Duke MBB v Clemson (1/25, 7:00 PM EST, ESPN2) Pre-game and In-game Thread

[luvdahops]

I would take the under, but barely. I think "5" is the best guess. That's what Torvik thinks, and I suspect Pomeroy does too.

While I think this is a fair assessment, the optimist in me believes that if we get Trevor back soon and have no more injuries or missed games among our Top 6, 16-4 or even 17-3 is well within reach. Both Pomeroy and Torvik have us favored in all of our remaining games, and generally at >70% probability. Both see at Carolina (64% for KenPom, 62% for Torvik) as our toughest matchup.

[CDu]

While I think this is a fair assessment, the optimist in me believes that if we get Trevor back soon and have no more injuries or missed games among our Top 6, 16-4 or even 17-3 is well within reach. Both Pomeroy and Torvik have us favored in all of our remaining games, and generally at >70% probability. Both see at Carolina (64% for KenPom, 62% for Torvik) as our toughest matchup.

Oh absolutely. We should definitely be favored every game the rest of the way. There are a couple of challenging spots, but even 18-2 shouldn’t be off the table. If things click for the team and they stay healthy, even winning out isn’t impossible. It think 4-5 is the reasonable expectation. But it is certainly possible that this group has had its learning curve games and is ready to go on a tear.

[scottdude8]

and even more to the point, the fact that they were highly rated at the time is irrelevant, as the systems have no memory (you don't need to know what teams past ratings were to evaluate their current rating). I don't think even NET has memory.

Perhaps more importantly, the team sheet at the end of the year has no memory. It won't show that Duke beat AP #1 (at the time) Gonzaga, it'll show that Duke beat Gonzaga with their final NET ranking. Committee members may recall "at the time" rankings and this may factor into their "qualitative" evaluation of teams, but I don't believe that info factors in to the official info that's "provided" to them.

So if VaTech goes on a huge run in the ACC and finds themselves in the NET Top 30, our win over them at home will be Q1 and that's all the committee will see... there won't be an asterisk that VaTech was scuffling and lowly ranked at the time. (Not that I anticipate this happening, as a disclaimer, but just an interesting hypothetical.)

[MarkD83]

And why do we assume that FSU and Miami won't lose any games to teams with .500 or below ACC records? FSU could quite reasonably lose at Clemson, at UNC, vs Wake, at Duke, at UVa. Games vs VT and Notre Dame are losable too. They did, for example, lose by 22 to Wake in Winston-Salem, and to Syracuse at home. They are certainly not immune.

Miami has played extremely well in conference so far, but they did lose to UCF at home and got blown out by Dayton. And they struggled to beat FAU. They have several losable games left: @VT, @UVa, @Wake, @Louisville, @Syracuse, vs VT, vs ND.

It's certainly possible that those two teams avoid any pitfalls along the way. But I wouldn't assume that they will do so.

My point seems to be lost...right now based on the actual games played, there are 8 teams with better than 0.500. We always seem to complain about the balanced schedule but can never tell who has an advantage until games are played. So now as flawed as it may seem we have an actual set of data regarding who has the most games left against the top 8.

A non-hypothetical example, Miami does NOT have to play FSU again so there is no chance of another loss to FSU. Duke does have to play FSU so that is a possible extra loss.

You did bring up an interesting scheduling analysis that points out an obvious critical game.

"FSU could quite reasonably lose at Clemson, at UNC, vs Wake, at Duke, at UVa". Duke has the same set of games and an equal likelihood of not playing well or playing great in these games. So the differentiation in this part of the schedule is the Duke/FSU game which we already knew.

[uh_no]

Perhaps more importantly, the team sheet at the end of the year has no memory. It won't show that Duke beat AP #1 (at the time) Gonzaga, it'll show that Duke beat Gonzaga with their final NET ranking. Committee members may recall "at the time" rankings and this may factor into their "qualitative" evaluation of teams, but I don't believe that info factors in to the official info that's "provided" to them.

So if VaTech goes on a huge run in the ACC and finds themselves in the NET Top 30, our win over them at home will be Q1 and that's all the committee will see... there won't be an asterisk that VaTech was scuffling and lowly ranked at the time. (Not that I anticipate this happening, as a disclaimer, but just an interesting hypothetical.)

for sure. I do want to make it clear that I think that despite wins and losses having not very much predictive value on their own, I think it's totally valid for it to be considered in something like NET. Teams SHOULD be rewarded for winning games, and they should be rewarded for scheduling competitive games.

[brevity]

No, not really. We moved 2 road games early on to right into the middle of the schedule where we already had a stretch of road games.

Had the ACC schedule played out like normal, it would have been: HAAHAHAHHAAHAHHAAAHA. That's pretty balanced.

Instead, it looks like this: HHAHAHHAAAHAAHHAAAHA. We now have a 5 in 6 away stretch AND a 4 in 5 away stretch, split only by a pair of home games in mid-February. Prior to that, we had just the 4 in 5 away stretch at the end. Most more road-heavy now for the back half of the schedule.

Looks pretty funny either way.

[CDu]

"FSU could quite reasonably lose at Clemson, at UNC, vs Wake, at Duke, at UVa". Duke has the same set of games and an equal likelihood of not playing well or playing great in these games. So the differentiation in this part of the schedule is the Duke/FSU game which we already knew.

I would argue that the differentiation lies just as much in the quality of the teams. FSU, by virtue of having been generally worse than Duke all season, is more likely to lose some of those games than Duke is.

Now, we probably DO have a tougher remaining schedule (depending somewhat on whether or not UNC folds). But we are also the better team (based on overall resume) and general recent performance. So we should fare better even with the unbalanced schedule.

[Kedsy]

My point seems to be lost...right now based on the actual games played, there are 8 teams with better than 0.500. We always seem to complain about the balanced schedule but can never tell who has an advantage until games are played. So now as flawed as it may seem we have an actual set of data regarding who has the most games left against the top 8.

A non-hypothetical example, Miami does NOT have to play FSU again so there is no chance of another loss to FSU. Duke does have to play FSU so that is a possible extra loss.

You did bring up an interesting scheduling analysis that points out an obvious critical game.

"FSU could quite reasonably lose at Clemson, at UNC, vs Wake, at Duke, at UVa". Duke has the same set of games and an equal likelihood of not playing well or playing great in these games. So the differentiation in this part of the schedule is the Duke/FSU game which we already knew.

I don't think your point is lost. I think your reasoning is flawed, for two main reasons:

(1) You state the top 8 teams in the ACC are the teams with a current league record above 0.500. Very few people believe that. Louisville and Virginia are in the top 8 of the standings (above .500) but they are not considered to be "top 8" ACC teams by anybody who studies this stuff (Virginia Tech and either Clemson or Syracuse round out pretty much everybody's top 8 instead of L'Ville and UVa).

(2) You are implicitly assuming that Duke, FSU, and Miami would all have an equal chance at winning the same games in the same schedule. But that's absurd. If the three teams played the same schedule 100 times, Duke's overall winning percentage would undoubtedly be much better than the other two teams. Put another way, to say that both Duke and FSU could "quite reasonably lose at Clemson," while true on its face, suggests an equality in the chances that each team might lose that isn't there. Sure both could lose, but the chances of FSU losing that game are much higher than the chances of Duke losing that game.

Obviously nobody can say for certain what teams will win and/or lose which games. But using the current standings to judge the relative strength among teams is a poor way to judge or predict.

[MarkD83]

I would argue that the differentiation lies just as much in the quality of the teams. FSU, by virtue of having been generally worse than Duke all season, is more likely to lose some of those games than Duke is.

Now, we probably DO have a tougher remaining schedule (depending somewhat on whether or not UNC folds). But we are also the better team (based on overall resume) and general recent performance. So we should fare better even with the unbalanced schedule.

You are getting to the heart of my thoughts...FSU beat Duke so the objective fact is they are better than Duke. I will agree with you on paper and theoretically Duke at home should beat FSU, but the actual results are what I am trying to assess to see who has the easier schedule.

[CameronDuke]

You are getting to the heart of my thoughts...FSU beat Duke so the objective fact is they are better than Duke. I will agree with you on paper and theoretically Duke at home should beat FSU, but the actual results are what I am trying to assess to see who has the easier schedule.

To me, currently*, the only game Duke cannot lose in order to win the ACC regular season is the FSU at Duke game. More than any other remaining game, I feel like currently*, if they lost that one, they’re not winning the ACC regular season.

* it’s only January 24th. A lot can and will change.

[DavidBenAkiva]

Since we are having tangential conversations about the game on this thread...

Is this a trap game?

[CDu]

You are getting to the heart of my thoughts...FSU beat Duke so the objective fact is they are better than Duke. I will agree with you on paper and theoretically Duke at home should beat FSU, but the actual results are what I am trying to assess to see who has the easier schedule.

No, it is not an objective fact that they are better than Duke. It is an objective fact that they won the only game they've played so far against Duke. But (1) it was at home (which is a 3-4 point advantage on average), (2) it was just one game, and (3) the two teams have an overall resume that suggests otherwise.

[Kedsy]

FSU beat Duke so the objective fact is they are better than Duke.

We're getting into Bill Walton territory here. Are you actually suggesting (as the above quote seems to say) that after every upset, the upset winner is better than the team they beat?

So, like in 2018, Maryland-Baltimore County was a better team than Virginia?

[MarkD83]

I don't think your point is lost. I think your reasoning is flawed, for two main reasons:

(1) You state the top 8 teams in the ACC are the teams with a current league record above 0.500. Very few people believe that. Louisville and Virginia are in the top 8 of the standings (above .500) but they are not considered to be "top 8" ACC teams by anybody who studies this stuff (Virginia Tech and either Clemson or Syracuse round out pretty much everybody's top 8 instead of L'Ville and UVa).

(2) You are implicitly assuming that Duke, FSU, and Miami would all have an equal chance at winning the same games in the same schedule. But that's absurd. If the three teams played the same schedule 100 times, Duke's overall winning percentage would undoubtedly be much better than the other two teams. Put another way, to say that both Duke and FSU could "quite reasonably lose at Clemson," while true on its face, suggests an equality in the chances that each team might lose that isn't there. Sure both could lose, but the chances of FSU losing that game are much higher than the chances of Duke losing that game.

Obviously nobody can say for certain what teams will win and/or lose which games. But using the current standings to judge the relative strength among teams is a poor way to judge or predict.

Let me put this another way...I am trying to determine the "magic" number to finish first in the ACC. In baseball this "magic" number does not take into account who a team plays or any analysis of how good a team is. It just uses how many wins or opponent losses to clinch.

The next step you can take with the "magic" number is what teams are left on the schedule and mathematically can you reduce your magic number with other teams losses. (Look if Duke wins out this is a moot point.)

Extrapolate back to today and use the assumption most people have that 5 losses is the most that one can have to win the ACC. This means NCSU, Pitt and GT have been eliminated. Syr, VT and Clemson are one loss from being eliminated, but they are not the teams that need to lose multiple times for Duke to win the ACC. The other top 8 are...so in the schedule where do these other top 8 get to 6 losses without Duke getting to 6 losses.

[CameronDuke]

Since we are having tangential conversations about the game on this thread...

Is this a trap game?

No, not a trap game. If Duke loses tomorrow, it will be because they’re just not a very consistent basketball team this year. ESPN gives Duke an 88% chance of winning. Duke likely will be favored by double digits. Clemson is not a great basketball team. They’re very mediocre. They’re 63 overall on Kenpom. Their offense and defense both are mediocre (54 and 82 respectively). This is a game that if Duke loses, they honestly don’t deserve the ACC regular season title. How many trap games or Ls would it then take to get their attention to give every game they play the level of focus needed and required? Losing tomorrow would seriously make me consider how bought in the coaching staff has the team and how effective their communication with the team really is this year.

[MarkD83]

We're getting into Bill Walton territory here. Are you actually suggesting (as the above quote seems to say) that after every upset, the upset winner is better than the team they beat?

So, like in 2018, Maryland-Baltimore County was a better team than Virginia?

No but UVA had no chance to win the NCAA after the loss to UMBC. (See my magic number analogy in a separate post).

I am trying to figure out the math of when teams are mathematically eliminated from winning the regular season and where they get those losses.
Duke losing to FSU objectively hurt Duke's chance of winning the ACC regular season and helped FSUs.

So where does Miami get 4 more losses?
Where does FSU get 4 more losses?

The best way to see those is in the games where the Top 8 play each other because there is a guaranteed loss for a team that is mathematically still in it.

[CDu]

Let me put this another way...I am trying to determine the "magic" number to finish first in the ACC. In baseball this "magic" number does not take into account who a team plays or any analysis of how good a team is. It just uses how many wins or opponent losses to clinch.

The next step you can take with the "magic" number is what teams are left on the schedule and mathematically can you reduce your magic number with other teams losses. (Look if Duke wins out this is a moot point.)

Extrapolate back to today and use the assumption most people have that 5 losses is the most that one can have to win the ACC. This means NCSU, Pitt and GT have been eliminated. Syr, VT and Clemson are one loss from being eliminated, but they are not the teams that need to lose multiple times for Duke to win the ACC. The other top 8 are...so in the schedule where do these other top 8 get to 6 losses without Duke getting to 6 losses.

We get what you are trying to do. We just disagree with your logic at this stage. You appear to be basing too much on too small a sample of games to date.

We don't know how good Miami and FSU really are. Torvik's metrics suggest that FSU has been playing like the #62 team in the country in January. That puts them behind Duke (#12), Miami (#23), Notre Dame (#28), and Wake Forest (#37), NC State (#50), and UNC (#60). That is despite 3 of their six ACC games this month being wins over Duke and Miami.

Your "magic number" idea needs more data before it is worthwhile to start thinking about it. Let the teams sort themselves a bit more before trying to make declarative statements about how many losses will eliminate one from the ACC 1 seed.

[uh_no]

Since we are having tangential conversations about the game on this thread...

Is this a trap game?

always.

[DavidBenAkiva]

always.

This might be the most correct answer.

[CDu]

No but UVA had no chance to win the NCAA after the loss to UMBC. (See my magic number analogy in a separate post).

But that's not the argument you made. And it's a lazy way of replying to the point Kedsy made. Put another way: Wake beat FSU. Are they objectively better than FSU? Duke beat Wake. Are they objectively better than Wake? FSU beat Duke. Are they objectively better than Duke? The answer to those three questions can't all be "yes." Which is what Kedsy and I are saying. A single outcome does not create an objective fact that the winner is better than the loser.

I am trying to figure out the math of when teams are mathematically eliminated from winning the regular season and where they get those losses.

We get that. But your approach is flawed.

Duke losing to FSU objectively hurt Duke's chance of winning the ACC regular season and helped FSUs.

This is correct. But not terribly useful for what you are trying to do.

So where does Miami get 4 more losses?

Any number of places. They have 12 conference games remaining. It's possible that they've "flipped the switch" and won't lose 4 more times. It's also quite possible that they overperformed in two games (@Duke and vs UNC) and are due for a regression towards the team that stunk in December. We just don't know yet. Because we don't have enough of a sample of them being "good" to know for sure how good they are.

Torvik suggests they're expected to lose @VT, @UVa, @Wake, @Louisville, and @ Syracuse. They might win some of those games, but they also might lose some others. In aggregate, he expects them to go 13-7 in conference.

Where does FSU get 4 more losses?

Ditto to the above, with the additional point that FSU has played discernibly worse than Miami with the exception of eeking out wins against them twice. But everything else about their resume says they're playing worse than Miami (and much worse than Duke) in January.

Torvik suggests they're likely to lose @UNC and @Clemson, and VERY likely to lose @Duke (11-point spread). They then have a toss-up at UVa, and challenging matchups vs Wake (who destroyed them in Winston-Salem) and vs VT. In aggregate, he expects them to go 13-7 in conference.

The best way to see those is in the games where the Top 8 play each other because there is a guaranteed loss for a team that is mathematically still in it.

But the point is that we don't know yet who the top-8 are. 13 teams are within 3 losses of each other at this point, and 7 teams (teams 7-13) are within a single loss of the "top 8". There are just way too many games left, and way too many teams who haven't faced another team in the conference yet.