2022-23 NET/Bracketology Discussion

[Nugget]

Following on the St Mary's & UConn discussion, and also (from one of the other threads), the fact that Ohio State is #29 in the NET despite having a record of 11-10: I understand that NET takes into account (from the NCAA's own website, https://www.ncaa.com/news/basketball...ings-explained) "strength of schedule, game location, scoring margin, net offensive and defensive efficiency, and the quality of wins and losses."

If one were to hypothesize the most extreme example - a team that finished the season 0-25, with all 25 losses coming on the road to Q1 teams, losing by 1 point in each game, while maintaining overall high net ratings on offense and defense - could that team finish high up in the NET rankings?

I realize this is an absurd example, but I'm just trying to understand how much weight the system is giving to performance (including against expectations), excluding wins and losses.

Your hypothetical is not likely to occur given that the NET Rankings do have not only an efficiency-type component (which would help the 0-25 team in your example), but also have a results-based component (the "Team Value Index") in which the losses would count and drag down your 0-25 team's rating. But, I don't think there is any way to better "understand how much weight" the system is giving to the relative components, as the NCAA hasn't disclosed the weighting as between the Team Value Index component vs. the adjusted efficiency component.

Having said that, it's always worth remembering as Scott has emphasized that the NET Rankings themselves don't have intrinsic power in setting selection and seedings but, rather, they are mostly an ordering tool for the committee members to use to look at actual wins and losses results, but with obvious focus on Quad 1 wins and Quad 3-4 losses.

Here's a few snips from a recent (late 2022) NCAA release explaining the components of the NET and their (limited) role in the Selection Committee's decisions:

https://www.ncaa.com/news/basketball...ctive%20metric.

"In May 2020, the NCAA announced there will be changes made to the NCAA Evaluation Tool to increase accuracy and simplify it by reducing a five-component metric to just two. The remaining factors include [1] the Team Value Index (TVI), which is a result-based feature that rewards teams for beating quality opponents, particularly away from home, as well as [2] an adjusted net efficiency rating. The adjusted efficiency is a team’s net efficiency, adjusted for strength of opponent and location (home/away/neutral) across all games played…

In addition, the overall and non-conference strength of schedule [ratings provided to the NCAA selection committee] has been modernized to reflect a truer measure for how hard it is to defeat opponents. The strength of schedule is based on rating every game on a team's schedule for how hard it would be for an NCAA tournament-caliber team to win. It considers opponent strength and site of each game, assigning each game a difficulty score. Aggregating these across all games results in an overall expected win percentage versus a team's schedule, which can be ranked to get a better measure of the strength of schedule. …

The number of Quadrant 1 wins and Quadrant 3/4 losses will be incredibly important when it comes time for NCAA tournament selection and seeding."

[dlmzzz]

Hey. Here is an updated version of the conference resumes after the weekend's games.

The week over week changes are fairly minor this week, with Wake slipping and Virginia Tech improving.

I modified one of the columns (Good Wins) to better reflect what the Team Sheets that the Selection Committee uses are expected to include. Top tier Quad 1 Wins are defined as Home wins against a Top 15 NET team, Neutral court wins against a Top 25 team, or Road wins against a Top 40 team. That change, along with slippage from teams that are on the border of being Top 25 teams, eliminates some of the Good Wins. Duke has zero Good Wins now, and Virginia and Miami are down to one. Strangely, the team that looks a little better using this metric is Pittsburgh, who has some road wins against teams barely in the Top 40 (NC State, Northwestern).

For a more complete example of the Team Sheets, here is a link to one for Duke:
https://www.warrennolan.com/basketba...heet?team=Duke

The current cut line for the ACC looks to be Virginia Tech (probably in the First Four in Dayton) probably in the bracket, with Wake likely outside the bracket. Palm's projections are showing that 8 ACC teams would make the field, with Wake and Syracuse being left out.

As usual, by the team you read this some numbers will have changed.

Team	Net	KenPom	Quad 1	Quad 2	Good Wins (H 1-15, N 1-25, A 1-4)	Bad Losses (Quad 3 & 4)	ACC Projection	NCAA Tourney %	Projected Seed
UVA	15	13	4-3	2-0	Baylor (13 N)	None	15-5	100.0%	4.1
Duke	24	28	3-6	2-0	None	None	12-8	95.2%	7.1
UNC	32	31	1-6	5-0	None	None	13-7	91.1%	7.7
NC State	38	52	3-4	3-1	None	None	13-7	89.2%	8.8
Miami	46	39	4-4	1-0	Virginia (15 H)	Georgia Tech (216 A)	13-7	91.0%	8.0
VA Tech	51	50	2-4	3-3	None	Boston College (194 A)	9-11	47.0%	9.5
Clemson	59	59	3-2	4-0	None	South Carolina (273 A), Loyola Chicago (279 N)	15-5	49.7%	10.2
Pitt	62	67	3-2	4-4	Virginia (15 H), NC State (38 A), Northwestern (39 A)	Florida State (201 H)	14-6	38.4%	10.2
Wake	72	74	2-5	2-2	None	LSU (136 N)	11-9	8.9%	10.9
Syracuse	100	86	0-4	1-3	None	Colgate (106 H), Bryant (179 H)	10-10	0.5%	12.3

[CDu]

For what little (nothing) it's worth - don't shoot me, Mtn Devil - the bracket matrix shows all kinds of range for Duke. The average has us as the last 6 seed (average of 6.4). But the range is a bit wild: anywhere from a 5 seed to a 10 seed. Kind of illustrates the jumbled mess that happens after the top 20 or so.

Obviously lots of changes will happen between now and Selection Sunday, and of course none of these guys will be involved in the decision. But it probably gives a reasonable idea of where we are. Most have us in the 6-7 range, with some 5s, a few 8s, and a stray 9 and a stray 10.

[Mtn.Devil.91.92.01.10.15]

For what little (nothing) it's worth - don't shoot me, Mtn Devil - the bracket matrix shows all kinds of range for Duke. The average has us as the last 6 seed (average of 6.4). But the range is a bit wild: anywhere from a 5 seed to a 10 seed. Kind of illustrates the jumbled mess that happens after the top 20 or so.

Obviously lots of changes will happen between now and Selection Sunday, and of course none of these guys will be involved in the decision. But it probably gives a reasonable idea of where we are. Most have us in the 6-7 range, with some 5s, a few 8s, and a stray 9 and a stray 10.

Honest question - when making a bracket projection, are the esteemed prognosticators saying essentially "what would happen if today was Selection Sunday?" or "here's what we thinm things will look like by Selection Sunday?"

It's two very different things in my book. I have zero interest in what anyone says about "if they were building brackets TODAY" - because they aren't. We have to play UVA, and UNC (2x) and have a nice tidy little tournament coming up.

Now, if they are saying something like "Duke is improving and will be getting a player back from injury and we expect them to beat UVA and split with UNC, giving them a 4/5 seed pending an ACC tournament run" - I'm paying closer attention.

Slightly.

[UrinalCake]

Top tier Quad 1 Wins are defined as Home wins against a Top 15 NET team, Neutral court wins against a Top 25 team, or Road wins against a Top 40 team.

Is “top tier quad 1” a separate designation from regular quad 1? I’ve never heard that used before with respect to the selection committee.

[CDu]

Honest question - when making a bracket projection, are the esteemed prognosticators saying essentially "what would happen if today was Selection Sunday?" or "here's what we thinm things will look like by Selection Sunday?"

It's two very different things in my book. I have zero interest in what anyone says about "if they were building brackets TODAY" - because they aren't. We have to play UVA, and UNC (2x) and have a nice tidy little tournament coming up.

Now, if they are saying something like "Duke is improving and will be getting a player back from injury and we expect them to beat UVA and split with UNC, giving them a 4/5 seed pending an ACC tournament run" - I'm paying closer attention.

Slightly.

It is the former. Doing the latter would be way too subjective. Not that there isn’t subjectivity in the former, but there is way more subjectivity in the latter.

[CDu]

Is “top tier quad 1” a separate designation from regular quad 1? I’ve never heard that used before with respect to the selection committee.

There is Q1 and Q1-A. The Q1-A is that top-tier Q1.

[Mtn.Devil.91.92.01.10.15]

It is the former. Doing the latter would be way too subjective. Not that there isn’t subjectivity in the former, but there is way more subjectivity in the latter.

Ugh. That's what I assumed. Thanks.

In that case there's absolutely zero metric to evaluate these brackets, correct?

[scottdude8]

Honest question - when making a bracket projection, are the esteemed prognosticators saying essentially "what would happen if today was Selection Sunday?" or "here's what we thinm things will look like by Selection Sunday?"

It's two very different things in my book. I have zero interest in what anyone says about "if they were building brackets TODAY" - because they aren't. We have to play UVA, and UNC (2x) and have a nice tidy little tournament coming up.

Now, if they are saying something like "Duke is improving and will be getting a player back from injury and we expect them to beat UVA and split with UNC, giving them a 4/5 seed pending an ACC tournament run" - I'm paying closer attention.

Slightly.

I believe most do it as “if the tournament was today”. Some, like Torvik, do more of a projection.

[CDu]

Ugh. That's what I assumed. Thanks.

In that case there's absolutely zero metric to evaluate these brackets, correct?

Correct, none. Nor would there be any reasonable metric to evaluate the other way either. Because teams can do better or worse than expected over remaining games for any number of reasons that are out of the prognosticators’ control.

[Mtn.Devil.91.92.01.10.15]

Correct, none. Nor would there be any reasonable metric to evaluate the other way either. Because teams can do better or worse than expected over remaining games for any number of reasons that are out of the prognosticators’ control.

But it would be more interesting to see who could more accurately predict who will end up on what line at the end of January.

Somewhat.

[CDu]

But it would be more interesting to see who could more accurately predict who will end up on what line at the end of January.

Somewhat.

Not really. Because injuries can happen. Teams can suddenly get better. It would just be an exercise in random guessing.

[Mtn.Devil.91.92.01.10.15]

Not really. Because injuries can happen. Teams can suddenly get better. It would just be an exercise in random guessing.

Like the brackets we all put $20 on in March.

[CDu]

Like the brackets we all put $20 on in March.

Yes, exactly like that. It would be a slightly educated dart throw.

[CDu]

With the win last night, we moved up to #22 in the NET today (up one spot from #23).

[jipops]

With the win last night, we moved up to #22 in the NET today (up one spot from #23).

It’s going to be interesting to see where we stand in the NET after this 3 game gauntlet.

[CDu]

It’s going to be interesting to see where we stand in the NET after this 3 game gauntlet.

If we can take one of the next three games, I will be pleased. That would put us in position to finish the ACC season at 13-7. Anything more than that is gravy.

[DarkstarWahoo]

Taking a more macro look at it, this is shaping up to be a really power-conference-heavy field this year. The Big 12 is going to get eight or nine bids. The B1G will probably get seven.

Looking at the next rung down the ladder:

American: Houston is #1 in the NET and Memphis is #42. I can't imagine they'll get more than two bids, with Memphis probably having to sweat out Selection Sunday if a random team wins the tournament.
A-10: One bid this year. VCU and Saint Louis have shot themselves in the foot too many times.
CAA: Charleston is a nice story, but the rest of the league is really dragging them down. Four teams with sub-300 NET rankings. I think Charleston (currently #61 in NET) needs to win the tournament to get in.
CUSA: Florida Atlantic is probably safe even with a conference tournament loss, provided they keep it going. The win at Florida is really nice for them. Maybe two bids here.
Missouri Valley: They're back in one-bid territory. The best team they have is #74 Bradley.
Mountain West: Really good this year. They could get five teams in (SDSU, Boise, New Mexico, Nevada, Utah State).
Pac-12: UCLA is #5, Arizona #10, Utah #50. Maybe three bids here, with the possibility of a fourth if someone like Stanford wins the tournament. Utah needs to improve its resume or they'll have serious rooting interests in the conference tournament.
IWCC: Zags and St. Mary's are locks. They would get a third if someone else won the tournament.
Randos: It's a good year for Christian basketball with Liberty and Oral Roberts both inside the top 45 in NET. Shouts to Ritchie McKay. But they have a combined zero Q1 wins and two Q2 wins, and their leagues won't provide any opportunities to bolster their resumes, so they need to win their tournaments. (They did play each other this year, for what it's worth, and ORU won.)

[Dr. Maturin]

Charleston is an interesting case. The AP voters sure like them a lot more than the predictive metrics. The CAA hasn't gotten an at-large bid since 2011, and I agree it will be tough if they don't win the AQ. It also shows you how hard it is for smaller conference teams to get a shot. KenPom projects them to go 28-3... that's really hard to do! Per the resume metrics, they are 30th in Strength of Record and 32nd in Wins Above Bubble, which would be at-large territory. But the Committee's reliance on the quadrant system makes it much easier to get in if you're a middling power conference team than if you are in a league like the CAA. Charleston only got one chance at a Q1 win, but a ton of chances to take a damaging loss.

[tommy]

UNC took a nice little hit from the loss to Pitt last night, falling from 37 to 43. Shame, that.