Page 1 of 2 12 LastLast
Results 1 to 20 of 28
  1. #1
    scottdude8's Avatar
    scottdude8 is offline Moderator, Contributor, Zoubek disciple, and resident Wolverine
    Join Date
    Jan 2010
    Location
    Storrs, CT

    An interesting hypothetical if you're worried about the tournament

    I put this as part of a long form piece on the front page, but I want to recapitulate it briefly here since I think it'll be salient to a lot of the conversations going on in various threads on the board.

    First, my premise: For Duke to not make the tourney this year, they'll likely need to have a worse resume than Notre Dame did last year. Notre Dame was one of the "Last 4 In" out of the ACC, and the conference was undeniably worse last year than it is this year. So, what were the key parts of Notre Dame's resume? Let's focus on three features: 11 total losses, 1 "bad" (Q3/4) loss, and 2 Q1 wins.

    Now, the first important point: It is extremely likely Duke will have at least 2 Q1 wins and less than 1 "bad" loss by the end of the season. The Xavier and Iowa neutral site wins are safely in Q1, and while the home Ohio State win looks precarious as our 3rd Q1 win, the home Miami win could easily ascend. We also will get a handful more opportunities before things end. Meanwhile, we only have 3 projected Q3/Q4 games left, two of which are at home. The other is this Saturday against Georgia Tech, so if we can make it out of Atlanta unscathed we'll likely be OK as it pertains to bad losses.

    With that in mind, I posit that for Duke's resume to be worse than Notre Dame's last year, it'll require losing at least five more regular season games, which would give us 11 regular season losses. If we go look at Torvik's projections, the likelihood of us losing all five of the games we're least likely to win is less than 5%.

    Obviously that number is a lot higher than we'd like, but it points to the fact that we've built a pretty nice buffer between us and the bubble, to the point that we'd have to be exceptionally bad/unlucky down the stretch to be really sweating come Selection Sunday. And this is arguably the most pessimistic scenario: if we were to say win one of @UVA, @UNC, @Miami, or home UNC, that'll likely be another Q1 win that'll give us even more of a buffer in case something extra weird happens.

    Now, let me be clear: I'm not saying that there's no chance we won't make the tourney. Torvik isn't omniscient... for instance, you might think we're going to lose all of our remaining road games because of our recent performances, which would change this hypothetical. What I'm trying to show is that, when you compare our resume to what the resumes of actual bubble teams actually look like, a LOT would have to go wrong for us to be in legitimate danger.

    I hope this provides some solace to the board, and for those who I've disagreed with on this topic, I hope that this hard analysis provides some concrete support to my argument. I look forward to continued debate
    Scott Rich on the front page

    Trinity BS 2012; University of Michigan PhD 2018
    Duke Chronicle, Sports Online Editor: 2010-2012
    K-Ville Blue Tenting 2009-2012

    Unofficial Brian Zoubek Biographer
    If you have questions about Michigan Basketball/Football, I'm your man!

  2. #2
    Quote Originally Posted by scottdude8 View Post
    I put this as part of a long form piece on the front page, but I want to recapitulate it briefly here since I think it'll be salient to a lot of the conversations going on in various threads on the board.

    First, my premise: For Duke to not make the tourney this year, they'll likely need to have a worse resume than Notre Dame did last year. Notre Dame was one of the "Last 4 In" out of the ACC, and the conference was undeniably worse last year than it is this year. So, what were the key parts of Notre Dame's resume? Let's focus on three features: 11 total losses, 1 "bad" (Q3/4) loss, and 2 Q1 wins.

    Now, the first important point: It is extremely likely Duke will have at least 2 Q1 wins and less than 1 "bad" loss by the end of the season. The Xavier and Iowa neutral site wins are safely in Q1, and while the home Ohio State win looks precarious as our 3rd Q1 win, the home Miami win could easily ascend. We also will get a handful more opportunities before things end. Meanwhile, we only have 3 projected Q3/Q4 games left, two of which are at home. The other is this Saturday against Georgia Tech, so if we can make it out of Atlanta unscathed we'll likely be OK as it pertains to bad losses.

    With that in mind, I posit that for Duke's resume to be worse than Notre Dame's last year, it'll require losing at least five more regular season games, which would give us 11 regular season losses. If we go look at Torvik's projections, the likelihood of us losing all five of the games we're least likely to win is less than 5%.

    Obviously that number is a lot higher than we'd like, but it points to the fact that we've built a pretty nice buffer between us and the bubble, to the point that we'd have to be exceptionally bad/unlucky down the stretch to be really sweating come Selection Sunday. And this is arguably the most pessimistic scenario: if we were to say win one of @UVA, @UNC, @Miami, or home UNC, that'll likely be another Q1 win that'll give us even more of a buffer in case something extra weird happens.

    Now, let me be clear: I'm not saying that there's no chance we won't make the tourney. Torvik isn't omniscient... for instance, you might think we're going to lose all of our remaining road games because of our recent performances, which would change this hypothetical. What I'm trying to show is that, when you compare our resume to what the resumes of actual bubble teams actually look like, a LOT would have to go wrong for us to be in legitimate danger.

    I hope this provides some solace to the board, and for those who I've disagreed with on this topic, I hope that this hard analysis provides some concrete support to my argument. I look forward to continued debate
    Awesome post, Scott. And certainly cannot disagree with the research and premise. Not sure if you saw yesterday, but I actually pulled these exact stats for all 4 Last Four In teams, as well as the First Four Out.

    My only counter-argument to this would be, are not all bubbles created differently? i.e. the bubble for 2023 may be stronger or weaker than 2022, and so on?

    There is some solace in the fact that, not only did Notre Dame get in last year with a much weaker resume (and one could argue Rutgers was even weaker than ND, outside the 7 Q1 wins), but they knocked off 6 seed Alabama and nearly knocked of 3 seed Texas Tech.

    As we know, anything can happen in the tournament, when it's ALL about matchups.

  3. #3
    scottdude8's Avatar
    scottdude8 is offline Moderator, Contributor, Zoubek disciple, and resident Wolverine
    Join Date
    Jan 2010
    Location
    Storrs, CT
    Quote Originally Posted by bshrader View Post
    Awesome post, Scott. And certainly cannot disagree with the research and premise. Not sure if you saw yesterday, but I actually pulled these exact stats for all 4 Last Four In teams, as well as the First Four Out.

    My only counter-argument to this would be, are not all bubbles created differently? i.e. the bubble for 2023 may be stronger or weaker than 2022, and so on?

    There is some solace in the fact that, not only did Notre Dame get in last year with a much weaker resume (and one could argue Rutgers was even weaker than ND, outside the 7 Q1 wins), but they knocked off 6 seed Alabama and nearly knocked of 3 seed Texas Tech.

    As we know, anything can happen in the tournament, when it's ALL about matchups.
    I did miss you doing that, which is a shame... good on ya!

    You'd 100% be right that all bubbles are created differently. For what it's worth, the last team in according to the consensus via BracketMatrix right now is West Virginia. It's hard to compare ourselves to them given that the NET being so in love with the Big 12 is going to skew things a lot this year... but WVU has 8 losses, is 1-4 on the road, and has 3 Q1 wins. You could argue given the Q1 similarity and similar lack of bad losses, that means we have a two game "cushion" on WVU. And they're certainly gonna lose a lot more games in the Big 12.
    Scott Rich on the front page

    Trinity BS 2012; University of Michigan PhD 2018
    Duke Chronicle, Sports Online Editor: 2010-2012
    K-Ville Blue Tenting 2009-2012

    Unofficial Brian Zoubek Biographer
    If you have questions about Michigan Basketball/Football, I'm your man!

  4. #4
    Quote Originally Posted by scottdude8 View Post
    I did miss you doing that, which is a shame... good on ya!

    You'd 100% be right that all bubbles are created differently. For what it's worth, the .last team in according to the consensus via BracketMatrix right now is West Virginia. It's hard to compare ourselves to them given that the NET being so in love with the Big 12 is going to skew things a lot this year... but WVU has 8 losses, is 1-4 on the road, and has 3 Q1 wins. You could argue given the Q1 similarity and similar lack of bad losses, that means we have a two game "cushion" on WVU. And they're certainly gonna lose a lot more games in the Big 12.
    Which is the exact reason I feel like a win on Saturday is absolutely necessary, not only for team confidence in getting a road win, but keeping that "bad loss" off the resume. Cause with GT at 180 in the NET, I don't see much chance that game becomes a Q2

  5. #5
    Speaking of WVU, it's interesting that they're currently the last projected team in the field and they sit at 29th in the NET.

    They have a chance to pick up a Q1 win tonight in Lubbock.

    EDIT: Not only a Q1 opportunity, but a Q1 opportunity against a team that's 10-9 overall and 0-7 in conference.
    Last edited by bshrader; 01-25-2023 at 07:43 PM.

  6. #6
    Join Date
    Feb 2007
    Location
    Princeton, NJ
    Quote Originally Posted by scottdude8 View Post
    I put this as part of a long form piece on the front page, but I want to recapitulate it briefly here since I think it'll be salient to a lot of the conversations going on in various threads on the board.

    First, my premise: For Duke to not make the tourney this year, they'll likely need to have a worse resume than Notre Dame did last year. Notre Dame was one of the "Last 4 In" out of the ACC, and the conference was undeniably worse last year than it is this year. So, what were the key parts of Notre Dame's resume? Let's focus on three features: 11 total losses, 1 "bad" (Q3/4) loss, and 2 Q1 wins.

    Now, the first important point: It is extremely likely Duke will have at least 2 Q1 wins and less than 1 "bad" loss by the end of the season. The Xavier and Iowa neutral site wins are safely in Q1, and while the home Ohio State win looks precarious as our 3rd Q1 win, the home Miami win could easily ascend. We also will get a handful more opportunities before things end. Meanwhile, we only have 3 projected Q3/Q4 games left, two of which are at home. The other is this Saturday against Georgia Tech, so if we can make it out of Atlanta unscathed we'll likely be OK as it pertains to bad losses.

    With that in mind, I posit that for Duke's resume to be worse than Notre Dame's last year, it'll require losing at least five more regular season games, which would give us 11 regular season losses. If we go look at Torvik's projections, the likelihood of us losing all five of the games we're least likely to win is less than 5%.

    Obviously that number is a lot higher than we'd like, but it points to the fact that we've built a pretty nice buffer between us and the bubble, to the point that we'd have to be exceptionally bad/unlucky down the stretch to be really sweating come Selection Sunday. And this is arguably the most pessimistic scenario: if we were to say win one of @UVA, @UNC, @Miami, or home UNC, that'll likely be another Q1 win that'll give us even more of a buffer in case something extra weird happens.

    Now, let me be clear: I'm not saying that there's no chance we won't make the tourney. Torvik isn't omniscient... for instance, you might think we're going to lose all of our remaining road games because of our recent performances, which would change this hypothetical. What I'm trying to show is that, when you compare our resume to what the resumes of actual bubble teams actually look like, a LOT would have to go wrong for us to be in legitimate danger.

    I hope this provides some solace to the board, and for those who I've disagreed with on this topic, I hope that this hard analysis provides some concrete support to my argument. I look forward to continued debate
    There are some pretty significant logic and math problems with this post. The most obvious one is that, while we probably won't lose all 5 of the least likely to win games, we also probably won't win all of the 5 most likely to win games. I believe that 5 more regular season losses (6-5 finish) is slightly worse than the computer models suggest is the most likely outcome, but not by much.

  7. #7
    scottdude8's Avatar
    scottdude8 is offline Moderator, Contributor, Zoubek disciple, and resident Wolverine
    Join Date
    Jan 2010
    Location
    Storrs, CT
    Quote Originally Posted by freshmanjs View Post
    There are some pretty significant logic and math problems with this post. The most obvious one is that, while we probably won't lose all 5 of the least likely to win games, we also probably won't win all of the 5 most likely to win games. I believe that 5 more regular season losses (6-5 finish) is slightly worse than the computer models suggest is the most likely outcome, but not by much.
    Of course… it wasn’t meant to be exhaustive by any means. It was meant to show that the most likely path to 5 more losses would be very unlikely. It wasn’t intended to show all the paths to 5 losses.
    Scott Rich on the front page

    Trinity BS 2012; University of Michigan PhD 2018
    Duke Chronicle, Sports Online Editor: 2010-2012
    K-Ville Blue Tenting 2009-2012

    Unofficial Brian Zoubek Biographer
    If you have questions about Michigan Basketball/Football, I'm your man!

  8. #8
    I don't think it is unlikely at all that Duke will lose 5 more games. Both KenPom and Torvik project Duke to lose 4 more games. On Torvik's win matrix page, his projections give Duke a 43.6% chance of losing at least 5 more games.

    That is not to say that Duke would miss the Tournament at 20-11(11-9). I'm not sure. Right now, Duke has an unspectacular resume without any major blemishes. Usually, teams on the wrong side of the bubble have something that really jumps out: a lot of losses, some ugly losses, a poor strength of schedule, lack of quality wins. Right now Duke is 3-6 in Q1 games and undefeated in all other games (the OSU and Wake games could drop to Q2, the others probably will stay Q1). Our SOS is 28th per NET (I'm using Warrennolan.com for this info). The consensus at Bracket Matrix has Duke as a 6 seed. In other words, Duke is comfortably in the Tournament as of today.

    But that can change. Duke has a Q3 landmine this weekend against Georgia Tech that it would do well to avoid. Hopefully it can be a "get right" game for the guys.

  9. #9
    Join Date
    Jan 2012
    Location
    Maryland
    Quote Originally Posted by bshrader View Post
    Speaking of WVU, it's interesting that they're currently the last projected team in the field and they sit at 29th in the NET.

    They have a chance to pick up a Q1 win tonight in Lubbock.

    EDIT: Not only a Q1 opportunity, but a Q1 opportunity against a team that's 10-9 overall and 0-7 in conference.
    Can somebody explain how a team (Texas Tech) with zero Q1 and Q2 wins has a NET ranking that is 77?

  10. #10
    Quote Originally Posted by DCDevil9194 View Post
    Can somebody explain how a team (Texas Tech) with zero Q1 and Q2 wins has a NET ranking that is 77?
    Because NET really doesn't have anything to do with Q1 and Q2 wins. NET is a performance metric kind of like KenPom/Sagarin/Torvik. I don't think the NET formula has ever been revealed, but usually it tracks pretty closely to those other metrics. A team doesn't move up in NET by racking up Q1 wins, but rather by playing well relative to the competition. For example, if a team played several Q3 games and won them by 40 points, their NET might go up. Similarly, close losses to highly ranked teams could cause a team's NET to rise.

    Also, it's often said that what matters is not so much your team's NET but rather the NET of your opponents. NET is a performance metric, but the Selection Committee generally rates teams by their resumes, rather than their performance (e.g. Houston last year was #2 in both KenPom and Torvik last year, but only got a 5 seed because the Committee was less impressed with their resume). One way the Committee weighs resumes is by looking at teams' records in Q1 games, for example. So really, the most important use of NET is in determining the difficulty of a team's schedule and trying to give credit accordingly. It's not a perfect system by any means.

  11. #11
    Join Date
    Feb 2007
    Location
    Earth
    I think that is a good analysis, but rule #1: the NCAA is consistently inconsistent in all sports. Ask the lacrosse team how a purportedly data-driven selection process worked out for them last year. Duke basketball is a lightning rod.

    If you are looking for excuses to keep Duke out of the tournament, away/neutral record is a concern. I would love to see more than 6 away/neutral wins in the regular season to feel safe. (Duke currently has 4.) Another related excuse can be the ACC Tournament seed. Duke is on track for the 8/9 seed in the ACC Tournament right now, and I'm not sure the NCAA will pass over more than one of seeds 5-7 to get to Duke. While that should improve with the remaining home schedule, Duke needs to move above the logjam. Conference tournament upsets and performance counted nada last year. Much of DBR, including myself, expected Duke to be a #3 seed behind TN.

  12. #12
    Join Date
    Feb 2007
    Location
    NC
    Quote Originally Posted by Dr. Maturin View Post
    I don't think it is unlikely at all that Duke will lose 5 more games. Both KenPom and Torvik project Duke to lose 4 more games. On Torvik's win matrix page, his projections give Duke a 43.6% chance of losing at least 5 more games.

    That is not to say that Duke would miss the Tournament at 20-11(11-9). I'm not sure. Right now, Duke has an unspectacular resume without any major blemishes. Usually, teams on the wrong side of the bubble have something that really jumps out: a lot of losses, some ugly losses, a poor strength of schedule, lack of quality wins. Right now Duke is 3-6 in Q1 games and undefeated in all other games (the OSU and Wake games could drop to Q2, the others probably will stay Q1). Our SOS is 28th per NET (I'm using Warrennolan.com for this info). The consensus at Bracket Matrix has Duke as a 6 seed. In other words, Duke is comfortably in the Tournament as of today.

    But that can change. Duke has a Q3 landmine this weekend against Georgia Tech that it would do well to avoid. Hopefully it can be a "get right" game for the guys.
    Yeah, I'd say there is about a 40% chance we'll lose 5 more games. But where I think the analysis still holds some water is that it isn't JUST losing 5 more games that would need to happen. We would need to lose 5 more AND to have at least one of those be a bad loss, while also not offsetting that with any additional Q1 wins. That combination seems rather unlikely.

    Now there is certainly a chance we could miss the tournament. But I think the likelihood of that is pretty small.

  13. #13
    scottdude8's Avatar
    scottdude8 is offline Moderator, Contributor, Zoubek disciple, and resident Wolverine
    Join Date
    Jan 2010
    Location
    Storrs, CT
    Quote Originally Posted by CDu View Post
    Yeah, I'd say there is about a 40% chance we'll lose 5 more games. But where I think the analysis still holds some water is that it isn't JUST losing 5 more games that would need to happen. We would need to lose 5 more AND to have at least one of those be a bad loss, while also not offsetting that with any additional Q1 wins. That combination seems rather unlikely.

    Now there is certainly a chance we could miss the tournament. But I think the likelihood of that is pretty small.
    Thanks CDu, this is the point I was trying to make... if you agree with my premise that we won't be in danger until our resume starts to look worse than Notre Dame's last year (considering the conference is significantly better this year), then what I tried to show (with some admittedly incomplete analysis, but I wasn't writing a peer-reviewed paper, haha) is that the road to get there is pretty narrow. As an example, I outlined a particular scenario that Torvik's analysis says would be the most likely way we get to 5 losses: considering how unlikely that is, it serves just as a proof of concept. If the most straightforward path to 5 losses had a 15-20% chance in Torvik's analysis, then I think we'd be sweating a lot more.

    One of these days I'm going to put aside my neuroscience research and direct my time to putting those skills to full use on my DBR posts... if someone can figure out how I can make money from that, that is
    Scott Rich on the front page

    Trinity BS 2012; University of Michigan PhD 2018
    Duke Chronicle, Sports Online Editor: 2010-2012
    K-Ville Blue Tenting 2009-2012

    Unofficial Brian Zoubek Biographer
    If you have questions about Michigan Basketball/Football, I'm your man!

  14. #14
    Quote Originally Posted by scottdude8 View Post
    I did miss you doing that, which is a shame... good on ya!

    You'd 100% be right that all bubbles are created differently. For what it's worth, the last team in according to the consensus via BracketMatrix right now is West Virginia. It's hard to compare ourselves to them given that the NET being so in love with the Big 12 is going to skew things a lot this year... but WVU has 8 losses, is 1-4 on the road, and has 3 Q1 wins. You could argue given the Q1 similarity and similar lack of bad losses, that means we have a two game "cushion" on WVU. And they're certainly gonna lose a lot more games in the Big 12.
    With WVU's win in Lubbock last night, they have moved up to 24th in the NET rankings. Bracketmatrix hasnt been updated for today yet, but I would have to assume this would move them off the "last team in" label.

  15. #15
    Join Date
    Feb 2007
    Location
    NC
    Quote Originally Posted by bshrader View Post
    With WVU's win in Lubbock last night, they have moved up to 24th in the NET rankings. Bracketmatrix hasnt been updated for today yet, but I would have to assume this would move them off the "last team in" label.
    Yep, I would suspect this would help them in that regard.

    Duke also got a boost without playing, moving up from 32 to 30 (leapfrogging UNC in the process).

    We'll probably want to cheer for Miami to continue to do well (except against us of course), as they are within striking distance of making our win in Cameron a Q1 win.

  16. #16
    Quote Originally Posted by CDu View Post
    Yep, I would suspect this would help them in that regard.

    Duke also got a boost without playing, moving up from 32 to 30 (leapfrogging UNC in the process).

    We'll probably want to cheer for Miami to continue to do well (except against us of course), as they are within striking distance of making our win in Cameron a Q1 win.
    The Wake Forest game in Cameron next week is teetering near the Q2/3 line for now. It's in the Q2 territory as of today. Wake gets State in Winston this weekend, so if they can hold serve, I would suspect that game vs Wake will be a Q2 game for us.

  17. #17
    Join Date
    Feb 2007
    Location
    NC
    Quote Originally Posted by bshrader View Post
    The Wake Forest game in Cameron next week is teetering near the Q2/3 line for now. It's in the Q2 territory as of today. Wake gets State in Winston this weekend, so if they can hold serve, I would suspect that game vs Wake will be a Q2 game for us.
    It doesn't matter what they do against State. The Q1/Q2 will not be final until Selection Sunday, not based on the standings at the time of the game. So Wake has lots of opportunities to improve or get worse.

  18. #18
    Quote Originally Posted by CDu View Post
    It doesn't matter what they do against State. The Q1/Q2 will not be final until Selection Sunday, not based on the standings at the time of the game. So Wake has lots of opportunities to improve or get worse.
    Yea I know. Lots of time for adjustments in the NET and others sites. Just for discussion purposes.

  19. #19
    scottdude8's Avatar
    scottdude8 is offline Moderator, Contributor, Zoubek disciple, and resident Wolverine
    Join Date
    Jan 2010
    Location
    Storrs, CT
    Quote Originally Posted by CDu View Post
    Yep, I would suspect this would help them in that regard.

    Duke also got a boost without playing, moving up from 32 to 30 (leapfrogging UNC in the process).

    We'll probably want to cheer for Miami to continue to do well (except against us of course), as they are within striking distance of making our win in Cameron a Q1 win.
    Yup. I'll probably do an in-depth rooting interest post/article soon (I, for one, enjoy having rooting interests in random games, it makes them a bit more engaging personally). But some of the easy ones include:
    - Rooting for our non-conference wins (Xavier, Iowa, and (ugh) Ohio State). Xavier and Iowa are pretty safely Q1 wins since they were on a neutral court, but Ohio State is teetering since it was at home.
    - Miami, since they could conceivably make it into Q1 if they remain towards the top of the ACC.
    - NC State, since they could also make it into the Top 30 and give us another Q1 home opportunity.
    - Wake, since that's the loss that's most vulnerable to falling into Q2 at the moment.
    Scott Rich on the front page

    Trinity BS 2012; University of Michigan PhD 2018
    Duke Chronicle, Sports Online Editor: 2010-2012
    K-Ville Blue Tenting 2009-2012

    Unofficial Brian Zoubek Biographer
    If you have questions about Michigan Basketball/Football, I'm your man!

  20. #20
    Join Date
    Nov 2022
    Location
    Atlanta
    I'm not as worried about Duke getting into the tournament as I am about them exiting it very quickly.

Similar Threads

  1. (Hypothetical) NBA College Alumni 3 on 3 Tournament
    By Big Pappa in forum Elizabeth King Forum
    Replies: 15
    Last Post: 05-08-2012, 04:35 PM
  2. I am worried about Purdue because...
    By hurleyfor3 in forum Elizabeth King Forum
    Replies: 76
    Last Post: 03-24-2010, 02:12 PM
  3. sagegrouse, you worried???
    By grossbus in forum Off Topic
    Replies: 4
    Last Post: 03-11-2010, 12:07 AM
  4. Is anyone else worried about the UNC game?
    By bluepenguin in forum Elizabeth King Forum
    Replies: 36
    Last Post: 02-03-2010, 07:36 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •