2022-23 NET/Bracketology Discussion

[dlmzzz]

If Duke is #25 in the net, shouldn't Duke be considered a "good win" for Clemson and also for Wake Forest?

Yep, you are right. At one point in my data gathering Duke was 27th, and at another point it was 25. I updated the Net ranking but missed the Good Wins column.

[dlmzzz]

Thanks. Interesting that we're considered "better" than Clemson according to the ranks, yet our projected ACC record is so much worse (15-5 vs. 12-8).

The stains on your resume from a bad loss (bad performances against bad teams in general) are one of the biggest differences in how the rankings perceive Duke & Clemson. It is also why I think the key thing Duke needs to do to ensure our tourney position is to avoid bad losses.

That's why I say that as long as we win against Wake (H), Notre Dame (H), and Louisville (H), we are going to be OK. The Away game at GA Tech could also be a potential bad loss, if they slip in the rankings.

[House P]

I thought this would be a good time to post what the resumes of the top conference teams (no, I don't know why I am including Syracuse in this list) looked like at this point in the season.

Amending your excellent chart to add Torvik's "Tourney Cast" estimates for the likelihood each team makes the tourney and their projected seed.

Duke with a 93.1% chance of making the tourney seems about right to me. Probably pretty close to 100% if Roach returns soon (and plays well) and probably a bit lower if he doesn't.

Team	Net	KenPom	Quad 1	Quad 2	Good Wins (Top 25)	Bad Losses (Quad 3 & 4)	ACC Projection	NCAA Tourney %	Projected Seed
UVA	15	12	3-3	2-0	Baylor (19 N), Illinois (23 N)	None	15-5	99.9%	4.3
Duke	25	29	3-4	1-1	Xavier (17 N), Ohio State (22 H)	None	12-8	93.1%	6.8
NC State	29	36	1-3	5-1	None	None	12-8	97.3%	6.8
UNC	33	22	1-6	1-0	Ohio State (22 N)	None	13-7	90.2%	7
Miami	39	39	4-2	2-1	Rutgers (16 H), Virginia (15 H)	Georgia Tech (155 A)	13-7	93.2%	7.3
Clemson	49	54	2-1	4-0	None	South Carolina (272 A), Loyola Chicago (283 N)	15-5	67.9%	9.4
VA Tech	55	45	2-1	2-4	None	Boston College (226 A)	9-11	50.2%	9.3
Pitt	58	62	3-1	3-5	Virginia (15 H)	None	14-6	54.8%	9.3
Wake	76	74	1-3	2-1	None	Loyola Marymount (99 N)	11-9	25.7%	10.3
Syracuse	114	86	0-2	3-2	None	Colgate (122 H), Bryant (160 H)	11-9	0.4%	12.3

[CDu]

Amending your excellent chart to add Torvik's "Tourney Cast" estimates for the likelihood each team makes the tourney and their projected seed.

Duke with a 93.1% chance of making the tourney seems about right to me. Probably pretty close to 100% if Roach returns soon (and plays well) and probably a bit lower if he doesn't.

Team	Net	KenPom	Quad 1	Quad 2	Good Wins (Top 25)	Bad Losses (Quad 3 & 4)	ACC Projection	NCAA Tourney %	Projected Seed
UVA	15	12	3-3	2-0	Baylor (19 N), Illinois (23 N)	None	15-5	99.9%	4.3
Duke	25	29	3-4	1-1	Xavier (17 N), Ohio State (22 H)	None	12-8	93.1%	6.8
NC State	29	36	1-3	5-1	None	None	12-8	97.3%	6.8
UNC	33	22	1-6	1-0	Ohio State (22 N)	None	13-7	90.2%	7
Miami	39	39	4-2	2-1	Rutgers (16 H), Virginia (15 H)	Georgia Tech (155 A)	13-7	93.2%	7.3
Clemson	49	54	2-1	4-0	None	South Carolina (272 A), Loyola Chicago (283 N)	15-5	67.9%	9.4
VA Tech	55	45	2-1	2-4	None	Boston College (226 A)	9-11	50.2%	9.3
Pitt	58	62	3-1	3-5	Virginia (15 H)	None	14-6	54.8%	9.3
Wake	76	74	1-3	2-1	None	Loyola Marymount (99 N)	11-9	25.7%	10.3
Syracuse	114	86	0-2	3-2	None	Colgate (122 H), Bryant (160 H)	11-9	0.4%	12.3

Agreed. Both in terms of current likelihood and in terms of the likelihood depending somewhat on Roach’s return.

[jv001]

The stains on your resume from a bad loss (bad performances against bad teams in general) are one of the biggest differences in how the rankings perceive Duke & Clemson. It is also why I think the key thing Duke needs to do to ensure our tourney position is to avoid bad losses.

That's why I say that as long as we win against Wake (H), Notre Dame (H), and Louisville (H), we are going to be OK. The Away game at GA Tech could also be a potential bad loss, if they slip in the rankings.

This is the Duke schedule for the rest of the season:
Home games: Miami, Wake, Cheats, ND, Louisville, VT, NC State.
Away games: VT, GT, Miami, VA, Cuse, Cheats.

Our record in the ACC 4-3 and the projections that have been listed here, project Duke to end the regular season at 12-8. Looking at the remaining games it seems this is the way they think the games will go, Wins= Louisville (H), Wake (H), VT (H), NC State (H), GT (A), VT (A), Cuse (A) and ND. Thats 8 wins.

Losses: Virginia (A), Miami (H) Cheats (H), Miami (A) and Cheats (A). That's 5 losses and leaves us with a 12-8 record.

We could beat Miami at home and the Cheats at home. But the Miami game is 2 days after the Cheats game in Miami. Then the next game is against Virginia away. Tough stretch there.

Looking at the Virginia schedule they could do better than the projections of 15-5. Another year the Cheats get Clemson in Cheatville.

GoDuke!

[bshrader]

This is the Duke schedule for the rest of the season:
Home games: Miami, Wake, Cheats, ND, Louisville, VT, NC State.
Away games: VT, GT, Miami, VA, Cuse, Cheats.

Our record in the ACC 4-3 and the projections that have been listed here, project Duke to end the regular season at 12-8. Looking at the remaining games it seems this is the way they think the games will go, Wins= Louisville (H), Wake (H), VT (H), NC State (H), GT (A), VT (A), Cuse (A) and ND. Thats 8 wins.

Losses: Virginia (A), Miami (H) Cheats (H), Miami (A) and Cheats (A). That's 5 losses and leaves us with a 12-8 record.

We could beat Miami at home and the Cheats at home. But the Miami game is 2 days after the Cheats game in Miami. Then the next game is against Virginia away. Tough stretch there.

Looking at the Virginia schedule they could do better than the projections of 15-5. Another year the Cheats get Clemson in Cheatville.

GoDuke!

Except the issue with these projections is that this team has shown no propensity to win on the road, outside of a near-collapse win in Chestnut Hill.

It is more than conceivable, and based solely on results we've seen thus far, could be called probable, that we lose at VT, at Cuse and even at GT. Let alone the road games at Miami, UVA and Carolina, which are looking like sure losses.

The issue with the metrics is that they don't consider trajectory of a team. We needed a herculean 2nd half to beat Pitt at home. Our trajectory isnt...great IMO.

[jv001]

Except the issue with these projections is that this team has shown no propensity to win on the road, outside of a near-collapse win in Chestnut Hill.

It is more than conceivable, and based solely on results we've seen thus far, could be called probable, that we lose at VT, at Cuse and even at GT. Let alone the road games at Miami, UVA and Carolina, which are looking like sure losses.

The issue with the metrics is that they don't consider trajectory of a team. We needed a herculean 2nd half to beat Pitt at home. Our trajectory isnt...great IMO.

I bet it rains on you every day. No rainbows. You should root for Gonzaga because they usually have easy schedules.

GoDuke!

[bshrader]

I bet it rains on you every day. No rainbows. You should root for Gonzaga because they usually have easy schedules.

GoDuke!

What's a rainbow?? 🤣🤣

[scottdude8]

With the usual caveats about Joe Lunardi taken under consideration, I find it interesting that he has 7 ACC teams in his current field, the third most amongst conferences: https://www.espn.com/espn/feature/st...ld-predictions

If this holds, it's a pretty strong indication that the metrics are significantly less down on the ACC than last year (when only five teams made it), which would give us more wiggle room down the stretch: we'll have more chances for impressive wins and fewer opportunities for really bad losses. Something to keep an eye on.

[Mtn.Devil.91.92.01.10.15]

With the usual caveats about Joe Lunardi taken under consideration, I find it interesting that he has 7 ACC teams in his current field, the third most amongst conferences: https://www.espn.com/espn/feature/st...ld-predictions

If this holds, it's a pretty strong indication that the metrics are significantly less down on the ACC than last year (when only five teams made it), which would give us more wiggle room down the stretch: we'll have more chances for impressive wins and fewer opportunities for really bad losses. Something to keep an eye on.

I've endeavored to not comment on this thread until I had something positive to say, and here we are! I won't credit Lunardi for his analysis, but I absolutely agree that having nine teams in the mix in late January (liberal interpretation of "late") is an empirically good thing for both Duke and the conference. Especially given how the generationally good Louisville is laying turds across their schedule. Of course, I wholly expect the Cardinals to put up a surprising performance on February 20th.

[Kedsy]

Our record in the ACC 4-3 and the projections that have been listed here, project Duke to end the regular season at 12-8. Looking at the remaining games it seems this is the way they think the games will go, Wins= Louisville (H), Wake (H), VT (H), NC State (H), GT (A), VT (A), Cuse (A) and ND. Thats 8 wins.

Losses: Virginia (A), Miami (H) Cheats (H), Miami (A) and Cheats (A). That's 5 losses and leaves us with a 12-8 record.

You're not entirely correct about the way the computers think this will go. They think we're likely to lose @Virginia, @Miami, @Va Tech, and @UNC. They think we're likely to win all of the other individual games, but because the likelihood (per Torvik) of Duke winning those games ranges from 58% (NCSU) to 97% (Louisville), that we'll probably drop one game in which we're favored to win.

Torvik has us as a 5.4 point favorite (70% chance of win) in our home game against Miami and a 3.4 point favorite (63% chance of win) in our home game against UNC.

[Bluedog]

This is the Duke schedule for the rest of the season:
Home games: Miami, Wake, Cheats, ND, Louisville, VT, NC State.
Away games: VT, GT, Miami, VA, Cuse, Cheats.

Our record in the ACC 4-3 and the projections that have been listed here, project Duke to end the regular season at 12-8. Looking at the remaining games it seems this is the way they think the games will go, Wins= Louisville (H), Wake (H), VT (H), NC State (H), GT (A), VT (A), Cuse (A) and ND. Thats 8 wins.

Losses: Virginia (A), Miami (H) Cheats (H), Miami (A) and Cheats (A). That's 5 losses and leaves us with a 12-8 record.

We could beat Miami at home and the Cheats at home. But the Miami game is 2 days after the Cheats game in Miami. Then the next game is against Virginia away. Tough stretch there.

Looking at the Virginia schedule they could do better than the projections of 15-5. Another year the Cheats get Clemson in Cheatville.

GoDuke!

You're not entirely correct about the way the computers think this will go. They think we're likely to lose @Virginia, @Miami, @Va Tech, and @UNC. They think we're likely to win all of the other individual games, but because the likelihood (per Torvik) of Duke winning those games ranges from 58% (NCSU) to 97% (Louisville), that we'll probably drop one game in which we're favored to win.

Torvik has us as a 5.4 point favorite (70% chance of win) in our home game against Miami and a 3.4 point favorite (63% chance of win) in our home game against UNC.

Not to pile on, but said another way, the individual game predctions/likelihood cannot be looked at in a vaccum when projecting "total record." If a team has an 80% chance of beating EVERY opponent for 5 games, there's still only a 32.8% chance (.8^5) they get through that stretch undefeated even though they are HEAVY favorites in each individual game. So, the computer models would likely predict 4-1 for a team even if they're given an 80% chance to win each of those games (and thus the computers/model still projects the team to win each game at the individual level).

[Natty_B]

With the usual caveats about Joe Lunardi taken under consideration, I find it interesting that he has 7 ACC teams in his current field, the third most amongst conferences: https://www.espn.com/espn/feature/st...ld-predictions

If this holds, it's a pretty strong indication that the metrics are significantly less down on the ACC than last year (when only five teams made it), which would give us more wiggle room down the stretch: we'll have more chances for impressive wins and fewer opportunities for really bad losses. Something to keep an eye on.

Would be kinda nuts if both UNC and Duke are too "down" to be seeded in Greensboro. Also if UNC is at Greensboro then I want nothing to do with that site.

[sagegrouse]

With the usual caveats about Joe Lunardi taken under consideration, I find it interesting that he has 7 ACC teams in his current field, the third most amongst conferences: https://www.espn.com/espn/feature/st...ld-predictions

If this holds, it's a pretty strong indication that the metrics are significantly less down on the ACC than last year (when only five teams made it), which would give us more wiggle room down the stretch: we'll have more chances for impressive wins and fewer opportunities for really bad losses. Something to keep an eye on.

Maybe it's just that some have learned that Nov-Dec upsets may not mean much. Moreover, it's actually a big deal that the ACC best the Big Ten.

[CDu]

Maybe it's just that some have learned that Nov-Dec upsets may not mean much. Moreover, it's actually a big deal that the ACC best the Big Ten.

These two sentences feel somewhat contradictory, as the ACC beat the Big Ten... in December.

[tommy]

Maybe it's just that some have learned that Nov-Dec upsets may not mean much. Moreover, it's actually a big deal that the ACC best the Big Ten.

I don't think the fact that the ACC beat the Big 10 in the Challenge means a rat's pattooie to the committee. They look at the credentials of individual teams, their records, strength of schedule, good wins, bad losses, etc. Why should the outcome of the Nebraska-Boston College game or the Iowa-Georgia Tech game, from November no less, matter to Duke's resume? They don't.

[sagegrouse]

These two sentences feel somewhat contradictory, as the ACC beat the Big Ten... in December.

Only somewhat. I would downgrade big upsets -- as, CL by SC and Loy Chi; wake by Loy Marymount; Syracuse by Bryant -- or, at least, limit their overall weight. And their "weight" spreads throughout the conference, not just affecting a single team.

[dlmzzz]

Hi. I wanted to post an updated version of the conference resumes after the weekend's games. I am including the info House P added (thank you!) that shows Torvik's projections of likelihood fo making the tourney, and current projected seed.

Pre-weekend results, Palm projects Wake as one of the last teams in the field, so that kind of resume is where the cut line is. His projection says that 8 ACC teams would make the field, with VA Tech and Syracuse being left out.
Lunardi (I know, I know) projects Wake in the Next Four Out (#7 below the cut line), with 7 ACC teams in the field.

Team	Net	KenPom	Quad 1	Quad 2	Good Wins (Top 25)	Bad Losses (Quad 3 & 4)	ACC Projection	NCAA Tourney %	Projected Seed
UVA	13	11	4-3	3-0	Baylor (14 N), UNC (24 N)	None	15-5	100.0%	3.8
Duke	29	31	3-5	2-0	Xavier (23 N), Ohio State (19 H)	None	12-8	96.0%	6.8
UNC	32	24	1-6	2-0	Ohio State (19 N)	None	13-7	92.8%	7.4
NC State	36	39	1-4	5-1	None	None	12-8	93.7%	7.9
Miami	44	41	4-3	1-0	Rutgers (20 H), Virginia (11 H)	Georgia Tech (155 A)	13-7	86.7%	8.5
VA Tech	55	48	2-5	2-2	None	Boston College (183A)	9-11	29.8%	9.8
Clemson	59	60	2-2	4-0	None	South Carolina (244 A), Loyola Chicago (237 N)	15-5	51.0%	10.2
Pitt	65	69	4-1	1-5	Virginia (11 H)	Florida State (142 H)	13-7	36.5%	10.2
Wake	71	76	1-4	3-1	None	LSU (116 N)	12-8	21.2%	10.5
Syracuse	100	80	0-3	1-2	None	Colgate (129 H), Bryant (187 H)	11-9	0.7%	12

[Mtn.Devil.91.92.01.10.15]

I've been steering mostly clear of this thread, but I can't tell you the joy it gives me to see "Lunardi (I know, I know)" in this conversation.

If Duke can take care of business, we could be anywhere from a 2 seed to an 4 seed. If we struggle, I could see us as low as 6 or 7.

[scottdude8]

I've been steering mostly clear of this thread, but I can't tell you the joy it gives me to see "Lunardi (I know, I know)" in this conversation.

If Duke can take care of business, we could be anywhere from a 2 seed to an 4 seed. If we struggle, I could see us as low as 6 or 7.

Agreed. The landscape is so muddled that we can certainly still put ourselves in a nice position come March. A lot of work to do, but it’s within reach.