# ACC Strength of Schedule

If you’re new around here, this will be the third year that we’ve examined the ACC regular season schedules to see who got the easier/harder regular-season schedules. In the aftermath of ACC expansion and unbalanced schedules, it’s always been obvious that the schedules will vary in difficulty and I was curious to see how much difference existed. So exactly how do we rate the strength of schedules?

Let’s start with what we don’t do (anymore). We don’t simply compile the won/loss records of each team and declare which teams had the hardest or easiest schedules. To be completely honest, I did compile the records the first year that I looked at this. However, since this technique has an obvious deficiency, we broke down the schedules further to take a closer look at the relative difficulty of the schedules.

The main deficiency with a simple compilation is that the teams at the bottom can’t play themselves (thus inflating their opponent’s records) and they lose a lot which further inflates their opponent’s records. Conversely, the teams at the top don’t play themselves which limits how many wins their opponents can compile. However, a technique to address this deficiency is already available in the RPI calculation.

We haven’t discussed the RPI much this year, but the RPI calculation consists of three parts:
- a team’s winning percentage
- their opponents winning percentage
- their opponents’ opponents’ winning percentage.

The RPI calculation removes the results between team “A” and team “B” when calculating the opponents’ winning percentage. This simple adjustment keeps the SOS calculation from being skewed in any direction by the team itself. Starting last year, we “stole” this philosophy and started calculating “adjusted wins” which removes the games that a given team lost when totaling their opponents’ wins.

For example, UNC won 14 conference games this year, with two of those wins coming against State. Those two losses are subtracted, leaving UNC with 12 wins, when totaling the wins for State’s opponents. Refer to last year’s entry for a more in-depth explanation.

I’ll quit boring our long-time readers now and proceed with this year’s results:

Anyone that knows their way around a spreadsheet is welcome to check my results and report any errors. The total wins calculation is nice in that there is some built-in, error checking ability. As far as I can tell, no such capability exists with the adjusted win calculations. In any event, I really don’t think that there are any mistakes here (have you ever known me to be bashful about my conclusions?). I would be more than happy to discuss any differences that anyone comes up with or any questions about the technique.

Now for a few observations:

- For the second, consecutive year, State ended up with the hardest schedule. Since State was scheduled to play Duke, UNC, and Clemson twice each this year, a lot of people have been expecting this result ever since the schedules were announced.

- Notice how UNC and Duke’s positions changed after calculating the adjusted wins. File this observation under “giving the devil his due”.

- Interestingly enough, the hardest and easiest schedules didn’t change after calculating the adjusted wins.

- Ignoring NCSU, WF, and VT for a moment, the remaining nine teams have remarkably even schedules. This cluster of nine teams is tighter than anything we’ve seen before.

Let’s look at a few issues associated with SOS:

Does the NCAAT Selection Committee Consider SOS?

Probably not directly. As discussed above, a team’s RPI calculation includes a SOS component and we have discussed before how the RPI affects NCAAT selections. But there are several problems with attempting to give SOS a higher weight in the selection process:

- How would you propose to give someone credit for playing tough teams, but losing?
- How can losing to good teams be considered a positive accomplishment?
- Do you expect the Selection Committee to project more wins into the record of a team that played a tough schedule?

Bottom Line: Who a team plays doesn’t necessarily tell you very much about how good a team is.

Who you beat is what tells you how good a team is. Referring once again to the Dance Card, wins again top-25 and top-50 teams as well as conference wins clearly correlate with getting an at-large bid to the NCAAT. So it appears obvious that the Selection Committee does look at SOS in that they consider wins against top teams when selecting the at-large bids.

A further example can be shown with the seeding that UVA received last year. UVa finished tied with UNC for the regular-season conference title (in part because they had the easiest schedule last year) and was seeded much higher than expected from their RPI. This episode strongly suggests that the Selection Committee seeded UVA based on their conference wins, without bothering to see who those wins came against. (Note that State received a similar, but significantly smaller, boost in seeding in 2004.)

The simple fact that this comes up over and over again proves that you just can’t fix stupid. About the only recurring topic that is sillier than suggesting a 22 game conference schedule is the assertion that State must leave the RBC Center and return to Reynolds.

Based on some reports from about this time last year, the ACC may go to an 18-game schedule with the next TV contract. In other words, the ACC will play more conference games if someone will pay for it.

Is FSU in danger of missing the NCAAT because their conference schedule was one of the strongest? Is VT on the verge of making the NCAAT because of their easy conference schedule?

This issue (actual or theoretical) is what generally drives the discussion of conference SOS. So let’s examine this from two different directions.

1) It’s not hard to evenly divide the ACC into three separate groups this year….the Upper 25% (UNC, Duke, Clemson), the Bottom 25% (State, UVA, BC), and the middle 50%. Since FSU and VT both fall into the middle 50%, it should be fairly easy to make some more detailed observations about the relative strength of their schedules.

If you look at the games these two teams played against the top 25% of the conference, we see that VT played the absolute minimum number of games against these three teams, while FSU played five games. Thus these two extra games against the top of the conference explain the two-game difference in FSU and VT’s records….doesn’t it?

Well, not exactly. Let’s take a closer look at VT and FSU to see where their wins and losses came from:

It would be impossible to argue that a significant difference in schedule difficulty doesn’t affect a team’s W/L record. However by taking a little closer look at the schedule breakdown, we see that FSU’s history of inconsistent play is evident again this year. FSU’s position on the bubble has a lot more to do with losses to USF (#168), Providence (#98), NCSU (#95), and two losses to WF(#87) than their four losses to UNC, Duke, and Clemson.

2) I was expecting VT to have a rough year…but they’ve played hard and won more games than nearly anyone outside of B’burg would have thought likely. However, their 0-6 record against the RPI Top-50 really puts their record into perspective. I think that even the most ardent Hokie fan would have to admit that they wouldn’t be nearly as close to making the NCAAT if they had played a significantly harder ACC schedule.

To be honest, I’ve delayed this entry several days trying to figure out what the final answer to this FSU/VT issue is. So far I’ve come up with the following:

- There is no practical way to even out the conference schedules.

- Whining about unbalanced conference schedules (or ACC expansion) is a waste of time.

- Projecting wins or losses based on a hypothetical schedule is worthless.

- Even though it is impossible to quantify, a significant difference in SOS will affect a team’s W/L record and thus will affect a bubble team’s standing come Selection Sunday.

So I have come up with a conclusion that is less than completely satisfying….blaming a trip to the NIT on your conference schedule is a lot like blaming a loss on the referees. Piss-poor referees can (and do) have an influence on a game…but they are not the only reason that your team lost.

HISTORICAL PERSPECTIVE

For our three-year summary, I decided to normalize the results against the toughest schedules seen during the 12-team ACC era.

Be Sociable, Share!

### 32 Responses to “ACC Strength of Schedule”

1. kyjelly
03/12/2008 at 12:06 PM #

Yes it is true no one did us a favor with having to play those 3 twice ,but it would not have mattered either,and we played a pretty much cup cake non league schedule which really prepared us for the league.

2. MattN
03/12/2008 at 12:18 PM #

1) No matter how you slice it, we woefully underperformed in every way possible this season.

2) 18 game schedule is better than 16, as long as we don’t drop two of our better OOC teams to do it. Drop teams like William&Mary and SC State, but keep the Cincinnatis, Seton Halls and Davidsons.

3) No matter how you slice it, we woefully underperformed in every way possible this season.

4) RPI of 95?!?!? Wow….just, wow.

3. Howler
03/12/2008 at 12:36 PM #

Great Info SFN. I am (almost) literally sitting here and applauding your in depth analysis. Maybe next year, the team will put as much time and effort into winning basketball games.

4. RAWFS
03/12/2008 at 1:01 PM #

^ I have to agree with that. The ECU loss was a huge flapping red flag and most intelligent fans knew it.

But hey, at least NC State contributed the “economic development of Eastern NC” by visiting Greenville. The “economic” phrase is the latest catchphrase that ECU uses when they whine in the legislature about making UNC and NCSU schedule them in football.

5. LRM
03/12/2008 at 1:10 PM #

Got to beat the teams you play, it’s that simple. That’s why the NCAAT is so pure.

We all want Duke and Carolina twice every year so eventually we have to start beating them regularly. Until then, it doesn’t really matter — Bottom 25% teams don’t play in Final Fours regardless of how tough their schedule is.

6. packbackr04
03/12/2008 at 1:13 PM #

im just glad our last 2 OOC games were against stiff competetion, and the week off prior to playing UNC in Chapel Hill helped too. that really helped us prepare to play the best team in the country. Wasnt it Will and Mary and presbyterian taht we played as “tune-ups” for UNC? who in the hell devised that plan? What a shitpot idea that was.

but again, as you said, it really doesnt matter b/c this team wouldve folded like a deck of cards at the first sign of trouble regardlees.

7. BJD95
03/12/2008 at 1:14 PM #

An 18-game slate would certainly flatten the “sine wave” distibution of schedule variance.

A quick look shows that the 6 major conferences are now split – 3 have 16-game conference slates (ACC, SEC, Big 12), while 3 have upped it to 18 (Big East, Big 10, Pac 10). The trend is definitely moving towards 18.

8. Trip
03/12/2008 at 1:22 PM #

Although, like stated above, we can’t argue about the schedule when we lost to ECU/NO and outside of 2 teams we’ve won our games by an average of what… 3 points?

Excellent analysis though, props given from myself.

9. Big Worm
03/12/2008 at 1:31 PM #

“When you lose to ECU, it’s hard to argue that the schedule was against you.”

By virtue of the RPI accounting for strength of schedule, and there being no magical win total that automatically gets you in to the Dance, schedules are neither “for” or “against” teams. Determining whether your win total warrants inclusion into the Dance is based on a sliding scale, of which strength of schedule is explicitly (RPI calculation) and implicitly (subjective judgment) a huge part.

With our strength of schedule this year, it is quite conceivable that we make the Dance having finished the season 7-9 in the ACC with an RPI in the very low 40′s (where it sat halfway through the ACC season). Halfway through the ACC, we had more wins against top-100 teams than all but around 7 or 8 other teams in the country.

For us, given our projected strength of schedule, 7-9 + 1 ACC tourney win (TOTAL: 8 ACC wins) likely gets us in. VT (RPI 58) likely requires 9-7 + 1 ACC tourney win (TOTAL: 10 ACC wins), and Maryland (8-8, RPI 70) likely would require two more wins (TOTAL: 10 ACC wins) to get in.

In short, teams earn their way in to the tourney. Even with an unbalanced ACC schedule, teams playing a difficult schedule do not necessarily have a harder road to the dance since the selection committee acknowledges strength of schedule and incorporates it into the primary metric (RPI) they use to rate teams.

10. Trip
03/12/2008 at 1:32 PM #

“who in the hell devised that plan? What a shitpot idea that was.”

I don’t know, but I bet that guy sure does know basketball! /snicker.

11. LRM
03/12/2008 at 1:34 PM #

Is anyone else worried that an 18-game schedule only raises the likelihood that we could finally lose 14 games?

12. GoldenChain
03/12/2008 at 1:46 PM #

As always, outstanding work.

You know, I’ve been hearing since the 1st of the season that unx had the #1 SOS in the country and that ours was very low. Couldn’t quite figure that out.

I don’t know if anyone saw the Ken Tysiac column yesterday where he rated the ACC teams this year, he “gave” State a ‘D’.
Personally I thought that was generous being that we were preseason voted to finish 3rd and ended up dead last!

13. choppack1
03/12/2008 at 1:50 PM #

I thought we went 2-14 once w/ Les Robinson?

I’m all for the 18 game ACC schedule. You play 11 teams, so that would be what – 7 teams twice (14 games), 4 teams once. That beats playing 5 teams twice, 6 teams once.

Once again, Swoff behind the curve – big shocker, huh?

14. Rochester
03/12/2008 at 1:51 PM #

we could finally lose 14 games

Been there. Done that. 2-14 in 1992-93 under Les Robinson.

By the way, nice win by Villanova today over Syracuse. That ought to prop up our out-of-conference schedule a little for Selection Sunday. Too bad Rider didn’t win their conference. That would have been a huge boost for us as the surprise at-large team of the year.

15. Texpack
03/12/2008 at 1:52 PM #

I like the idea of an 18 game conference schedule mainly because it will make the regular season results more reflective of the actual quality of the teams in the league. As for our OOC schedule, I would say that it was one of the better ones we’ve played in the last 10 years. I’m still shaking my head (just like Sidney) wondering how we could perform that badly. Hook ‘Em Horns! (A No. 1 seed in Houston will increase the market value of my sweet sixteen tickets.)

16. nycfan
03/12/2008 at 2:02 PM #

Very nice work.

17. nycfan
03/12/2008 at 2:13 PM #

FYI, I took your adjusted win # and multiplied it by each team’s conference winning % to translate the results into a very rough ACC RPI … it came out like this:

Team AOW Win% ACC RPI
UNC 117 0.875 102.375
Dook 117 0.813 95.063
Clemson 115 0.625 71.875
VPI 109 0.563 61.313
MIA 116 0.500 58.000
Umd 114 0.500 57.000
FSU 117 0.438 51.188
GaT 114 0.438 49.875
WFU 109 0.438 47.688
Uva 115 0.313 35.938
NCSU 123 0.250 30.750
BC 116 0.250 29.000

There is very little difference between the conference standings and the ACC RPI, except a shuffle at 7, 8 & 9 among FSU (who moves from 9 to 7), WFU (8 to 9) and GaT (7 to 8). So FSU gets a bit of a bump. Also, State and BC flop places.

18. Noah
03/12/2008 at 2:42 PM #

If Villanova doesn’t get in, will we have beaten anyone this year who made the tournament?

And State fans will be forgiven for not remembering the 1992-93 season. That entire season should be erased, “Eternal Sunshine of the Spotless Mind” style from our memories. Holy god, what a nightmare.

19. choppack1
03/12/2008 at 2:57 PM #

20. Sweet jumper
03/12/2008 at 3:02 PM #

I am a lifelong State fan and a Davidson graduate. At least I have one team to cheer for during March madness.

21. Noah
03/12/2008 at 3:28 PM #

Davidson. And I forgot Miami…they should be in the tournament. And maybe ‘nova. So three teams? Did I forget anyone?

22. SaccoV
03/12/2008 at 4:02 PM #

VaWolf82, how does this same data change when only 11 teams were in the mix? Don’t misconstrue the possible hint at a Herb-bashing curve to the thread. I was wondering how much the complete addition of Boston College has changed the overall picture of the conference SOS. Because in my estimation, the addition of BC has been a detriment to conference as a whole. The days of six and seven NCAA tourney bids could be a thing of the past for the ACC.

VaWolf82: I’m not sure what you are asking. There was only one year with 11 teams, but BC won the BE regular season that year (2005). BC made the NCAAT in 2006 and 2007 with double-digit ACC wins. I don’t see how you can conclude that adding BC has been a detriment to the conference.

In 2004, six of nine ACC teams made the NCAAT (most before expansion). But there are several years where only three received bids. I don’t really see how expansion can be construed to reducing the number of bids.

23. choppack1
03/12/2008 at 4:27 PM #

SaccoV – How do you figure – they’ve been to the NCAAs 2 out of 3 years (assuming they don’t win the ACC tourney this year.) In football, from a performance standpoint,they’ve carried their weight.

There’s one team that hasn’t done squat since expansion – and this year they will likely tie NCAA tourney bids w/ us.

NC State shouldn’t talking about anyone not pulling their weight. We f’ing suck in most stuff that matters.

24. VaWolf82
03/12/2008 at 4:44 PM #

With our strength of schedule this year, it is quite conceivable that we make the Dance having finished the season 7-9 in the ACC with an RPI in the very low 40’s

All available evidence would say that a team with a losing conference record would need several wins in their conference tournament to make it into the NCAAT.

Besides that, State’s OOC SOS has dropped steadily since December.

25. McPete
03/12/2008 at 7:21 PM #

Virginia Tech may make it in as well.