2022-23 NET/Bracketology Discussion

[duke2x]

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.

As long as all 4 local teams don't get to play TN in Greensboro as an underdog, I'm happy. We're pretty far off the Greensboro track right now, and I think the win @UVA is mandatory to put it back in the discussion. I would pick any Big 4 team against Rick Barnes in Greensboro.

If there's good news about the current top 25, Duke is less likely to get a hostile or western venue unless we fall to the 7 line. (Yes, it's possible with a 4/5/6). The spots in Orlando and Albany look like 4/5 seeds this year.

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.

We can be lower than 6 or 7 if we struggle, but I do think a 3 seed is possible with 9 losses going into Selection Sunday.

[Mtn.Devil.91.92.01.10.15]

We can be lower than 6 or 7 if we struggle, but I do think a 3 seed is possible with 9 losses going into Selection Sunday.

We absolutely could be lowee than a 6 or 7. We could lose every remaining game and miss the tournament. 7 seed is just the low end of what I see as likely for this team.

[CDu]

We absolutely could be lowee than a 6 or 7. We could lose every remaining game and miss the tournament. 7 seed is just the low end of what I see as likely for this team.

So you don’t see a realistic scenario where the team drops any lower than they are right now? Because right now they are a 6 or a 7 seed.

Seems like the “if we struggle” should be attached to a lower seed than we currently would be.

[Mtn.Devil.91.92.01.10.15]

So you don’t see a realistic scenario where the team drops any lower than they are right now? Because right now they are a 6 or a 7 seed.

Seems like the “if we struggle” should be attached to a lower seed than we currently would be.

I didn't say it was unrealistic. I said it was the low end of my expectations. So yes, I expect this team to improve their seeding between now and the end of the year, not to do worse and have their seed lowered.

Again, I acknowledge that it is absolutely conceivable or possible that we drop lower, or even miss the tournament. It's within the realm of possibility that we don't win another game this season. My expectation ranges from 6/7 up to 3, possibly 2.

Part of this is because our team is very talented and very young and our coach is clearly learning along side the players. I don't think this team is even close to the ceiling if we gel and start to maximize our potential.

Another part of my expectation is that there's not a big group of elite teams that are likely to run the table and hold their top four seeds. Sure, the top 16 teams might solidify their rankings over the next 6 weeks, but it looks much more volatile than that from my seat.

So, in conclusion, my expectation runs the gamut from "we improve drastically and get a tasty seed on Selection Sunday" to "we remain middling and hold to a 6/7 seed." I don't see us ending up on the bubble.

[CDu]

I didn't say it was unrealistic. I said it was the low end of my expectations. So yes, I expect this team to improve their seeding between now and the end of the year, not to do worse and have their seed lowered.

Again, I acknowledge that it is absolutely conceivable or possible that we drop lower, or even miss the tournament. It's within the realm of possibility that we don't win another game this season. My expectation ranges from 6/7 up to 3, possibly 2.

Part of this is because our team is very talented and very young and our coach is clearly learning along side the players. I don't think this team is even close to the ceiling if we gel and start to maximize our potential.

Another part of my expectation is that there's not a big group of elite teams that are likely to run the table and hold their top four seeds. Sure, the top 16 teams might solidify their rankings over the next 6 weeks, but it looks much more volatile than that from my seat.

So, in conclusion, my expectation runs the gamut from "we improve drastically and get a tasty seed on Selection Sunday" to "we remain middling and hold to a 6/7 seed." I don't see us ending up on the bubble.

Fair enough. In that case, though, I probably wouldn't have used the words "if we struggle," though. I think therein lies the confusion, to which the previous poster and I were commenting. Because "if we struggle" the rest of the way, we will end up on the bubble.

[Mtn.Devil.91.92.01.10.15]

Fair enough. In that case, though, I probably wouldn't have used the words "if we struggle," though. I think therein lies the confusion, to which the previous poster and I were commenting. Because "if we struggle" the rest of the way, we will end up on the bubble.

Shall we continue to litigate my stated opinion? I think I've clarified it pretty thoroughly, but if you want we can move to PMs and I'll continue to reiterate how I feel about the team.

Of course, since it's my opinion there's no statistic to reference to see if it's right or wrong. If Duke loses every game for the rest of the season and misses the tournament, I suppose you could say "gotcha," but again it's reflective of my opinion about this team.

This team could run the table and not lose again this season. We could have injuries that derail our players or two players could shut it down and decide they want to do push ups until the NBA draft - either of those scenarios could mean we don't win again this season.

My opinion is that we end up somewhere in the middle, continue to improve and our NCAA moves up a bit.

I'd prefer not to continue this conversation because it seems incredibly boring for other DBR readers who probably understand what I'm saying and don't want to hear me state it many more times. But if you want to, feel free to PM.

[CDu]

Shall we continue to litigate my stated opinion? I think I've clarified it pretty thoroughly, but if you want we can move to PMs and I'll continue to reiterate how I feel about the team.

Of course, since it's my opinion there's no statistic to reference to see if it's right or wrong. If Duke loses every game for the rest of the season and misses the tournament, I suppose you could say "gotcha," but again it's reflective of my opinion about this team.

This team could run the table and not lose again this season. We could have injuries that derail our players or two players could shut it down and decide they want to do push ups until the NBA draft - either of those scenarios could mean we don't win again this season.

My opinion is that we end up somewhere in the middle, continue to improve and our NCAA moves up a bit.

I'd prefer not to continue this conversation because it seems incredibly boring for other DBR readers who probably understand what I'm saying and don't want to hear me state it many more times. But if you want to, feel free to PM.

Ease up man. I'm not litigating your opinion. I was just clarifying where the confusion resided and making a friendly suggestion. It wasn't a "gotcha" response. You made a post that - on its face - didn't seem sensical. Hence the two replies that you got (one from me, one from someone else). You clarified what you were saying, and thank you for that.

I don't begrudge you your opinion. I personally am pretty agnostic about what our team could do the rest of the way. Mainly because I have no idea how much the freshmen will improve. So I have no strong opinion about the accuracy of your opinion.

I just think the wording of your opinion was a little off, which led to confusion. That's all.

[Mtn.Devil.91.92.01.10.15]

I just think the wording of your opinion was a little off, which led to confusion. That's all.

You got me. I'm guilty. Game over.

I hope the three subsequent replies and clarifications have ended this potential confusion. If anyone wants to better understand how I feel, please PM me.

Go Duke.

[scottdude8]

Relevant to the above discussion is that, if we're thinking about competing for a 6-7 seed (or heck, even a 4-5 seed) is that the teams we're competing against won't be nearly as good as those that we're used to competing against for 1-2 seeds. Last year, 6 and 7 seeds from major conferences roughly averaged around 11 losses. It's not like when we're hunting for a 1 seed where every loss feels debilitating because our competition is rarely going to lose... 6-7 seeds are going to have a lot of 1-1 weeks, just like we have thus far this season.

Again with the obvious caveats about Lunardi, he did say something surprisingly cogent at some point last year when talking about the bubble, which was that bubble teams are expected to lose big time games... that's why there may not be much movement on the bubble if, for instance, a team on the right side of the bubble loses a Q1 game while a team on the wrong side beats up on a team at the bottom of the conference. We're nowhere near bubble talk, obviously, but just an example of how we may have to adjust our perspective if we do find ourselves more in the 4-8 range than the 1-4 range.

[scottdude8]

Duke is amidst arguably the biggest stretch of the season, and after faltering at Clemson getting our captain back in a win over Miami was huge. While not a Q1 game with Miami at #46 in the NET, that win has the potential to eke into that top quadrant if the Hurricanes keep playing well.

Right now, Duke stands at #29 in the NET, which is obviously much lower than we're used to being. However, our 3 Q1 wins are a nice differentiator in comparison to other teams around us. An interesting side note that the committee may have to consider this year is how the Big 12 is skewing the quadrant system: because the NET is so high on that conference, six of the ten teams with 5 or more Q1 wins are all from the Big 12 (the others are Purdue, Alabama, Arizona, and Xavier, the best teams from four other Power conferences). If this pattern continues it'll be quite interesting to see how the committee seeds the Big 12.

Each of Duke's 3 best non-conference wins remain Q1. Despite faltering, the NET still loves Ohio State, with the 8 loss Buckeyes at #24. Iowa has rebounded from a really bad period and is now at #38, fairly safely in the Top 50. Xavier, meanwhile, is lower in the NET (#25) than in the human rankings, but beating them on a neutral floor remains one of the better non-conference wins in the country.

Meanwhile, the ACC is still going to provide plenty of resume opportunities for us. Virginia is at #13, giving us a marquee opportunity when we face them on the road. UNC is #32, right on the edge of giving us a home Q1 opportunity... NC State is also nearing that border, as they stand at #36 (although that'll depend heavily on their health moving forward). Miami (#46), Virginia Tech (#55), Clemson (#59), Pitt (#65), and Wake (#71) are all inside the important Top 75. And don't look now, but Syracuse is within shouting distance of that as well, at #101 (and there tends to be a lot more volatility the further down you get in the rankings).

[bshrader]

Duke is amidst arguably the biggest stretch of the season, and after faltering at Clemson getting our captain back in a win over Miami was huge. While not a Q1 game with Miami at #46 in the NET, that win has the potential to eke into that top quadrant if the Hurricanes keep playing well.

Right now, Duke stands at #29 in the NET, which is obviously much lower than we're used to being. However, our 3 Q1 wins are a nice differentiator in comparison to other teams around us. An interesting side note that the committee may have to consider this year is how the Big 12 is skewing the quadrant system: because the NET is so high on that conference, six of the ten teams with 5 or more Q1 wins are all from the Big 12 (the others are Purdue, Alabama, Arizona, and Xavier, the best teams from four other Power conferences). If this pattern continues it'll be quite interesting to see how the committee seeds the Big 12.

Each of Duke's 3 best non-conference wins remain Q1. Despite faltering, the NET still loves Ohio State, with the 8 loss Buckeyes at #24. Iowa has rebounded from a really bad period and is now at #38, fairly safely in the Top 50. Xavier, meanwhile, is lower in the NET (#25) than in the human rankings, but beating them on a neutral floor remains one of the better non-conference wins in the country.

Meanwhile, the ACC is still going to provide plenty of resume opportunities for us. Virginia is at #13, giving us a marquee opportunity when we face them on the road. UNC is #32, right on the edge of giving us a home Q1 opportunity... NC State is also nearing that border, as they stand at #36 (although that'll depend heavily on their health moving forward). Miami (#46), Virginia Tech (#55), Clemson (#59), Pitt (#65), and Wake (#71) are all inside the important Top 75. And don't look now, but Syracuse is within shouting distance of that as well, at #101 (and there tends to be a lot more volatility the further down you get in the rankings).

Admittedly, I am no analytics expert, but is there any explanation in the discrepancy for St. Mary's between the AP and the NET rankings? How in the world is a team unranked in the AP poll, yet 6th in the NET rankings, with 1 Q1 win and 2 Q3 losses?? Is there a plausible explanation?

[CDu]

Admittedly, I am no analytics expert, but is there any explanation in the discrepancy for St. Mary's between the AP and the NET rankings? How in the world is a team unranked in the AP poll, yet 6th in the NET rankings, with 1 Q1 win and 2 Q3 losses?? Is there a plausible explanation?

AP voters tend to consider W/L record rather than metrics. And there is some inertia involved. If a team isn't well thought of going into the season, it takes a lot of work for them to get noticed by the voters.

St. Mary's does really well metrically (i.e., their scoring differential is very good relative to what an average team would be expected to do). But they don't have the huge resume wins, and they have enough losses for a "low-major" conference team that the AP voters don't like them yet. They beat San Diego State on a neutral floor and lost close against Houston (neutral site) and New Mexico State (at home). So they haven’t gotten the inertia to move in their favor.

Also, they are the second-highest vote-getter in the "also receiving votes. So it isn't like they are totally off the radar. They are essentially #27 in the AP.

Essentially, they have a similar problem to what Duke had in 2021 (only they are in a better position overall): they are just 1-4 in games decided by 5 or less. If they were 20-2 instead of 18-4 (so if they hadn't lost in OT to Washington and hadn't lost to Colorado State by 2), they'd probably be somewhere near the top-10. The metrics ignore the W/L result and only care about the points per possession, so those losses (which really hurt them in the voters' minds) don't make a marginal difference than getting a 1pt win would make metrically.

[jv001]

Admittedly, I am no analytics expert, but is there any explanation in the discrepancy for St. Mary's between the AP and the NET rankings? How in the world is a team unranked in the AP poll, yet 6th in the NET rankings, with 1 Q1 win and 2 Q3 losses?? Is there a plausible explanation?

That is a very good question and I'd love to hear the answer. Like you I'm not an analytics expert, but it doesn't make sense.

GoDuke!

Well, CDu answered the question. Are AP voters certain writers?

[scottdude8]

Admittedly, I am no analytics expert, but is there any explanation in the discrepancy for St. Mary's between the AP and the NET rankings? How in the world is a team unranked in the AP poll, yet 6th in the NET rankings, with 1 Q1 win and 2 Q3 losses?? Is there a plausible explanation?

AP voters tend to consider W/L record rather than metrics. And there is some inertia involved. If a team isn't well thought of going into the season, it takes a lot of work for them to get noticed by the voters.

St. Mary's does really well metrically (i.e., their scoring differential is very good relative to what an average team would be expected to do). But they don't have the huge resume wins, and they have enough losses for a "low-major" conference team that the AP voters don't like them yet. They beat San Diego State on a neutral floor and lost close against Houston (neutral site) and New Mexico State (at home). So they haven’t gotten the inertia to move in their favor.

Also, they are the second-highest vote-getter in the "also receiving votes. So it isn't like they are totally off the radar. They are essentially #27 in the AP.

Essentially, they have a similar problem to what Duke had in 2021 (only they are in a better position overall): they are just 1-4 in games decided by 5 or less. If they were 20-2 instead of 18-4 (so if they hadn't lost in OT to Washington and hadn't lost to Colorado State by 2), they'd probably be somewhere near the top-10. The metrics ignore the W/L result and only care about the points per possession, so those losses (which really hurt them in the voters' minds) don't make a marginal difference than getting a 1pt win would make metrically.

That is a very good question and I'd love to hear the answer. Like you I'm not an analytics expert, but it doesn't make sense.

GoDuke!

Well, CDu answered the question. Are AP voters certain writers?

Yes, CDu's answer is right on. While the AP and NET rankings are certainly correlated, there's nothing concrete connecting them: the AP rankings are based off of writer's perspectives of the teams, while the NET is purely numerical. There are other discrepancies: look at College of Charleston, for example.

There have always been some outliers in the NET rankings, which is why they are just one of many tools used to seed teams. As just one example, Houston was No. 2 in the final NET rankings last year, but were the 18th overall seed in the tournament. Severe outliers like St. Mary's have tended to even out over the course of the season, so it wouldn't surprise me if they start slowly falling down the rankings... but, even if they don't, just because they're in the NET Top 10 doesn't mean they're legitimate contenders for a top seed.

[Bluedog]

Important to note that while NET "raw rank" likes St. Mary's, the NCAA committee from a SEEDING perspective looks at Wins/Losses (like AP voters), so St. Mary's would be equally penalized with a lower seed that the computer metrics suggest given their Q1 and Q2 records.

^what scottdude said

[House P]

Admittedly, I am no analytics expert, but is there any explanation in the discrepancy for St. Mary's between the AP and the NET rankings? How in the world is a team unranked in the AP poll, yet 6th in the NET rankings, with 1 Q1 win and 2 Q3 losses?? Is there a plausible explanation?

Others have already hit upon the fact that the Dork polls like what St. Mary's has done and the NET is highly correlated with the Dork polls.

As to why the Dork polls like St. Mary's, CDu pointed out that their four losses have come by a total of 15 points and two of these losses are currently Q1 games (#1 Houston and New Mexico who is #26 in the NET, #23 in Torvik and #44 in KenPom).

The other side of the coin is that the rest of their schedule is fairly devoid of true cupcakes (only 2 games* vs teams in the bottom half of the NCAA) and they have taken care of business in these games with a 19 point average margin of victory.

Another way to look at it is that St. Mary's has really only played 2-3 "supar" games all year. Torvik gives them a game score of 50 vs Washington, 55 vs Colorado St, and 80 vs New Mexico. Otherwise, the rest of their games have a game score at 87 or higher (which is roughly equivalent to a top 25 team) and there have been 15 times when they have reached a game score of 90 (roughly equivalent to a top 10-15 team). For contrast, Duke has only reached a 90+ game score eight times so far.



*St. Mary's also inexplicitly played the Academy of Art this year, but that game doesn't count for the NET rankings. The Academy of Art actually gave them a close game than potential tourney teams Vermont, North Texas, and Hofstra.

[dlmzzz]

I thought I should point out that last night's loss to Va Tech had almost no change on our projections. Still projected to end up with a 12-8 record in the ACC, same as before. That is projected as coming in from 6th to 8th in the conference, a 93% chance of making the tourney, and as a 7 seed in the NCAAs. No sky fell last night.

I believe the key for that projection to hold is to make sure we win against:

• GA Tech on the road
• Wake at home
• Notre Dame at home
• Louisville at home

In other words, no bad (Quad 3 or 4) losses.

Sine the projection expects that we will go 12-8 in conference (and we are currently 5-4), that means the expectation is that we would 3-4 against the remaining Quad 1 & 2 games.

[tommy]

Hard to understand how St Mary's (NET 6 now) and UConn (NET 7) are still so highly rated, with both being above Kansas.

UConn has lost 6 of its last 8 games. Yes, most versus decent-to-good teams, but still. They have the great Thanksgiving win over Alabama, but come on. Lost 6 of 8 and still #6 in the NET? I know it's always said that there's noise in these rankings that takes awhile to work out of the system, but the season is 2/3 over. The noise should be out by now.

St Mary's has one Q1 win and 2 Q3 losses and they're #7? How can that be? Kansas has a nation-high 8 Q1 wins and no Q3 losses, playing in the toughest conference in the country.

These seem kinda silly to me.

[scottdude8]

Hard to understand how St Mary's (NET 6 now) and UConn (NET 7) are still so highly rated, with both being above Kansas.

UConn has lost 6 of its last 8 games. Yes, most versus decent-to-good teams, but still. They have the great Thanksgiving win over Alabama, but come on. Lost 6 of 8 and still #6 in the NET? I know it's always said that there's noise in these rankings that takes awhile to work out of the system, but the season is 2/3 over. The noise should be out by now.

St Mary's has one Q1 win and 2 Q3 losses and they're #7? How can that be? Kansas has a nation-high 8 Q1 wins and no Q3 losses, playing in the toughest conference in the country.

These seem kinda silly to me.

Important to note here: quadrant wins are calculated using the NET, but don't directly affect the NET. So the fact that St. Mary's only has 1 Q1 win and Kansas has 8 doesn't factor into the NET rankings directly. That said, the complex formula they use does supposedly factor in "beating quality teams" (https://www.ncaa.com/news/basketball...ings-explained), but that's just one of multiple factors that contribute, and it isn't clear what the exact formula is (i.e., how each of the various components are weighed).

It's worth noting that the NET does seem to track KenPom's rankings more closely than the human polls: Saint Mary's is #7 in KenPom, UConn is #6, while Kansas is #9. This follows from the fact that efficiency ratings supposedly are a major part of the NET.

Finally, and perhaps most importantly, the NET has been historically rough early in the season, but gradually equilibrates towards season's end. Over the past few years there has always been a team or two that is oddly highly rated in December/January, and those outliers aren't nearly as noticeable at the end of the season. And all that matters is your end of season NET ranking, including for the calculation of quadrant wins. It doesn't matter for us, for example, what Ohio State was ranked in the NET when we played them in December, but only what their final ranking is. Obviously there are positives (you might beat an unranked team early in the season that is one of the surprises of the season, and you want credit for that) and negatives (you could beat a team before a major contributor goes down with an injury) to this, but I think it's better than basing your "quality wins" based on what a team's AP ranking was at the time.

[pcal6vb]

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. I sometimes wonder the same thing about other computer rankings e.g. KenPom. To me, they're a useful tool to provide context to how a team has performed, their strengths/weaknesses, and how potent/efficient they can be on offense and defense, but should always be supplementary to wins and losses (and conference championships) when awarding tournament selection and seeding.