At its summer meeting this week in Philadelphia, the NCAA Division I Women's Basketball Committee approved championship and selection enhancements, reviewed tournament

www.ncaa.org

 

[YKCornelius]

FWIW, Autumn Johnson, who publishes her own form of bracketology on the NCAA.com site, does indeed have Virginia in the field at this time.

Charlie Creme's biggest competition. See the link below.

2026 March Madness women's bracket predictions: Autumn Johnson's first NCAA tournament projections

With March Madness approaching, Autumn Johnson revealed her first NCAA tournament bracket predictions of the season with UConn, UCLA, South Carolina and Texas grabbing the No. 1 seeds.

www.ncaa.com

 

[NycUcWbbFan]

VA’s case illustrates that the NET’s use as probabilities can contradict its construction:

I believe Charlie’s mathematical reasoning is faulty.

Yes, there are 68 slots for the tournament and 31 of those slots will go to AQ conference champs. That leaves 37 at large slots.

However, you ant just look at the top 37 ranked teams because at least 5 (and possibly more) of the at ranked teams will be filled by the P/5 champs who have already qualified as AQ’s. They can’t be both a AQ and an at large! Thus, a ranking of 37 is not the cutoff to determine at large bids. It must be expanded by least five ranking places to 42..

If you add in a Princeton as an AQ who could also qualify as an at large, it expands the ranking to 43.

I’m not sure I have stated by reasoning clearly. Please feel free to correct me if I am wrong.

At bottom, I don’t think we’re saying anything different.

The “37” in NET (Virginia) vs “37” At-Large (non-AQ) is just coincidence. The NCAA/ Torvik pegs the bubble team for WAB purposes at “45”, i.e. there are Auto-bids (AQ) among the Top NET-45sh.

• I imagine the Selection Committee can quickly unanimously agree which Top NET non-AQs to add to the Top NET AQs for inclusion;
• And by this step, which Top NET non AQs go through extra scrutiny (other factors) to determine the rest of the 37 non-AQs for inclusion;
• Seeding, of course, is separate.
• See: How the field of 68 DI women’s teams is picked for March Madness

VA is one such case for extra scrutiny:

• A NET-39 win (Clemson) (Quad 2) is its signature win;
• It also has a NET-41 non-competitive loss at home (60-79) against Syracuse (Quad 2);
• It is 0-4 against Q1; 1-1 against Q2; 2-0 against Q3; 11-1 against Q4;
• Its Q4 loss is to NET-203 (UMBC) at home.

I understand how the NET works and the Committee, with such understanding, will question VA’s NET-37 as being buoyed by its cupcake schedule.

 

[visitingcock]

At bottom, I don’t think we’re saying anything different.

The “37” in NET (Virginia) vs “37” At-Large (non-AQ) is just coincidence. The NCAA/ Torvik pegs the bubble team for WAB purposes at “45”, i.e. there are Auto-bids (AQ) among the Top NET-45sh.

• I imagine the Selection Committee can quickly unanimously agree which Top NET non-AQs to add to the Top NET AQs for inclusion;
• And by this step, which Top NET non AQs go through extra scrutiny (other factors) to determine the rest of the 37 non-AQs for inclusion;
• Seeding, of course, is separate.
• See: How the field of 68 DI women’s teams is picked for March Madness

VA is one such case for extra scrutiny:

• A NET-39 win (Clemson) (Quad 2) is its signature win;
• It also has a NET-41 non-competitive loss at home (60-79) against Syracuse (Quad 2);
• It is 0-4 against Q1; 1-1 against Q2; 2-0 against Q3; 11-1 against Q4;
• Its Q4 loss is to NET-203 (UMBC) at home.

I understand how the NET works and the Committee, with such understanding, will question VA’s NET-37 as being buoyed by its cupcake schedule.

I don’t think WE are saying anything very different . I Just question CHARLIE’S analysis.

 

[ConnMotors]

USC is almost certainly a tournament team, assuming they don't self-destruct. They are currently ranked number 20 in Massey. I don't have the stats at my fingertips, but I think there's only been one team in history ranked that high the didn't make it in and they weren't from one of the top four conferences.

They have nine games left in the regular season. I see USC going 5 and 4. My opinion but they don’t belong unless they win games against the upper tier Big 10 teams.

 

[jake0414]

Despite playing in the Big East, vs. UCLA in the Big Ten and others in the SEC, Connecticut is still 2nd in the country in Wins above Bubble, only 0.3 behind UCLA.

Wins Above Bubble

Torvik’s WAB figures differ from the WAB data posted on the NCAA website:

NCAA Stats as of 27 Jan 2026

 

[fomalhaut]

Torvik’s WAB figures differ from the WAB data posted on the NCAA website:

NCAA Stats as of 27 Jan 2026

Thanks! I hadn't seen this. Still second, though further behind UCLA than on Torvik. Connecticut's Non-conference WAB, though, is easily the best.

 

[NycUcWbbFan]

I don’t think WE are saying anything very different . I Just question CHARLIE’S analysis.

I think Creme does know NET-37 VA is not at the fault-line of the 37 At-Large (non AQ) bids.

He picked VA as the case to illustrate the Selection Committee process — ostensibly to tie into the newsy bit that WAB is new to the dashboard / team sheets this year.

However, choosing WAB and Q1 as his main reasoning is not the strongest case he could have made arguing for VA’s exclusion.

• And I think he knows that.

 

[fomalhaut]

They have nine games left in the regular season. I see USC going 5 and 4. My opinion but they don’t belong unless they win games against the upper tier Big 10 teams

Hardly anyone in the Big Ten has won games against the upper tier, depending on how you define that. USC has wins against Washington and Nebraska in conference and NC State out of conference. They're 2.56 above the bubble in the stats jake0414 just pointed us to.

No one has beaten UCLA or Iowa in conference. No one but those two has beaten Ohio State. Wisconsin has a win against Michigan State,
and Washington against Michigan.

Flipping it around: Maryland's best win so far is against ...USC.
Ohio State is in easily, with wins against Maryland and two good OOC wins against West Virginia and TCU.
Michigan has only beaten Wisconsin and Illinois (plus ND when they were still ranked).
Michigan State has only beaten Washington and Nebraska (same as USC!).

I just don't see "unless they win games against the upper tier" as being a qualifier for the NCAA tournament in a power conference.

 

[RockyMTblue2]

These kind of metrics can be manipulated in scheduling and will be.

 

[Phil]

They have nine games left in the regular season. I see USC going 5 and 4. My opinion but they don’t belong unless they win games against the upper tier Big 10 teams.

I think UCLA wins over Rutgers, Wisconsin, Northwestern, Penn State and Indiana are highly likely although of course anything can happen. That's five. I don't disagree that losses are expected against UCLA and Ohio State.
That leaves Illinois and Iowa. Are those the two you think will be losses for USC that don't fall into the very likely to win and very likely to lose buckets?

I note that Massey gives USC a slight edge over Iowa but I can not only see that going either way I could see that being a large loss to Iowa. That leaves Illinois, which is had some good games notably win over Maryland. We will know the results of that in another hour.

Five wins would the USC 16 wins which may not be enough. They do play one of the toughest schedules in the country so I'm thinking 17 or 18 wins and they are in but that still in dangerous territory.

 

[packwrap]

USC really needs to get to 9-9 in the league, and win a game or 2 in league tourney to lock up a spot.

That would required at minimum 2 good wins still in league play. Iowa game sure interesting since it's in LA.