Bracketology

[Wahoo2000]

It is waaaaaay too early to say what will or won't be difficult when it comes to seeding. I remember in mid-January, 2010, after Duke lost to NC State, many DBR posters said with conviction that it was then impossible for us to get a #1 seed, because Kansas, Texas, Kentucky, and (I think) Villanova were locked in. As it happened, Texas (#1 at the time) ended up an 8-seed. Villanova ended up the #2 in Duke's region.

Nothing is etched in stone at this point.

It can get way later than that to not be "etched in stone". In 2015 we were 28-1 going into the last game of the regular season. We had been #2 for over half of the year (KY was undefeated and #1 all regular season), and I was just CERTAIN we had a 1 seed pretty much locked up. Mangok Mathiang, one of the worst shooters in the history of ACC basketball, hit what amounted to a buzzer beater from a step beyond the elbow to beat us at Louisville in a game played where we'd already locked the ACC regular season and #1 seed for the conference tournament. We still won a game in the ACC tourney, and didn't have a bad loss (to UNC in the semis). Finished 29-3. Somehow we ended up a 2 seed (don't think I've EVER seen a team with our resume get a 2 seed, though many people say that there were more elite teams in 2015 than any other year in the last 20-30 seasons - Kenpom metrics actually bear that out), and it COST us - getting matched up with MSU in the second round, who took us out. Not that I'm still bitter about it or anything.... :-D

[wobatus]

I'm less concerned with Lunardi... my concern is how the selection committee makes its decisions.

I'm not gonna lie... I'm still pissed off about last year. There is NO WAY the Cheats deserved a #1 seed over us. No effing way. And then to put us in a position to have to play a home game against South Carolina in the 2nd round...

Still makes my blood boil when I think about it.

I think it's because you lost the ACC regular season by 3 games to UNC. Even with an unbalanced schedule, a bit much. Loss at home to last year's NC State team almost counts as 2 (kiding, no offense Packfan, but you weren't too good last year as you'd acknowledge; Keatts will get'em right). And somewhat moots the unbalanced schedule argument. OK, 2-1 head to head. And 15-7 against ACC teams versus 15-5 for UNC if you count the ACC tournament. So quite close with the unbalanced schedule, but again the NC State loss hurts. Georgia Tech wasn't a great loss for UNC but they played a bit better than thought in the league (blow-out loss to you guys excepted), and that was a road loss. Anyway, the 2-1 head to head, so against everyone else in the ACC Duke went 13-6, UNC 14-3. Close. I can see the argument. But it wasn't a "no way no how" thing I think.

As for South Carolina, very tough to play them in SC as 7 seed, no doubt. Although Arkansas, 8th seed in UNC's bracket, was 12-6 SEC, 25-9 pre-tournament. South Carolina was 12-6, 22-10. And had gone 2-5 its last 7 pre-NCAA's. Arkansas had gone 8-3 its last 11 pre-tournament. Including a win at South Carolina, true road game. And almost beat UNC.

Given how it played out hard to say the choice was all that wrong. Even if it may have had some bearing on how things played out. I get your blood boiling. And that you think there's no way UNC should have had the number 1 seed over Duke. It's hard for me to be objective but to me it's a very close call and not necessarily the wrong one.

[wobatus]

Part of why I somewhat buy in to the NC State delusional storyline is because of how Herb Sendek was treated. Yeah, I know he could not beat Duke or UNC (8-38 record against them), but the guy was run out of town after making 5 consecutive NCAA tourneys and had a winning record in ACC play in 4 of his final 5 season. The one year he didn't have a winning record in that span, his team still made the NCAA Sweet 16.

I get that there was a perception that his NC State teams were doomed to be good but never great and that merely making the Big Dance and never dancing for very long can really grate on a fan base, but he sure seemed like he was doing a better job than the respect that he got from fans. I'll say this, you would probably LOVE to have that level of failure back again, wouldn't you?

-Jason "it should also be noted that Herb's teams never felt all that interesting or exciting... still, winning counts for something" Evans

It's early yet but I think Keatts will get them past the Herb level. Heck even Gott got them to 2 Sweet 16s. After 2015 he'd just knocked off a number 1 seed in Villanova and everyone was saying Jay Wright couldn't win the big game. Time flies. Anyway, I think Keatts will be really good. You can already see the philosophy. NC St is 22nd in opponent turnovers per possession, 25th in OR%, thus 23rd in extra scoring chances per game. They are 25th in effective possession ratio. They are just ok at 64th in turnovers per possession on offense. Johnson suspended but they only had 9 TOs against UNC-Greensboro (Wes Miller also doing a good job) with him out. Still, last year Wilmington had the lowest turnover per possession rate in the country (right ahead of Notre Dame). And 3rd in extra scoring chances per game at 7.3 (right behind West Virginia and UNC). But last year they didn't really turn people over that much, contrary to Keatts' rep (71st in opponent TO%). Different ways to get to the same result. UNC pounded the offensive boards. West Virginia turned you over and pounded the boards. UNC-Wilmington did a little of that but protected the ball like gold.

So Keatts seems to know what he is doing. Gott had his moments and recruited well, but was more roll-the-ball-out-there.

Sendek to me is kind of like Dixon at Pitt. You don't know what you got 'til it's gone. And Dixon had even more success, a regular in the upper echelon of Kenpom until recently. But similar lack of tourney success. He seems to have turned TCU right around.

[bullettoothtony]

I think it's because you lost the ACC regular season by 3 games to UNC. Even with an unbalanced schedule, a bit much. Loss at home to last year's NC State team almost counts as 2 (kiding, no offense Packfan, but you weren't too good last year as you'd acknowledge; Keatts will get'em right). And somewhat moots the unbalanced schedule argument. OK, 2-1 head to head. And 15-7 against ACC teams versus 15-5 for UNC if you count the ACC tournament. So quite close with the unbalanced schedule, but again the NC State loss hurts. Georgia Tech wasn't a great loss for UNC but they played a bit better than thought in the league (blow-out loss to you guys excepted), and that was a road loss. Anyway, the 2-1 head to head, so against everyone else in the ACC Duke went 13-6, UNC 14-3. Close. I can see the argument. But it wasn't a "no way no how" thing I think.

As for South Carolina, very tough to play them in SC as 7 seed, no doubt. Although Arkansas, 8th seed in UNC's bracket, was 12-6 SEC, 25-9 pre-tournament. South Carolina was 12-6, 22-10. And had gone 2-5 its last 7 pre-NCAA's. Arkansas had gone 8-3 its last 11 pre-tournament. Including a win at South Carolina, true road game. And almost beat UNC.

Given how it played out hard to say the choice was all that wrong. Even if it may have had some bearing on how things played out. I get your blood boiling. And that you think there's no way UNC should have had the number 1 seed over Duke. It's hard for me to be objective but to me it's a very close call and not necessarily the wrong one.

I can't remember all the details but I'll say a couple of things.

The unbalanced schedule was a huge factor in the standings. Duke played almost every team in the top 7 or 8 of the standings twice I think. Carolina played something like 4 or 5 road games against teams in the top ten or eleven of the standings and lost all of them or all but one of them. And going into the NCAA Tournament, Duke had more RPI Top 25 wins, and more RPI Top 50 wins.

Again, all of this is from memory... I just can't remember every damn detail but I do remember that the only thing in Carolina's favor was fewer conference losses (and I think the committee has said repeatedly over the years that conference standings aren't factored in because of unbalanced schedules). Factor in the head-to-head, injury issues,the ACC Tournament Championship, and more RPI wins and it was a slam dunk. And every talking head (i.e. people who cover/analyze the sport for a living and probably knew more about it than both of us) I can recall on Selection Sunday agreed with me.

My apologies for the thread hijack but I had to respond. I'll say no more on the matter.

[UrinalCake]

Yeah Duke had more top 25 wins by far, something like 12 compared to 7 or 8 for UNC (i’m doing this all from memory). And UNC’s were all at home. We had more top-25 wins away from home just in the ACC tournament than UNC had all year. Basically we had more good wins and also more bad losses, whereas UNC was more even keel. They had no bad losses except for GT but also no good wins except the one win over us. And really, NC State was our only BAD loss (sub-75 RPI).

Keep in mind that our team had injuries all season long, which the selection committee decided to completely ignore even though they factor it in for every other team. The Kansas loss came without three freshmen, VT was without Grayson (that was a suspension, so maybe ignore that one), Louisville and FSU were without Amile, etc.

We played a final four team in what was essentially a true road game in the round of 32. If UNC has gotten that draw, and played the way they did against Arkansas, I am certain they would have been the ones going gone.

[pfrduke]

I recall one of the issues being that they didn't ever do a head-to-head comparison between Duke and UNC. They concluded that we were 7th and our resume was not better than 5 (Kentucky) or 6 (Arizona) and they concluded that UNC at 4 was better than Kentucky at 5, so Duke/UNC head-to-head never happened. Weirdly, it was a little bit harder to make the case that we deserved a 1 over UK or UA than it was over UNC, yet UK and UA couldn't get ahead of UNC.

[BandAlum83]

I recall one of the issues being that they didn't ever do a head-to-head comparison between Duke and UNC. They concluded that we were 7th and our resume was not better than 5 (Kentucky) or 6 (Arizona) and they concluded that UNC at 4 was better than Kentucky at 5, so Duke/UNC head-to-head never happened. Weirdly, it was a little bit harder to make the case that we deserved a 1 over UK or UA than it was over UNC, yet UK and UA couldn't get ahead of UNC.

The transitive property as applied to Bracketology. It will get you every time!

[wobatus]

Yeah Duke had more top 25 wins by far, something like 12 compared to 7 or 8 for UNC (i’m doing this all from memory). And UNC’s were all at home. We had more top-25 wins away from home just in the ACC tournament than UNC had all year. Basically we had more good wins and also more bad losses, whereas UNC was more even keel. They had no bad losses except for GT but also no good wins except the one win over us. And really, NC State was our only BAD loss (sub-75 RPI).

Keep in mind that our team had injuries all season long, which the selection committee decided to completely ignore even though they factor it in for every other team. The Kansas loss came without three freshmen, VT was without Grayson (that was a suspension, so maybe ignore that one), Louisville and FSU were without Amile, etc.

We played a final four team in what was essentially a true road game in the round of 32. If UNC has gotten that draw, and played the way they did against Arkansas, I am certain they would have been the ones going gone.

I don't think you can say "no good wins". Holding serve at home against good teams counts.

Whatever alchemy goes into selections (and they don't just rely on RPI, but also look at kenpom, etc), it is likely the NC State loss did you in. It's simply a worse loss than at Georgia Tech. A home loss to a team that ended up RPI 140 versus a road loss to RPI 78 (at least that's how Teamranking has it). The kenpom ranks were 109 and 77.

Either way you play the games in front of you. UNC beat Arkansas (KP #35), Butler (KP #25), Kentucky (KP #4), Oregon (KP #10), Gonzaga (KP #1). That's all of course hindsight, but they didn't have a cakewalk. RPI actually had it at season's end as Arkansas 28, Butler 12, Kentucky 2, Oregon 7, Gonzaga 6.

A lot of luck in that run. As there often is.

Duke had a lot of injury woes. Not sure how important you view Theo Pinson to UNC's team last year, but he missed their first 4 losses. Isaiah Hicks missed the loss in Cameron. Kenny Williams missed the last 2 losses. I know, cry me a river.

If the schedules were reversed the regular season would have looked different and I assume Duke would have had a 1 seed. It was pretty close as is. This year the schedule is reversed somewhat. UNC goes to Virginia, Notre Dame, Florida State, Louisville and Syracuse this year. Syracuse in February. Maybe I'll make that trek. Did you know Syracuse gets more annual snowfall than Buffalo?

Like UNC last year, Duke has opened with a road loss. But hold serve at home against good teams and the regular season title runs through Durham.

Oh my, just checked RPI and it's Duke # 1, UNC #2. That's at ESPN and NCAA. Teamrankings has it reversed. Early season wonkiness, but still kinda funny.

[wobatus]

I don't think you can say "no good wins". Holding serve at home against good teams counts.

Whatever alchemy goes into selections (and they don't just rely on RPI, but also look at kenpom, etc), it is likely the NC State loss did you in. It's simply a worse loss than at Georgia Tech. A home loss to a team that ended up RPI 140 versus a road loss to RPI 78 (at least that's how Teamranking has it). The kenpom ranks were 109 and 77.

Either way you play the games in front of you. UNC beat Arkansas (KP #35), Butler (KP #25), Kentucky (KP #4), Oregon (KP #10), Gonzaga (KP #1). That's all of course hindsight, but they didn't have a cakewalk. RPI actually had it at season's end as Arkansas 28, Butler 12, Kentucky 2, Oregon 7, Gonzaga 6.

A lot of luck in that run. As there often is.

Duke had a lot of injury woes. Not sure how important you view Theo Pinson to UNC's team last year, but he missed their first 4 losses. Isaiah Hicks missed the loss in Cameron. Kenny Williams missed the last 2 losses. I know, cry me a river.

If the schedules were reversed the regular season would have looked different and I assume Duke would have had a 1 seed. It was pretty close as is. This year the schedule is reversed somewhat. UNC goes to Virginia, Notre Dame, Florida State, Louisville and Syracuse this year. Syracuse in February. Maybe I'll make that trek. Did you know Syracuse gets more annual snowfall than Buffalo?

Like UNC last year, Duke has opened with a road loss. But hold serve at home against good teams and the regular season title runs through Durham.

Oh my, just checked RPI and it's Duke # 1, UNC #2. That's at ESPN and NCAA. Teamrankings has it reversed. Early season wonkiness, but still kinda funny.

Damn, Duke did have it rough last year. @ Louisville, Virginia, Notre Dame, Florida State (plus at home), @ UNC (+ at home), @ Virginia Tech, @ Syracuse (always tough in mid-winter).

By the way, Syracuse usually wins the Golden Snowball (don't eat it):

https://en.wikipedia.org/wiki/Golden_Snowball_Award

[wobatus]

The transitive property as applied to Bracketology. It will get you every time!

Kentucky must have loved that since they beat UNC on a neutral court during the OOC.

[brevity]

17 years later, NCAA Selection Committee enters the 21st century.

For the first time, committee members will see rankings from ESPN's BPI and Strength of Record, KPI, Jeff Sagarin Ratings and KenPom.com next to the RPI on the official team sheets used to assess each program.

They'll also see the average score of the six advanced metrics for each team, which could be repackaged as a new, official tool that replaces the RPI next season, said David Worlock, media coordinator for the NCAA tournament.

"Fans of the University of North Carolina need not worry," Worlock added. "Roy Williams will always be our favorite tool."

[duke4ever19]

Lunardi's latest bracket has us as a 1-seed, but the rest of our bracket is hella scary:

2-seed: Michigan St.

3-seed: Oklahoma

4-seed:Arizona

5-seed: Kentucky

6-seed: Gonzaga

9-seed: Texas

I know it's all make-believe -- especially in January -- but just pretending for a second that this was the bracket we were given . . . that would be insane.

[moonpie23]

that is DEFINITELY made up purposely for clickbait and fear/joy mongering.

[Kedsy]

I know it's all make-believe -- especially in January -- but just pretending for a second that this was the bracket we were given . . . that would be insane.

that is DEFINITELY made up purposely for clickbait and fear/joy mongering.

If you really look at it, why is that bracket insane?

#1 seed expected to be #1 to #4: Duke is #4 in KenPom; #1 in RPI; #4 in AP rankings
#2 seed expected to be #5 to #8: Michigan State is #5 in KenPom; #26 in RPI; #6 in AP (the #26 implies they may not deserve a #2, but supposedly reliance on RPI will be lessened this year...)
#3 seed expected to be #9 to #12: Oklahoma is #19 in KenPom; #8 in RPI; #12 in AP (here, KenPom implies they may not deserve a #3 and the average would suggest a #4, but it's pretty close)
#4 seed expected to be #13 to #16: Arizona is #22 in KenPom; #17 in RPI; #11 in AP (based on these, probably ought to be a #5)
#5 seed expected to be #17 to #20: Kentucky is #33 in KenPom; #18 in RPI; #29 in AP (based on these, probably ought to be a #7)
#6 seed expected to be #21 to #24: Gonzaga is #7 in KenPom; #57 in RPI; #15 in AP (so much variance it's hard to call this one, average would make them a #7)
#9 seed expected to be #33 to #36: Texas is #40 in KenPom; #42 in RPI; worse than #38 in AP (based on these looks more like a #10 or #11)

So, other than Duke, you could make a decent argument that every single team in this bracket is seeded too high. In other words, this bracket is probably too easy, not too hard.

[Bluedog]

If you really look at it, why is that bracket insane?

#1 seed expected to be #1 to #4: Duke is #4 in KenPom; #1 in RPI; #4 in AP rankings
#2 seed expected to be #5 to #8: Michigan State is #5 in KenPom; #26 in RPI; #6 in AP (the #26 implies they may not deserve a #2, but supposedly reliance on RPI will be lessened this year...)
#3 seed expected to be #9 to #12: Oklahoma is #19 in KenPom; #8 in RPI; #12 in AP (here, KenPom implies they may not deserve a #3 and the average would suggest a #4, but it's pretty close)
#4 seed expected to be #13 to #16: Arizona is #22 in KenPom; #17 in RPI; #11 in AP (based on these, probably ought to be a #5)
#5 seed expected to be #17 to #20: Kentucky is #33 in KenPom; #18 in RPI; #29 in AP (based on these, probably ought to be a #7)
#6 seed expected to be #21 to #24: Gonzaga is #7 in KenPom; #57 in RPI; #15 in AP (so much variance it's hard to call this one, average would make them a #7)
#9 seed expected to be #33 to #36: Texas is #40 in KenPom; #42 in RPI; worse than #38 in AP (based on these looks more like a #10 or #11)

So, other than Duke, you could make a decent argument that every single team in this bracket is seeded too high. In other words, this bracket is probably too easy, not too hard.

I think the reason one might view it as a "bracket of death" is because many believe Oklahoma, Arizona, Kentucky and Gonzaga have a lot talent on their teams and have the POTENTIAL to beat top quality opponents -- but just haven't always consistently done so throughout the season. Basically, pre-season expectations are high for those teams, which are a decent indicator of NCAA success shockingly. That's why 538 uses pre-season rankings in their projections and why an 8-seed Kentucky who made the Final Four wasn't THAT huge of a shock/upset compared to what one would typically expect of an 8-seed. So, I'd agree that that would be a killer bracket with Duke, MSU, and Zona being within the top 4 pre-season all in the same bracket with other slalwarts such as Gonzaga and Kentucky as lower seeds.

[Nugget]

I think the reason one might view it as a "bracket of death" is because many believe Oklahoma, Arizona, Kentucky and Gonzaga have a lot talent on their teams and have the POTENTIAL to beat top quality opponents -- but just haven't always consistently done so throughout the season. Basically, pre-season expectations are high for those teams, which are a decent indicator of NCAA success shockingly. That's why 538 uses pre-season rankings in their projections and why an 8-seed Kentucky who made the Final Four wasn't THAT huge of a shock/upset compared to what one would typically expect of an 8-seed. So, I'd agree that that would be a killer bracket with Duke, MSU, and Zona being within the top 4 pre-season all in the same bracket with other slalwarts such as Gonzaga and Kentucky as lower seeds.

Certainly one couldn't suggest anything odd about Mich St. being the #2 to our #1 under the circumstances. And, Oklahoma and Gonzaga do not strike me as objectively scary teams to have in your bracket.

The obvious outliers there to me are Arizona and Kentucky -- I'd want no part of either of them as #4/5 seeds in our region. It would be an open invitation for panic about a Sweet 16 flameout to a highly talented team that had underachieved in the regular season a la 2005, 2006 or 2011.

As a practical matter, this won't happen -- the Pac 12 is so weak that Arizona will win out (maybe they could lose 1 at ASU or Oregon) and probably end up around a 2 seed or 3 at worst.

[Kedsy]

I think the reason one might view it as a "bracket of death" is because many believe Oklahoma, Arizona, Kentucky and Gonzaga have a lot talent on their teams and have the POTENTIAL to beat top quality opponents -- but just haven't always consistently done so throughout the season. Basically, pre-season expectations are high for those teams, which are a decent indicator of NCAA success shockingly. That's why 538 uses pre-season rankings in their projections and why an 8-seed Kentucky who made the Final Four wasn't THAT huge of a shock/upset compared to what one would typically expect of an 8-seed. So, I'd agree that that would be a killer bracket with Duke, MSU, and Zona being within the top 4 pre-season all in the same bracket with other slalwarts such as Gonzaga and Kentucky as lower seeds.

I don't necessarily disagree with you on this. Putting four of the top five pre-season teams in one bracket (in addition to Duke, MSU, and AZ being #1, #2, and #3 in the pre-season AP poll, UK was #5) would be pretty wild. Personally, as a Duke fan, I wouldn't be so happy with this bracket.

However, it should also be noted that (a) Oklahoma was unranked in the pre-season, which could be seen as at least partially balancing the crazy pre-season expectations of the other four teams; (b) Duke would have to play at most one of Arizona/Kentucky and at most one of Michigan State/Oklahoma/Gonzaga -- if you removed, e.g., Arizona, Oklahoma, and Gonzaga from this bracket, would it still appear unusually tough?; and (c) I highly doubt the committee makes its decisions based on pre-season ranking. If Lunardi is attempting to recreate the committee's thinking (and he is), this bracket isn't particularly unreasonable.

[DavidBenAkiva]

I still think we're a month out from really looking at bracket projections with any reasonable sense of how they'll end up. Any of Duke, UVA, Purdue, and Villanova (least likely) could run into a stretch of 3-4 bad games.

Even if Duke wins on Saturday over UVA, there are still two games against UNC, a game at Clemson, at VA Tech, and a game at home against Louisville.

UVA has to play at Miami, Louisville and Florida State. They have a fairly easy schedule ahead of them, all things considered.

Purdue hosts Michigan tonight, Ohio State in a couple of weeks, and has to travel to Michigan State on February 10th.

Villanova has to go to Xavier, Providence and they host Butler (who has already defeated them once this year).

There's a fairly big gap between these top 4 teams today and the next group of schools. That being said, any of Michigan State, Kansas, Cincinnati, Oklahoma, or others could jump into a top seed if one of the current top teams falters. If the B1G teams all lose to each other, it's possible they could drop each other out of contention for a #1 seed. I could see a scenario where Kansas wins the Big 12 again, Duke beats UVA but Virginia wins the regular season ACC while MSU, Ohio State, and Purdue all get #2 seeds. That would be a very interesting and possibly controversial result. There are just too many moving parts to know what will happen from here.

[W&LHoo]

I still think we're a month out from really looking at bracket projections with any reasonable sense of how they'll end up. Any of Duke, UVA, Purdue, and Villanova (least likely) could run into a stretch of 3-4 bad games.

Even if Duke wins on Saturday over UVA, there are still two games against UNC, a game at Clemson, at VA Tech, and a game at home against Louisville.

UVA has to play at Miami, Louisville and Florida State. They have a fairly easy schedule ahead of them, all things considered.

Purdue hosts Michigan tonight, Ohio State in a couple of weeks, and has to travel to Michigan State on February 10th.

Villanova has to go to Xavier, Providence and they host Butler (who has already defeated them once this year).

There's a fairly big gap between these top 4 teams today and the next group of schools. That being said, any of Michigan State, Kansas, Cincinnati, Oklahoma, or others could jump into a top seed if one of the current top teams falters. If the B1G teams all lose to each other, it's possible they could drop each other out of contention for a #1 seed. I could see a scenario where Kansas wins the Big 12 again, Duke beats UVA but Virginia wins the regular season ACC while MSU, Ohio State, and Purdue all get #2 seeds. That would be a very interesting and possibly controversial result. There are just too many moving parts to know what will happen from here.

I think relatively few UVA fans are even looking at our schedule past this Saturday, which pretty dramatically alters the degree of difficulty of our road schedule. I honestly can't think of a place I'd like to see us play LESS than in Cameron.

[JasonEvans]

ESPN analytics tweeted the following this morning (along with a tweet giving Duke a 60% chance of winning the game with UVA, which sounds about right to me... maybe a little low):

Chance to earn a No. 1 seed in NCAA tournament, based on result of Saturday's game:
Virginia w/ win: 96%
Virginia w/ loss: 87%
Duke w/ win: 65%
Duke w/ loss: 38%

-Jason "the above percentages demonstrate that it is still really, really early to be talking about this stuff" Evans