Good (NET) News, Everyone! Updates to Duke's (and our rivals') resumes

[scottdude8]

Following my NET-driven "viewing guide" post of a few days ago, I figured it would be worthwhile to post some updates occasionally on the teams I spotlighted, as well as some new teams to keep an eye on as things develop.

As a whole, yesterday was a good day when it comes to our NET-driven NCAA Tourney resume! Most importantly, Georgetown's upset win on the road against Butler put them back in the NET Top 50. That makes our neutral site victory over the Hoyas a much-needed Q1 win. If this holds, it would be our third Q1 non-conference win, which looks much better in comparison to the four currently held by both Baylor and Kansas. FWIW, Gonzaga also has 3 Q1 non-conference wins.

Virginia Tech also got back on the winning track, giving them a boost in the NET rankings: the Hokies are now 76 in the NET rankings, one spot away from our road victory over them becoming a Q1 win. Meanwhile, Georgia Tech also remains in shouting distance of the NET Top 75, as they currently sit at 86th. It's also worth keeping an eye on both Virginia and N.C. State, as we'd like them to stay in the Top 75 for our upcoming road games against them to be of the Q1 variety: Virginia should be safe, as they stand at a NET ranking of 55 after breaking UNC's hearts (making that extra enjoyable!), while N.C. State has a little less room for error at NET No. 61.

If the Hokies can squeak back into the Top 75, the Hoyas stay in the Top 50, and Virginia and N.C. State don't collapse, that would give us the potential for 8 Q1 wins come ACC Tournament time. That would compare MUCH better to the 10 Q1 wins currently held by Kansas. However, a couple of those wins are tenuous for the Jayhawks: Oklahoma State is NET 69, and Iowa State is NET 70. Interestingly it was Texas, not Iowa State, who was inside the NET Top 75 when I wrote my article, but then the Longhorns got demolished by the Cyclones. That's probably for the best, though, as without projected first-round pick Tyrese Haliburton for the rest of the year, Iowa State seems a likely candidate for a backslide. Oklahoma State's upset win over Texas Tech unfortunately helped the Cowboys' NET ranking, though.

This means that, while Kansas currently has 10 Q1 wins, two of them are tenuous at best, and there aren't any clear possibilities for a Q2 win to jump into a Q1 victory on their resume. So if Kansas can't add another Q1 win at Baylor or at Texas Tech, the divide between our resume's could be closing quickly.

[SCMatt33]

I don’t think which side of the line these games fall on matter as much as much as people like to think. When you get into the committee room, they know the difference between Duke’s win over Kansas and Duke’s win over Georgetown, even though both count as a Q1 win right now. But you start seeing “blind resumes” this time of year, that treats all Q1 wins as equal and treats all Q2 wins as universally worse than a Q1 win. No one would ever argue that Duke’s Georgetown win is closer to the Kansas win than it is the VT win, but that’s what happens when you only look at the quad numbers. Duke has 4 wins that are borderline (Gtown, VT, GT and Syracuse) that could reasonably fall on one side of the line or the other come March. I think a better way to look at the quads are as mile markers. They make a good reference point to compare home wins vs road wins, and if one team has 10 wins on side of a marker while another has 2, that’s gonna look striking, as it’s unlikely that a team with 10 of em is gonna have 9 that are borderline and all falling on one side. Duke right now has 2 on each side of the line and chances are it will stay that way, but if it ends up as 3-1 or 1-3, that’s only gonna make a difference to people who buy into the “blind resume” silliness, and none of those people are on the committee

[UrinalCake]

Appreciate the breakdown, that’s good stuff. I would hope that if Duke and Kansas have comparable resumes, our head to head win over them would be the tie breaker for a 1-seed, even though I know that h2h is not a metric that is used and that game was a lifetime ago. I’m also hoping Michigan State can pull it together, our road win over them felt huge at the time but now it’s lost a lot of luster. It’s still a Q1 win by far, but may not be considered a “marquee” win by the Selection Committee. We don’t have many of those on our resume since the ACC is a train wreck and we lost our only game with Louisville.

[MarkD83]

Wouldn't it make sense to count a win or loss as Q1, Q2 etc based on the ratings of the teams when they played, rather than at the end of the year.

When Duke played G'town they still had their 2 best players, who were subsequently dismissed from the team. This was clearly a Q1 win regardless of what Gtown does the rest of the year without their best players.

[uh_no]

Wouldn't it make sense to count a win or loss as Q1, Q2 etc based on the ratings of the teams when they played, rather than at the end of the year.

When Duke played G'town they still had their 2 best players, who were subsequently dismissed from the team. This was clearly a Q1 win regardless of what Gtown does the rest of the year without their best players.

that would make our SFA win look so much worse.

IMO the value of having more data on teams far outweighs the consequence that in a small number of cases, a game is mis-valued...and it's not even close.

Injuries are a tough thing to deal with in a sane way, I agree, but discarding on average half a season's worth of data for every other game which wasn't so affected is not the right solution.

Given, the Q thing is stupid to begin with.

[scottdude8]

Everyone’s criticisms of the Q system and that the committee “should” view things more holistically are 100% valid. That said, every outward indication (and we can debate how honest the NCAA is about this, haha) is that they do look heavily at the quadrant system. So as silly as it may seem, it very well may affect our seeding in a few weeks.

[SCMatt33]

Everyone’s criticisms of the Q system and that the committee “should” view things more holistically are 100% valid. That said, every outward indication (and we can debate how honest the NCAA is about this, haha) is that they do look heavily at the quadrant system. So as silly as it may seem, it very well may affect our seeding in a few weeks.

Records against quads absolutely matter in general. If you think that Georgetown being 49 vs 51 and Duke having a 7-1 record vs a 6-1 record when it’s all said and done will change our seed, you’re putting too much weight on it. Does 7-1 vs 3-2 make a difference, absolutely. Does 10-1 vs 5-1 make a difference, yes. But the committee isn’t going to flip someone’s seed because Georgetown won or lost in the big east quarterfinals and changes whether they’re Q1 or Q2. They’re not going to show up on Wednesday with Georgetown as a Q2 and have Duke as a 2 seed, then have Georgetown win a game and move up and say “you know what, I think Duke’s a 1 seed now.” If two teams are in the same ballpark, they’re going to look at individual games and break them down further then that

Does it matter that in a down ACC year Duke is only going to have 8-10 total Q1 games vs another year where Duke may have 13-15 with the same schedule, yes, but that’s why we’re discussing a 3 loss Duke needing help for a 1 seed vs most years where 5-6 losses would have us in the conversation. It doesn’t mean that Georgetown is going to be some super awesome win that propels us past SDSU or Gonzaga because it’s Q1 or let’s Maryland pass us and kick us out of the east because it’s Q2. It’s a decent win against a bubble team today and it will be a decent win against a bubble team tomorrow.

Might it look different in the committee room if all four borderline games fall on one side or the other, perhaps, but if all four of them are playing well or poorly enough to definitively fall on one side or the other, there probably is a material difference in how good those teams really were.

[BD80]

What are the chances our games with unc become Q1, for us?



snort!

[TruBlu]

What are the chances our games with unc become Q1, for us?



snort!

Pretty good, once unc claims that their record is a typo.

[duke2x]

I don’t think which side of the line these games fall on matter as much as much as people like to think. When you get into the committee room, they know the difference between Duke’s win over Kansas and Duke’s win over Georgetown, even though both count as a Q1 win right now.

I agree. The 1 line is a little more political than the others. Gonzaga is a lock. SDSU is a lock if they are undefeated. The only open spot on the 1 line is going to come between Baylor and Kansas. We will be first in line for that spot if we keep winning, but we might have to win out.

[duke2x]

As this gets easily lost in the shuffle, here is a live link to the NET rankings. With no change to the top, Duke would still be #2 in the East. Maryland at #7 would be the 2nd team in Greensboro but in a different region.

https://www.ncaa.com/rankings/basket...l-net-rankings

[Bluedog]

As this gets easily lost in the shuffle, here is a live link to the NET rankings. With no change to the top, Duke would still be #2 in the East. Maryland at #7 would be the 2nd team in Greensboro but in a different region.

https://www.ncaa.com/rankings/basket...l-net-rankings

NET rankings are not to be confused with seed list...San Diego St is currently #1 in NET but would definitely not be the overall #1 seed. Yes, if we end as 6th or 7th on the SEED LIST, we'd likely be the #2 in the East.

[Acymetric]

Records against quads absolutely matter in general. If you think that Georgetown being 49 vs 51 and Duke having a 7-1 record vs a 6-1 record when it’s all said and done will change our seed, you’re putting too much weight on it. Does 7-1 vs 3-2 make a difference, absolutely. Does 10-1 vs 5-1 make a difference, yes. But the committee isn’t going to flip someone’s seed because Georgetown won or lost in the big east quarterfinals and changes whether they’re Q1 or Q2. They’re not going to show up on Wednesday with Georgetown as a Q2 and have Duke as a 2 seed, then have Georgetown win a game and move up and say “you know what, I think Duke’s a 1 seed now.” If two teams are in the same ballpark, they’re going to look at individual games and break them down further then that

Does it matter that in a down ACC year Duke is only going to have 8-10 total Q1 games vs another year where Duke may have 13-15 with the same schedule, yes, but that’s why we’re discussing a 3 loss Duke needing help for a 1 seed vs most years where 5-6 losses would have us in the conversation. It doesn’t mean that Georgetown is going to be some super awesome win that propels us past SDSU or Gonzaga because it’s Q1 or let’s Maryland pass us and kick us out of the east because it’s Q2. It’s a decent win against a bubble team today and it will be a decent win against a bubble team tomorrow.

Might it look different in the committee room if all four borderline games fall on one side or the other, perhaps, but if all four of them are playing well or poorly enough to definitively fall on one side or the other, there probably is a material difference in how good those teams really were.

I think you're giving more credit to the committee than is warranted. I absolutely believe 6-1 vs 7-1 quad wins (and the accompanying additional quad 2 loss) could impact seeding for two teams who are close, especially at the 1/2 line.

[scottdude8]

More good news: in the latest rankings, Virginia Tech is at No. 75, meaning they represent Duke's sixth Q1 win! Meanwhile, two of Kansas and Baylor's Q1 wins, at Iowa State and Oklahoma State, are looking tenuous, as the Cyclones and Cowboys fall at No. 73 and 74 in the NET.

If Virginia Tech continues to ascend, while Oklahoma State and Iowa State fall out of the Top 75, that would leave both Baylor and Kansas with 8, as opposed to 10, Q1 victories. It's a lot easier to compare resumes when Duke is within shouting distance of these guys in the Q1 category the committee seems to love, as opposed to 4-5 wins away.

I'd argue that Duke holds the "tiebreaker" over Kansas for a No. 1 seed considering the head-to-head contest (which has to mean SOMETHING to the committee). But in order for that to come into play, we need to have an argument that our resume is on par with Kansas'. In a scenario where Kansas loses two more games this year (one to either Baylor or Texas Tech, both on the road, and one in the Big 12 Tourney), and we lose one more game (ideally winning the ACC Tourney), that would likely give us a resume with one less loss and within a game or two in the Q1 category of Kansas. Despite all the nonsense from various talking heads saying Kansas and Baylor are both "locks" to be No. 1 seeds, in that scenario it would be hard to justify choosing Kansas over Duke and justifying it, IMHO.

So the rooting interest is clear: keep pulling for Virginia Tech (except when they play us on Saturday, obviously), and against Oklahoma State and Iowa State.

[UrinalCake]

In the "getting waaay ahead of myself" department, has a team ever won both the ACC regular season and the ACC Tournament and NOT gotten a #1 seed?

[scottdude8]

In the "getting waaay ahead of myself" department, has a team ever won both the ACC regular season and the ACC Tournament and NOT gotten a #1 seed?

You're not getting too ahead of yourself because I was thinking the same thing!

Doing a very quick look at recent years where there was a sole-regular season champion that also won the Tourney:
2018 Virginia was a No. 1 seed
2016 North Carolina was a No. 1 seed
2014 Virginia was a No. 1 seed
2013 Miami was NOT a No. 1 seed! They got the top No. 2 seed.

[uh_no]

In the "getting waaay ahead of myself" department, has a team ever won both the ACC regular season and the ACC Tournament and NOT gotten a #1 seed?

https://xkcd.com/1122/

(Note: this column is NOT about politics, but about overfitting)

It's not a perfect analogy, as it's about things that clearly have no bearing on the outcome...but in this case, winning titles obviously does...but the point I think most applies here is that just because something has always happened in the past doesn't mean it must happen now...with varying degrees of correlation.

I would venture to say that if we won both titles, we would probably be a 1...just by virtue of having won a bunch of games, and likely not having lost in a while...not due to the titles themselves.

That said, the ACC hasn't been this weak since pre-re-alignment, and generally never is...so i'm not sure we can take performance vs the league in much of a historical context.