KenPom and the Dorks 2022-23

[sagegrouse]

T-Rank has Duke #5 and the Cheats #4. Zags #1, Baylor #2 and the Vols #3. Houston, Texas, Kentucky, UCLA and Kansas round out the top ten.


If the rest of the ACC is this bad, our Strength of Schedule will not be very good.

K-Pom does not have the 2023 season up yet.

GoDuke!

The ACC was hugely underrated last season, based on a couple of bad losses in November/December. We were chalked up as perhaps the weakest of the P6 conferences; yet the ACC had three in the Elite 8 and two of the Final Four.

Surely that mistake by the rankers won't be repeated. But, of course it may, as long as folks just grind the numbers and don't watch the teams.

[jv001]

The ACC was hugely underrated last season, based on a couple of bad losses in November/December. We were chalked up as perhaps the weakest of the P6 conferences; yet the ACC had three in the Elite 8 and two of the Final Four.

Surely that mistake by the rankers won't be repeated. But, of course it may, as long as folks just grind the numbers and don't watch the teams.

I agree with you. The ACC was underrated last season. I don't fully understand the dynamics in these rankings, but I do like to take a look at them once in a while.

GoDuke!

[Bluedog]

The ACC was hugely underrated last season, based on a couple of bad losses in November/December. We were chalked up as perhaps the weakest of the P6 conferences; yet the ACC had three in the Elite 8 and two of the Final Four.

Surely that mistake by the rankers won't be repeated. But, of course it may, as long as folks just grind the numbers and don't watch the teams.

I agree the ACC was underrated but many would argue a single elimination tournament result also isn't the best arbiter of conference strength. (Although not sure why November/December games are either, I guess sample size is a bit larger but teams have evolved from then.) It's obvious UNC was underseeded...blech!

We need to bring back later season out of conference games but I guess that ship has sailed with expanding conferences...

[JasonEvans]

T-Rank has Duke #5 and the Cheats #4. Zags #1, Baylor #2 and the Vols #3. Houston, Texas, Kentucky, UCLA and Kansas round out the top ten.

The rest of the ACC:

Virginia #17
Miami #20
VT #23
ND #1
Clemson #57
FSU #71
BC #75
CUSE #78
Wake #84
NC State #88
GT #97
L'SVILLE #98
PITT #123

If the rest of the ACC is this bad, our Strength of Schedule will not be very good.

Left out the 3 in Notre Dame.. they are #31.

As for the assertion that our SOS will suffer from the conference being weak, I dunno. These rankings make every single ACC game except contests with Pitt a top 100 contest. Last year, the ACC had 6 teams fall outside the top 100.

Every one of the top 9 teams would be a quad 1 game, whether at home or on the road. The top 5 would even be quad 1 games if you played them at home.

Based on these rankings, the ACC seems pretty similar to the Big Ten (ACC better at the top, B10 better in the middle, both conferences with a few teams in the end of the top 100 and one team just outside the top 100). Given how far the league has been from the top conferences in recent years, I'll take it!

-Jason "SEC seems quite good this year and the B12 will be a juggernaut... Torvik has every single B12 team in his top 50" Evans

[MChambers]

From what I saw, the ACC was not at all underrated last year. It wasn't a very good conference. Just because UNC got hot at the end of the year and got some lucky breaks in the tournament doesn't mean that the Tar Heels were that good. Miami got farther in the tournament than anyone expected, but then got absolutely destroyed by Kansas.

I hope the ACC is better this year and in the years to come, but I don't really know what has changed that will cause that to happen.

[uh_no]

From what I saw, the ACC was not at all underrated last year. It wasn't a very good conference. Just because UNC got hot at the end of the year and got some lucky breaks in the tournament doesn't mean that the Tar Heels were that good. Miami got farther in the tournament than anyone expected, but then got absolutely destroyed by Kansas.

I hope the ACC is better this year and in the years to come, but I don't really know what has changed that will cause that to happen.

I grant UNC getting hot, but the miami run is way overstated. they beat a mediocre USC team, and a mediocre gift matchup in the sweet 16 in #43 Iowa St. The win over auburn was a *great* win, but a great win doesn't an entire season make.

[Dr. Maturin]

PowerBI? Blech...j/k...

Torvik also is the major competition for Kenpom of course. I guess there must be money in this college basketball analytics stuff or people just like doing it.

I will say when I had a login/subscription issue on kenpom, Ken himself was the one that emailed me to fix it...I thought that was pretty cool. Not some huge operation.

Major plug for KenPom here. It's the easiest $20 I spend every year. He's never increased the price, put in ads, or done anything to make the site less user-friendly. Just gives you the good stuff, and nothing else.

[uh_no]

PowerBI? Blech...j/k...

Torvik also is the major competition for Kenpom of course. I guess there must be money in this college basketball analytics stuff or people just like doing it.

I will say when I had a login/subscription issue on kenpom, Ken himself was the one that emailed me to fix it...I thought that was pretty cool. Not some huge operation.

people like doing it, but there is also money in it. KP offers a service to help teams schedule OOC opponents.

But think of people like kedsy, and even me sometimes, that like to dump stats. I like to do analysis. heck, nate silver started 538 almost just for fun.

[CDu]

I grant UNC getting hot, but the miami run is way overstated. they beat a mediocre USC team, and a mediocre gift matchup in the sweet 16 in #43 Iowa St. The win over auburn was a *great* win, but a great win doesn't an entire season make.

The Auburn win also isn’t even as great a win as it looks on paper. Auburn was essentially playing without their star (and 1st round pick) C Kessler, who had a separated shoulder and played just sparingly in that game.

[House P]

The ACC was hugely underrated last season, based on a couple of bad losses in November/December. We were chalked up as perhaps the weakest of the P6 conferences; yet the ACC had three in the Elite 8 and two of the Final Four.

In fairness, it was more than a couple bad losses in Nov/Dec. It is hard to look at the ACC's performance in non-conference games prior to March and conclude that the conference was anything but subpar for a P6 conference.

Prior to March, ACC teams not named Duke were (as far as I can tell)

13-29 vs P6 teams (avg KP rank 57)
6-12 vs teams from the AAC, WCC, and A10 (avg KP rank 143)
7-4 vs teams from the Patriot League and America East (avg KP rank 219).

The ACC was an OT buzzer beater for FSU vs Boston University and a pair of come-from-behind-in-the-final-minute wins for NC State and Pitt vs Colgate from having a 4-7 record vs the Patriot League and America East!

In terms of results by KenPom ranking, non-Duke ACC teams were

3-21 vs non-conference teams ranked in KenPom's Top 45
9-31 vs non-conference teams ranked in KenPom's Top 75
23-41 vs non-conference teams ranked in KenPom's Top 125
18-6 vs non-conference teams ranked 126-200
52-5 vs non-conference teams ranked 201+

I don't fully understand the dynamics in these rankings, but I do like to take a look at them once in a while.

Other than Duke, only one ACC team had a winning pre-March non-conference record vs Top 125 teams. Here are the ACC non-conference standings vs Top 125 non-conference teams prior to March.

Team	W	L	Win Pct
Clemson	4	3	57%
Pitt	3	3	50%
Wake	1	1	50%
Miami	2	3	40%
NC State	2	3	40%
UNC	2	3	40%
Va Tech	2	3	40%
Louisville	2	4	33%
Syracuse	2	4	33%
UVA	1	2	33%
FSU	1	3	25%
N Dame	1	4	20%
BC	0	2	0%
Ga Tech	0	3	0%

I might as well continue beating this dead horse, so here is one final way to look at it. The three best Nov/Dec non-conference wins for non-Duke ACC teams were

ND's 4 point win vs Kentucky (KP#6)
UNC's 21 point win vs Michigan (KP #27),
UVA's 18 point win vs Providence (KP #27).

The next highest ranked non-conference wins for the ACC are

Syracuses's 2 point win vs Indiana (KP #48)
Louisville's 14 point win vs Mississippi St (KP #49)
Pitt's 2 point win vs St. Johns (KP #55)
Miami's 6 point win vs North Texas (KP #57)
UNC's 13 point win vs Furman (KP #74)
Pitt's 4 point win vs Towson (KP #75)

I think it is fairly safe to say that the ACC is way down when wins vs North Texas, Furman, and Towson are candidates for the conference's 4th best non-conference win (by teams other than Duke).

I agree the ACC was underrated but many would argue a single elimination tournament result also isn't the best arbiter of conference strength. (Although not sure why November/December games are either, I guess sample size is a bit larger but teams have evolved from then.)

The ACC had 157 Nov/Dec non-conference games and 23 March non-conference games (and eight of the league's teams had no March non-conference games). These 23 games are meaningful evidence of how well the top seven teams in the league ended up, but I am not sure how much extra weight to give to 23 games in March (from 7 teams) vs 157 games in Nov/Dec from all 15 teams.

We need to bring back later season out of conference games but I guess that ship has sailed with expanding conferences...

I agree with this 100%

[Mtn.Devil.91.92.01.10.15]

Yeah... I think the ACC wasn't underrated last year. I think the ACC was historically bad last year.

A few wins in March don't make up for the body of work.

[sagegrouse]

Yeah... I think the ACC wasn't underrated last year. I think the ACC was historically bad last year.

A few wins in March don't make up for the body of work.

How about 14? The ACC record was 14-5, better than any other P6 conference.

[uh_no]

How about 14? The ACC record was 14-5, better than any other P6 conference.

a bit bogus as the acc only got five teams in, so avoided more losses and likely fewer wins per team

[sagegrouse]

No one else won 14 games. The Big Ten got nine and they were 9-9.

[Mtn.Devil.91.92.01.10.15]

How about 14? The ACC record was 14-5, better than any other P6 conference.

Look. Our conference was bad by any measure until March.

Not mediocre. Bad.

You can argue that March wins means the ACC was good, but by that measure you'd have to acknowledge that UNC was better than Duke.

(I'm sort of joking)

[rsvman]

We overachieved in March. But Miami was a pesky team off and on all year that was capable of beating anybody on any given night. You may recall that they gave us fits, and I insist that we had a pretty good team. Not everything is told in a win-loss record.

[Mtn.Devil.91.92.01.10.15]

We overachieved in March. But Miami was a pesky team off and on all year that was capable of beating anybody on any given night. You may recall that they gave us fits, and I insist that we had a pretty good team. Not everything is told in a win-loss record.

No. But wins against other conferences and a lack of top 25 teams all season definitely paint a picture.

We had a few underrated teams and a few that came in strong in March. But no one would accuse the ACC of being anything other than bad last year, prior to March.

Now, we can have a conversation about where a run in March negates 3 months of terrible results.

[UrinalCake]

But how do we know that the ACC was historically bad when they don't play any OOC games after December? The common response is that they got hot in March, but isn't it equally possible that they were cold in November?

Consider this theoretical exercise: let's say on January 1 last season every ACC team got to draft a current NBA player. Duke got Giannis, UNC picked up LeBron, UVA got Steph, etc. Then they played the conference season. What would the conference's rating have looked like on Selection Sunday? All of those bad OOC losses in November would still be counting against them, and beating up on each other would have had no impact. So we would still be rated as the worst P5 conference (or whatever we were) but obviously every team got drastically better and should have been favored to make the final four.

That's a ridiculous example, but my point being that the computer polls are inherently flawed when conferences don't play each other after that early period.

[Mtn.Devil.91.92.01.10.15]

But how do we know that the ACC was historically bad when they don't play any OOC games after December? The common response is that they got hot in March, but isn't it equally possible that they were cold in November?

Consider this theoretical exercise: let's say on January 1 last season every ACC team got to draft a current NBA player. Duke got Giannis, UNC picked up LeBron, UVA got Steph, etc. Then they played the conference season. What would the conference's rating have looked like on Selection Sunday? All of those bad OOC losses in November would still be counting against them, and beating up on each other would have had no impact. So we would still be rated as the worst P5 conference (or whatever we were) but obviously every team got drastically better and should have been favored to make the final four.

That's a ridiculous example, but my point being that the computer polls are inherently flawed when conferences don't play each other after that early period.

Well, sure. I'll give you your hypothetical exercise. But I don't understand your bigger point...

So, I acknowledge that if each ACC teach were injected with one Lebron player after the holidays, it wouldn't show up on the comparative conference rankings. What makes you believe that our conference teams made these "Lebron-ish" leaps after the first of the year?

I see a re6bad ACC last year. Like, historically poor. Much like every metric that is out there would indicate.

The fact that several teams made admirable runs is testament to some good coaching, some gritty play, some luck. Which is all good stuff. But doesn't mean that our conference wasn't a hot mess.

[Skydog]

..
Given the huge movements in college basketball these days though, these prognostications are total guesswork but still fun.

Several years ago Pomeroy did an analysis comparing preseason rankings to end of season rankings as predictors of NCAA Tourney performance. So on one side we had rankings based on 0 games, just based on talent of incoming and returning players, program/coach assessments, years of experience, etc. On the other side we have a full season of play data for every team.

Tested over several seasons which was the best predictor of how far a team went in the NCAA’s?

Yep, you guessed it — preseason rankings!

That was an *amazing* finding imo. It argues that much of the season data is noise and that when crunch time comes talent, programs & coaches are what count most.

I would like to see the study replicated with recent year data, where there is more turnover. Preseason rankings might do even better since talent takes a front seat when you have less experienced teams? Ir it could go the other way - need to see the younguns
play 25+ games to really assess them.