NCAA tournament seeding - facts and myths

[BRS24]

Some interesting items in this article, after the wins by UConn & Texas, and the lackluster wins by USC and UCLA -

Five things we learned in a week of women's NCAA hoops upsets

In a week of upsets, a perennial contender reestablished itself among the nation's elite and a defending champ looked vulnerable.

www.espn.com

 

[HuskyNan]

Nan, I know you’ve watched these seeding determinations play out over many seasons, as have I. I have never seen a top 5 team get beaten at home in February as badly as SC was beaten yesterday and still retain their seeding advantage. If I’m wrong we’ll know in a few weeks and you can tell me, “I told you so.”

The thing is, you could end up being right for the wrong reason. The Committee isn’t going to use one game to determine seeding but let’s imagine the Committee members wake up in a good mood and decide to bestow a #1 seed on UConn. It happens. i didn’t use the word capricious without reason or history to back it up

 

[Plebe]

Nan, I know you’ve watched these seeding determinations play out over many seasons, as have I. I have never seen a top 5 team get beaten at home in February as badly as SC was beaten yesterday and still retain their seeding advantage. If I’m wrong we’ll know in a few weeks and you can tell me, “I told you so.”

Flip the statement around. In the many seasons I've watched, I've never seen a team with a clearly superior resume get seeded below a team with a clearly inferior resume solely on the basis of one head-to-head result.

 

[oldude]

The thing is, you could end up being right for the wrong reason. The Committee isn’t going to use one game to determine seeding but let’s imagine the Committee members wake up in a good mood and decide to bestow a #1 seed on UConn. It happens. i didn’t use the word capricious without reason or history to back it up

So you’re suggesting that UConn could still be a #1 seed even if they lost to SC yesterday?

 

[Plebe]

The committee members are human beings. Imagine trying to explain how you seeded SC ahead of UConn after the Huskies beat the Gamecocks by 29 on their home court.

And the explanation is an easy one: "We base our seeding decisions on the entire season, not just one game."

 

[oldude]

Flip the statement around. In the many seasons I've watched, I've never seen a team with a clearly superior resume get seeded below a team with a clearly inferior resume solely on the basis of one head-to-head result.

UConn’s resume is hardly clearly inferior as their NET ranking makes clear. Let’s all put a cork in this until the final seedings are completed and made public.

 

[Plebe]

UConn’s resume is hardly clearly inferior as their NET ranking makes clear. Let’s all put a cork in this until the final seedings are completed and made public.

A NET ranking is not a resume. A resume is who you've beaten and who you've lost to.

In a holistic assessment of quality wins and significant losses it is not a close call.

 

[YKCornelius]

Nan, I know you’ve watched these seeding determinations play out over many seasons, as have I. I have never seen a top 5 team get beaten at home in February as badly as SC was beaten yesterday and still retain their seeding advantage. If I’m wrong we’ll know in a few weeks and you can tell me, “I told you so.”

Here's a question I can't answer, but maybe somebody has time to research: In the relatively short window where the selection committee has provided "reveals", have there even been any Top Five teams receiving anything close to such a loss at home and retain (or improve) the seeding they had in the first reveal?

 

[Centerstream]

UConn’s resume is hardly clearly inferior as their NET ranking makes clear. Let’s all put a cork in this until the final seedings are completed and made public.

Sorry @oldude it's the BY, no way that's gonna happen. In fact when the Week 16 AP is announced any minute now, let's see what that thread leads to.

 

[oldude]

A NET ranking is not a resume. A resume is who you've beaten and who you've lost to.

In a holistic assessment of quality wins and significant losses it is not a close call.

You’re bouncing back and force. First it was UConn’s resume is “clearly inferior.” Now we have to use a “holistic assessment” of quality wins. So again I suggest that rather than bouncing back and forth on this, let’s see what happens between now and the end of the season.

 

[Ralph]

You have to compare the entire resume, not one game. So Carolina has more good wins than UConn. Head to head helps but it’s not THE determining factor

View attachment 107077

But interestingly, at the left this shows our NET ranking has just gone from 2 to 1 and SC fell from 1 to 2. Not sure how I can claim this is "our" ranking, I should say it the the team's ranking. I know my rooting for them is vitally important to their winning, but still I am not a person who counts for the team limit of 15 players. I am not even a coach.

 

[JoePgh]

Hypothetically, suppose that South Carolina loses its upcoming regular season game to #8 Kentucky, but then wins everything else including the SEC Tournament. Would two "bad losses" (to UConn because of the MOV and to Kentucky as a lower-NET team) knock it off the #1 seeding line in the tournament?

I would think so, provided that none of the other contenders for a #1 seed stumble along the way. And that is despite the quality wins earlier in the SEC season.

 

[ERWUC65]

So you are suggesting that the Selection Committee criteria of “Head to Head” games has absolutely no bearing on UConn’s seeding vis a vis South Carolina, the defending national champions, despite a 29-point beat down to the Gamecocks on their home court?

Talk about the, “right to be misinformed……”

Here it is, hot off the press - The NCAA Division I Women's Basketball Committee revealed its top 16 seeds Sunday afternoon. Four hours later the list was ready for the shredder. While UConn's stunning win over South Carolina doesn't change the seed for either team. Apology accepted!

 

[maple]

Time to get out the tea leaves & ouija board

I’ve been doing my daily affirmations and manifestations this season.

 

[Dugway]

Seems to me that the season is not yet over. Not to mention the conference tournaments.

Is it not possible that a team that might look to be a #1 today could experience another loss or two (or more)? Might addition losses affect their standing? Is it not possible that a team that might look to be a #2 seed could win out all of the remaining games and their conference tournament, beating a whole bunch of top teams in the process and end up a #1?

I think I will just wait and watch the selection show in a month.

 

[oldude]

Here it is, hot off the press - The NCAA Division I Women's Basketball Committee revealed its top 16 seeds Sunday afternoon. Four hours later the list was ready for the shredder. While UConn's stunning win over South Carolina doesn't change the seed for either team. Apology accepted!

You are cherry-picking a comment from ESPN’s WBB pundits, none of whom are on the Selection Committee. But since you want to go that route, please carefully read Charlie Crème’s analysis under the subheading, “The number 1 seeds are still fluid.” Creme makes many of the same points that I have made. UConn is not a number 1 seed yet, but they’ve “closed the gap.” They will need help, but that help may exist. He points out that both USC & UCLA were losing yesterday deep into the 4th qtr. He also confirms my opinion that LSU is no longer in the running for a #1 seed.

While It is pretty tactless to make statements like, “apology accepted,” I would strongly suggest that you withhold such statements until the final and official tournament bracket is announced.

 

[Crossover]

So you are suggesting that the Selection Committee criteria of “Head to Head” games has absolutely no bearing on UConn’s seeding vis a vis South Carolina, the defending national champions, despite a 29-point beat down to the Gamecocks on their home court?

Talk about the, “right to be misinformed……”

You would flunk high school debate. There is an ocean of space between "absolutely no bearing" and "absolutely determinative".

 

[oldude]

You would flunk high school debate. There is an ocean of space between "absolutely no bearing" and "absolutely determinative".

I love how folks jump in here with a non-sequitur that has little relationship to the original discussion. For the record, my sarcastic use of the term “absolutely no bearing” relates to my being described as misinformed for suggesting that “head to head” competition, a factor confirmed to be part of the NCAA Selection Committee’s criteria, did in fact have a bearing on their selections.

 

[BRS24]

We'll never know what the committee does behind closed doors, however, if a team, such as S Carolina, doesn't right the ship between now and Selection Sunday, they could lose the #1 seed.

 

[Crossover]

I love how folks jump in here with a non-sequitur that has little relationship to the original discussion. For the record, my sarcastic use of the term “absolutely no bearing” relates to my being described as misinformed for suggesting that “head to head” competition, a factor confirmed to be part of the NCAA Selection Committee’s criteria, did in fact have a bearing on their selections.

You didn't just suggest that head to head competition has "a bearing" (which of course it does). You stated emphatically that it trumps everything else. Where do you see that in the Selection Committee's criteria?

 

[oldude]

You didn't just suggest that head to head competition has "a bearing" (which of course it does). You stated emphatically that it trumps everything else. Where do you see that in the Selection Committee's criteria?

Please show me where I supposedly said, “emphatically that it trumps everything else?” I actually sited 4 Selection Committee criteria that all supported UConn’s position ahead of SC. Again, you’re getting creative in your argument.

Honestly, I have no problem going back and forth with you on this. But let’s stick to the facts.

 

[ThisJustIn]

Gosh, I miss Sue Donohoe.
Remember when she opened up the election committee vault?

 

[Gus Mahler]

I saw today where oddsmakers (who? where?) now have UConn as the favorite to win the tournament, replacing SCar. Maybe the committee should take that into account.

 

[fomalhaut]

Calling some of the criteria squishy is putting it mildly.

 

[Glenn]

I note that the AP rankings--which are subject to much scorn here--are a better predictor of the Committee's ranknigs this year, than is the NET. So while the rankings are not part of the criteria--they appear to summarize the criteria better than the NET does.
1. I say AP is better because the AP rankings are more correlated with the selection committee, with a correclation coefficient of 0.89, compared to 0.69 for the NET--both are useful, but the AP rankings more so.
2. Last week's AP ranknigs had two teams, NC State and Kentucky in their top ten and they were not in the NET top 16--both are in the selection committee reveal, close to where AP had them ranked.
3. the NET loves UConnm because we win big. The NET heavily relies on offensive and defensive efficiency, which is a fancy pants way of doing margin of victory adjusted for quality of opponent. That is why we have been consistently high in the NET, now #1 and last week #2. But the Committee seeded us close to the AP ranking and below LSU (#10 in the NET).

	Committee	AP week 15	net
1. UCLA	1	1	5
2. South Carolina	2	4	1
3. Texas	3	3	3
4. Notre Dame	4	2	4
5. Southern California	5	6	6
6. LSU	6	5	10
7. UConn	7	7	2
8. NC State	8	10	19
9. TCU	9	11	9
10. Duke	10	13	8
11. North Carolina	11	12	14
12. Kansas State	12	14	7
13. Kentucky	13	8	21
14. Ohio State	14	9	15
15. Oklahoma	15	16	16
16. Tennessee	16	15	11
Ole miss		28	12
WVA		15	13
correlation		0.88	0.69

 

[Tufkas]

It's all pretty much a swag, considering how critical the health of certain players is. For me, UConn with Azzi Fudd healthy, feeling her best, is a clear #1 seed, if not the overall #1 seed. Her presence has a geometric effect on all aspects of the game, and her team. If she's unable to to play, UConn's a #2 at best, for me, as per the losses this campaign.

It's all just click-bait, until the final in-season version determines the brackets.

 

[Plebe]

You’re bouncing back and force. First it was UConn’s resume is “clearly inferior.” Now we have to use a “holistic assessment” of quality wins. So again I suggest that rather than bouncing back and forth on this, let’s see what happens between now and the end of the season.

No, I'm not. My use of the word "resume" clearly implies a holistic assessment—as opposed to focusing solely on one game, which is what you insist on doing—and to suggest I've contradicted myself only exposes a lack of comprehension.

You keep wanting to shelve the discussion but only if you can get in the last word with yet another spurious take.

 

[Vowelguy]

As I understand it. The NET is the actual starting point. The most perfectly fair placement for every team in the

No the NET is not the starting point. It is just one more piece of information.

 

[Vowelguy]

We'll never know what the committee does behind closed doors,

Though They have held mock sessions with reporters to show them how the process works.

 

[packwrap]

So you’re suggesting that UConn could still be a #1 seed even if they lost to SC yesterday?

There's too many elements favoring teams like SC, UCLA, USC to ND and TX right now. Mostly strength of schedule and quality wins.

Head to head is important, but that also yields crazy inputs. Utah will have to be ahead of ND. Tenn will have to be ahead of UConn. Also margin of victory generally is insignificant. If it's included, it encourages 4th q beatdowns, instead of gentlemanly tradition.

UConn does have 2 huge advantages though now for seeding. 1- Somehow we are #1 in NET already and SC #2. and 2- At current #6 seed, we can watch teams above us still play each other. With plenty of games left, reasonable chance someone above us stumbles. USC and UCLA almost went down yesterday!