PDA

View Full Version : Nonconference Scheduling from 2010 - Can this be right?



Mtn.Devil.91.92.01.10.15
07-29-2011, 02:34 PM
I was reading up on Jim Larranaga's new regime here (http://espn.go.com/mens-college-basketball/blog/_/name/katz_andy/id/6814106/jim-larranaga-relishes-challenge-miami) and found this gem buried near the end:

The ACC ranked 16th in nonconference scheduling last season. The Big East was No. 1. That was a major issue at the annual spring meetings. The ACC had four NCAA tournament teams. The Big East had 11.

16th? Can that possibly be true? Can anyone here name 15 other conferences without cheating?

We truly didn't deserve more than four teams if that's accurate. Very disappointing.

Anyone able to lend insight?

uh_no
07-29-2011, 02:43 PM
I was reading up on Jim Larranaga's new regime here (http://espn.go.com/mens-college-basketball/blog/_/name/katz_andy/id/6814106/jim-larranaga-relishes-challenge-miami) and found this gem buried near the end:

The ACC ranked 16th in nonconference scheduling last season. The Big East was No. 1. That was a major issue at the annual spring meetings. The ACC had four NCAA tournament teams. The Big East had 11.

16th? Can that possibly be true? Can anyone here name 15 other conferences without cheating?

We truly didn't deserve more than four teams if that's accurate. Very disappointing.

Anyone able to lend insight?

Part of the reason was certainly the fact that the rankings were turned upside down all year. Teams that we thought were going to be great like MSU faltered and teams that were supposed to be mediocre turned out to be great. Obviously we saw how this affected Duke, but I think it hurt the conference as a whole. Now, there are certain teams who seem to duck bigtime OOC games (cough cough VT cough), and I'm glad to see the ACC is trying to work on that (anytime I see ACC teams playing big games, its a good thing...unless they're getting the snot beat out of them....in which case maybe not). So I think some combination of lax scheduling and an outlier of a year in terms of rankings combined to put the ACC at the bottom. Something that helped the Big east, too, was their deep runs in several preseason tournaments. I think 11(?) of their teams finished first or second in their preseason tournament, meaning that each of those ended up with another game against a top (likely ranked) quality team.

Conference stuff is second fiddle though to how duke schedules and performs, and K schedules hard and wins, so you can't argue. We have a brutal OOC schedule this year with the champions tournament, the big10 challenge, and maui. If the rest of the ACC plays cupcakes for 2 months, it's unfortunate because it will lessen the value of playing those ACC teams later, but howeever THEY choose to schedule, we'll have played our share of games against top teams.

Bob Green
07-29-2011, 02:51 PM
Can anyone here name 15 other conferences without cheating?

Okay, I'm willing to try: 1) Big East 2) SEC 3) Big Ten 4) Big 12 5) C-USA 6) PAC 10 7) WAC 8) Mountain West 9) Ivy League 10) Big South 11) Sun Belt 12) Colonial Athletic Association 13) West Coast Athletic Conference 14) Missouri Valley Conference 15) Horizon 16) Ohio Valley Conference 17) Atlantic 10 18) Patriot 19) Big Sky

Okay, I named 19 conferences off the top of my head. That wasn't so hard. Now I'm off to ESPN to see if the ones I named actually exist. :cool:

pfrduke
07-29-2011, 04:39 PM
16th? Can that possibly be true? Can anyone here name 15 other conferences without cheating?


This is a sure sign of obsession, but here are the 31 conferences:

America East
Atlantic 10
Atlantic Coast
Atlantic Sun
Big East
Big Sky
Big South
Big Ten
Big Twelve
Big West
Colonial
Conference USA
Great West
Horizon
Ivy
Metro-Atlantic
Mid-American
Mid-Eastern
Missouri Valley
Mountain West
Northeast
Ohio Valley
Pacific 10/12
Patriot
Southeastern
Southern
Southland
Southwestern
Sun Belt
West Coast
Western

Reilly
07-29-2011, 04:47 PM
3 Atlantics
6 Bigs
2 Mids
4 Souths
2 Wests

Mtn.Devil.91.92.01.10.15
07-30-2011, 08:15 AM
Okay, I was being mostly facetious about not being able to name that many conferences.

But, seriously, is anyone else appalled and disturbed by the ACC ranking that low? I mean, I am willing to give some leeway for disappointing teams scheduled in advance, but I can think of ZERO excuse to be ranked 16th. Sure, I can see flipping around in the top four or five with the Big East, SEC, Big Twelve or Big Ten from year to year. Can the rest of the conference (i.e. "not Duke and UNC") really be scheduling THAT poorly?

I just find it incredibly embarrassing. From my (admittedly skewed) perspective, the ACC was the paragon of college basketball for the entirety of the 90's and most of the previous decade. Is this just the result of conference expansion and Seth Greenberg's annual quest to get enough wins to impress the committee?

CameronBornAndBred
07-30-2011, 09:17 AM
Is this just the result of conference expansion and Seth Greenberg's annual quest to get enough wins to impress the committee?
Well at least that annual quest is looking like it's no longer tradition. He knows being shut out of the tourney last year, despite beating us and having Dickie V personally guarantee his entrance to the dance, was because of his wimpy schedule. Hopefully their trip to the NIT woke up some other teams as well.


Seth Greenberg (http://www.dukebasketballreport.com/topic/sports/basketball/seth-greenberg-PESPT00008859.topic) has changed Virginia Tech basketball’s scheduling philosophy for the better.

Preparing for his ninth season as the Hokies (http://www.dukebasketballreport.com/topic/sports/college-sports/virginia-tech-hokies-ORSPT000225.topic)’ coach, Greenberg has constructed an ambitious non-conference road for Tech in 2011-12 that includes at least three games against teams that reached last season’s NCAA tournament (http://www.dukebasketballreport.com/topic/sports/basketball/ncaa-division-i-basketball-tournament-EVSPR0000083.topic).

http://www.dailypress.com/sports/teel-blog/dp-teel-time-greenberg-changes-hokies-scheduling-approach-for-the-better-20110706,0,2157288.story

sagegrouse
07-30-2011, 12:50 PM
These results are from the 2010-2011 regular season, where I looked at the out-of-conference games between all teams in the six BCS conferences. I did not weight the games by the supposed quality of the opposition. Nevertheless, the results show that the ACC teams played more games against other BCS teams than the other five BCS conferences.




Total Per Team
Conf Mbrs Games Wins Losses Games Wins Losses
ACC 12 50 24 26 4.2 2.0 2.2
Big 12 12 47 26 21 3.9 2.2 1.8
Big East 16 43 27 16 2.7 1.7 1.0
Big Ten 11 36 17 19 3.3 1.5 1.7
Pac 10 10 29 11 18 2.9 1.1 1.8
SEC 12 43 19 24 3.6 1.6 2.0

Totals 73 248 124 124 3.4 1.7 1.7



A couple of notes: These were results I put together for another purpose and have not been double-checked. Also, I would not criticize the Big East, which appears at the bottom in terms of total games, because I believe their teams play an 18-game schedule in conference. But I can't see why or how the ACC would be at the bottom of any list.

sagegrouse

Mtn.Devil.91.92.01.10.15
07-30-2011, 02:48 PM
I assume they were referring to strength of schedule, so I'm not sure how the OOC record would factor in...

sagegrouse
07-30-2011, 03:12 PM
I assume they were referring to strength of schedule, so I'm not sure how the OOC record would factor in...

Here's a quote from your original post:


The ACC ranked 16th in nonconference scheduling last season. The Big East was No. 1. That was a major issue at the annual spring meetings. The ACC had four NCAA tournament teams. The Big East had 11.

My table (because I already had the data for other purposes) addresses which conferences played how many teams from the other so-called power conferences. I included the W-L because is was already in my spreadsheet. Your point, I suppose, is that to be truly meaningful, one must have some power index to weight the individual games. That would certainly be helpful. (Now exactly which indices would you use for which point in time?)

But the referenced point was that the ACC ranked 16th among all conferences in strength of OOC schedules. 16th!?!? Holy cow!!! Big South? Southern Conference? The data I presented said, "Hmmmmm....... Then how is it that the ACC teams played more games against the other power conferences than the average team in any of those conferences?" So I titled my post "...the Charge Appears Bogus." I stand by it. I suppose one can construct a set of ratings for teams where the Southern Conference had a tougher OOC than the ACC. (Didn't most of those teams beat Wake last year?) Or, the ACC played only the patsies in the other power conferences (uhh, not possible since they all played each other). But I think it is highly, highly unlikely that the purported weakness of the ACC schedule was both accurate and reported accurately.

sagegrouse
'Your ever humble mathematician'
'I did learn once that STATISTICS are a property of the data set, not the result of misreporting by some misguided soul'

Mtn.Devil.91.92.01.10.15
07-30-2011, 03:26 PM
Ah, I misinterpreted what your data was intended to show. Never happened before! :)

I see your point that your numbers show that the ACC played lots of games against other top conferences. But, in order for this to show that the ACC didn't have such horrible SOS in comparison to other conferences, it would mean that it's a "better" game to play against a Depaul, Wake Forest, Indiana, Arizona St. or some other power conference cellar dweller na game against a Butler, Bucknell, or St. Mary's. Granted, this is the extreme example (I know that not all ACC games were against Indiana and that not all the other conferences played against Butler) but I think it's worth keeping in mind. Clearly, the difference between top conferences and the "lesser" conferences has diminished significantly in the last ten years.

I'm not a giant believe in strength of schedule anyway, and I guess the fact that we don't have a straightforward way to refute the claims that there were 15 conferences with better SOS than the ACC is illustrative of the vagueness. I just wish that the conference garnered a little more respect these days.

sagegrouse
07-30-2011, 04:33 PM
Ah, I misinterpreted what your data was intended to show. Never happened before! :)

I see your point that your numbers show that the ACC played lots of games against other top conferences. But, in order for this to show that the ACC didn't have such horrible SOS in comparison to other conferences, it would mean that it's a "better" game to play against a Depaul, Wake Forest, Indiana, Arizona St. or some other power conference cellar dweller na game against a Butler, Bucknell, or St. Mary's. Granted, this is the extreme example (I know that not all ACC games were against Indiana and that not all the other conferences played against Butler) but I think it's worth keeping in mind. Clearly, the difference between top conferences and the "lesser" conferences has diminished significantly in the last ten years.

I'm not a giant believe in strength of schedule anyway, and I guess the fact that we don't have a straightforward way to refute the claims that there were 15 conferences with better SOS than the ACC is illustrative of the vagueness. I just wish that the conference garnered a little more respect these days.

The ACC can't play all patsies in the other power conferences, unless only patsies play ACC teams, which is demonstrably untrue. Here is more data. We have the breakdown of the 50 power conference games played by the 12 ACC teams last season against the top four, the bottom four and the remainder (must be "the middle") of the other five power conferences:

Top four: 16
Bottom four: 17
Middle: 17

Looks to me that among the six power conferences, the ACC was above average in strength of OOC schedule: more games than any other conference (total and per school) and very well balanced across the strong and weak teams in the other conferences.

So, then how is the ACC 16th in out-of-conference scheduling, if it is demonstrably average or above against the five other strongest conferences?

I think Jim Larranaga is a great guy, but I wouldn't hire him to do my taxes.

sagegrouse

Mtn.Devil.91.92.01.10.15
07-30-2011, 05:53 PM
Additionally, looking on a handful of different online resources, the ACC had several teams listed in the top 150 of SOS, making it effectively impossible.

sagegrouse, I'm not in anyway disagreeing with your assessment. I'm just figuring that the article must have been based on SOMETHING.

Any ideas of what metric might validate the claim? Perhaps if most SOS take in conference games into account (which would clearly boost everyone's SOS playing against UNC and Duke) and they were using a method that ONLY measures out of conference games?

sagegrouse
07-30-2011, 06:53 PM
Additionally, looking on a handful of different online resources, the ACC had several teams listed in the top 150 of SOS, making it effectively impossible.

sagegrouse, I'm not in anyway disagreeing with your assessment. I'm just figuring that the article must have been based on SOMETHING.

Any ideas of what metric might validate the claim? Perhaps if most SOS take in conference games into account (which would clearly boost everyone's SOS playing against UNC and Duke) and they were using a method that ONLY measures out of conference games?

The quote is actually written by Andy Katz and is not directly attributed to Larranaga. Here's a passage from Katz's piece:


"Larranaga said the Hurricanes will have multiple seasons within one -- the nonconference, the early ACC without Johnson, the latter part of the ACC with him and the ACC tournament before possibly the postseason. To reach that postseason goal, the Hurricanes will have to make some noise in the nonconference schedule. The ACC ranked 16th in nonconference scheduling last season. The Big East was No. 1. That was a major issue at the annual spring meetings. The ACC had four NCAA tournament teams. The Big East had 11."

I dunno Katz's source or method. I think it's total BS, and I am trying to stamp it out before it becomes [gasp!!] conventional wisdom.

sagegrouse

SCMatt33
07-31-2011, 02:10 AM
I dunno Katz's source or method. I think it's total BS, and I am trying to stamp it out before it becomes [gasp!!] conventional wisdom.

I would guess that his source is ESPN's RPI calculations. An individual teams RPI is calculated using their record (adjusted for Home/Road/Neutral), their opponents record, and their opponents opponents record (neither of which are adjusted for location). I would assume that SOS is calculated from either the second component or the second and third component combined. I don't have ESPN's RPI numbers or SOS numbers, so I can't verify what Katz found. What we can use is Pomeroy's numbers. There is obviously some variation between Pomeroy's numbers and RPI numbers, and this varition is magnified because Pomeroy's numbers are post-tourney, and the RPI numbers are likely pre-tourney, but this should only significantly affect a few teams who made deep runs. Either way, it is unlikely that KenPom says that a team had a below DI average schedule while the RPI says that they have a top 100 schedule. Here are the team non-conference SOS ranks from the ACC according to KenPom:

UNC - 24
Duke - 64
FSU - 102
BC - 130
Clem - 180
VT - 189
Miami - 190
NC ST - 232
GT - 237
Md - 261
UVA - 291
Wake - 305

With 345 DI teams, the median SOS rank is 173. The mean SOS would be .500 which for Ken Pom occurs at 172. I don't know whether he adjusts to make the median so close to the mean or if it happens naturally given the large number of DI games. Either way, only 4 ACC teams ended with a Non-conference SOS in the upper half of DI and only Duke and UNC had a top 100 schedule, while Md, UVA, and Wake all had bottom 100 schedules. KenPom doesn't have any kind of conference SOS rankings, but looking at this data, it it entirely reasonable for Katz to have come up with 16th out of 31 for the ACC. In fact, if Ken Pom did have this kind of ranking, the ACC might have come out lower than 16th.

Even using advanced calculation methods, SOS numbers can be very decieving. Perception would tell us that a game against a top 10 team is much harder than one against a team ranked 50-60, a bubble team if you will. When you hear experts and fans alike debate schedules come selection time, you hear them complain about teams not playing enough ranked opponents, even if there are many top 50 and top 100 games on the schedule, while those who do play and beat several top 25 teams receive praise, even if there are few top 50 and top 100 games outside of those top 25 games. Games against "cupcakes" are then largely ignored. No one cares if you beat team 175 or team 340. If you're looking to get a bid or a higher seed, you should win both of those games easily so the difference is irrelevant. The computers though, see a huge difference between team 175 and team 340, while the difference between teams 15 and 50 smaller from the perspective of the computer. When we only look at "power conference" games, we ten to draw huge conclusions from who played the teams at the very top as opposed to those just below them, when the computers don't care that much.

The next question is how to fix such a problem. Again, people tend to think that the best solution is to start playing better top end games. For teams that really need it, it does help, but the real problem is at the bottom. I'm going to break down everybody's favorite example of bad scheduling, Va Tech.

First, we'll look at 2010, when their 307th ranked (kenpom) non coference SOS was largely blamed for their tourney snub. They played 5 guarantee games against terrible teams (all 297 or lower in KenPoms rankings): VMI, Charleston Southern, UMBC, Longwood, and NC Central. They played an away game at Campbell to open their arena. They started a "home and home" with UNCG (only giving them a game in 2011 as ACCT prep). They also had a home game against Brown as part of their preseason "tourney" which consisted of that and two scheduled games in Philly against Temple and Delaware. Of those teams, Temple was ranked 22 by Kenpom, and the UD, Brown, and UNCG were all 246 or lower. They drew Iowa (ranked 173) in the ACC/B10 challenge, a game in Mexico against Seton Hall, and home and home series games against Georgia and Penn State, the three of them ranking in the bottom quarter of the top 100.

In 2011, in which their Non Conf SOS wasa much improved, but still lackluster 189th , their home and home with Georgia was replaced by K-State (ranked 30th) and Penn State (37th) was much improved. Their preseason tourney was the 76 classic, in which they played only one bad team in Cal St. Northridge (230th), and two good teams in OkSt (74th) and UNLV (27th). They got a much better draw in the challenge with Purdue (9th) being over 160 spots better than their previous opponent. The replace the Mexico game against Seton Hall with a Bahamas game against Mississippi St. They played a semi-away game with St. Boneventure (152nd) for which they will get a home game next year. They returned the game with UNCG (268) for ACCT prep, and played 3 guarantee games plust the return from Campbell against teams ranging from 256th to 320th.

Analysing those two schedules, you see little change in the philosophy. The only noticible change was replace one gurantee game with a home and (semi)home series against the Bonnies and dropping one off the schedule altogether to leave them with 3 instead of 5. They still had 2 home and homes with BCS teams, a preseason tourney, the ACC/B10 game, a foreign game against a mediocre BCS team, the "home and home" with UNCG (which I imagine will either be continued given that the ACCT is staying put for awhile or turn into a guarantee). They played 13 non conference games, but since 3 were in a tourney, they should have been able to play 14. I don't know what happened, but I'm guessing that a guarantee game was cancelled on them and they couldn't replace it.

Here's the problem that Va Tech has. It's real easy for teams like Duke and UNC to put together high ranking schedules. We can get home and homes or neutral games with any top 10 program we want and can get guarantee games from teams in good conferences like C-USA, the A-10, and the CAA. Think about it. Duke can schedule a home/semihome series against a top end A-10 team like Temple, where Va Tech can get the same deal from a bottom end A-10 team like St. Bonaventure. Duke gets guarantee games from the likes of St. Louis, Tulsa, and UAB. Va Tech can't do any better than teams like Longwood (an independent), SC Upstate (A-Sun), and Mt. St. Mary's (NEC). That's not to say that Duke and other top teams should start scheduling worse to be benevolent or that Va Tech has to stop scheduling guarantee games with bad teams, because lets face it, ACC fans expect a certain number of home games and the athetic departments want to sell tickets to as many games as possible. In order to get better results, teams like Va Tech need to get better at scheduling, not try to schedule above themselves. What does that mean? They can't try to just put top teams on the schedule, as it's simply not going to happen. They have to almost be a progosticator and figure out who will be good for the next two years. They have to be able to pick out the teams like K-state who are their peers in other major conferences and will be going through an upswing when they get on the schedule. Coach K does a pretty good job at trying to pick out mid major teams who will be near the top of their leagues to play guarantee games with. A team like Va Tech needs to do this with low major conferences to get the team that's 190 in RPI instead of the one that 320. Instead of a home/semihome with a team at the bottom of the A-10, try and give a true home and home to a team near the top of the A-10 or CAA. I bet Richmond or VCU would play a home and home or a neutral site game with Va Tech and it would look a lot better than a game with the Bonnies. BC already does this by regularly playing local teams like UMass and Rhode Island. They don't get to play all of those game in Conte Forum, but it certainly helps their schedule and can help drum up local interest before conference season. With the rich get richer way of scheduling in preseason tourneys, conference challenges, and home and home series, where top programs will almost always play other top programs, teams like Va Tech will have a hard time regularly getting top notch schedules, but they can put themselves in a good spot to at least have a respectable schedule with just small improvements.

Mtn.Devil.91.92.01.10.15
08-01-2011, 07:29 AM
Think about it. Duke can schedule a home/semihome series against a top end A-10 team like Temple, where Va Tech can get the same deal from a bottom end A-10 team like St. Bonaventure. Duke gets guarantee games from the likes of St. Louis, Tulsa, and UAB. Va Tech can't do any better than teams like Longwood (an independent), SC Upstate (A-Sun), and Mt. St. Mary's (NEC).

First of all thanks very much to SCMatt for your analysis and thoughts. I think that you are mostly spot on with your post, but I wonder how much of the problem is that VaTech and other lower-tier ACC competitors "can't" do any better than these games and how much of it is that they simply don't want more L's on their ledger in March. I think the general perception is that these teams shy away from bigger games rather than being unable to reach agreements.

On the other hand, they probably are subject to the same philosophy from other schools. If you are a St. Louis or Tulsa, wouldn't you rather play against Duke? If you lose, no one is surprised. If you pull of the miracle win, it's a signature win for your school, might keep you in the NCAA tourney even if you lose your conference crown, and your players and fan base will remember it for all time. If you play VaTech all those things are diminished, but you have a better chance of pulling off the upset.

I still think this is another disappointing result from the expansion of the conference. In their championship 1982-1983 season, NC State's out of conference schedule included games against Michigan State, Louisville, West Virginia, Missouri, and Notre Dame. That same year Clemson went 10-20 and had opponents such as Texas A&M, Vanderbilt (twice), University of Washington, and South Carolina (twice) going 2-4. (Duke's program, on the cusp of a new era, went 11-17 and played a middle schedule with Colorado, California, and Louisville that year).

This year, Clemson played Seton Hall, South Carolina (once) and Michigan (Big Ten Challenge) going 1-2 on their way to a 22-12 record built against Long Beach State, Western Carolina, and Wofford.

There are fewer OOC games to fill regardless of who you want to schedule and teams are reluctant to schedule more challenging games when they can pad their resume with the Woffords of the world.

As I say, I'm just disappointed because it's definitely having it's affect on the general perception of the ACC. Basketball was our strong suit and a pillar of the conference. Football motivated the expansion and we are a now a middling football conference at best. Expansion doesn't appear to have harmed Duke and Carolina, but our conference "brand" seems to have suffered dramatically. I don't see the ACC getting more than four or five bids in the foreseeable future without a MAJOR change in philosophy.

johnb
08-01-2011, 12:52 PM
I

Here's the problem that Va Tech has. It's real easy for teams like Duke and UNC to put together high ranking schedules. We can get home and homes or neutral games with any top 10 program we want and can get guarantee games from teams in good conferences like C-USA, the A-10, and the CAA. Think about it. Duke can schedule a home/semihome series against a top end A-10 team like Temple, where Va Tech can get the same deal from a bottom end A-10 team like St. Bonaventure. Duke gets guarantee games from the likes of St. Louis, Tulsa, and UAB. Va Tech can't do any better than teams like Longwood (an independent), SC Upstate (A-Sun), and Mt. St. Mary's (NEC). That's not to say that Duke and other top teams should start scheduling worse to be benevolent or that Va Tech has to stop scheduling guarantee games with bad teams, because lets face it, ACC fans expect a certain number of home games and the athetic departments want to sell tickets to as many games as possible. In order to get better results, teams like Va Tech need to get better at scheduling, not try to schedule above themselves. What does that mean? They can't try to just put top teams on the schedule, as it's simply not going to happen. They have to almost be a progosticator and figure out who will be good for the next two years. They have to be able to pick out the teams like K-state who are their peers in other major conferences and will be going through an upswing when they get on the schedule. Coach K does a pretty good job at trying to pick out mid major teams who will be near the top of their leagues to play guarantee games with. A team like Va Tech needs to do this with low major conferences to get the team that's 190 in RPI instead of the one that 320. Instead of a home/semihome with a team at the bottom of the A-10, try and give a true home and home to a team near the top of the A-10 or CAA. I bet Richmond or VCU would play a home and home or a neutral site game with Va Tech and it would look a lot better than a game with the Bonnies. BC already does this by regularly playing local teams like UMass and Rhode Island. They don't get to play all of those game in Conte Forum, but it certainly helps their schedule and can help drum up local interest before conference season. With the rich get richer way of scheduling in preseason tourneys, conference challenges, and home and home series, where top programs will almost always play other top programs, teams like Va Tech will have a hard time regularly getting top notch schedules, but they can put themselves in a good spot to at least have a respectable schedule with just small improvements.

Thanks for the analysis.

I do have a question about it. Admittedly, Va Tech is unlikely to be able to get a home and away with UCLA or Kansas, but is it really true that they can't get a home and away with Cincinnati, Xavier, Louisville, Kentucky, Penn State, etc, which are places that have good teams and appear to be close by? It's fine to have a few patsies, but are they really the only team in a power conference that can't get a decent OOC sked?

SCMatt33
08-01-2011, 01:24 PM
Thanks for the analysis.

I do have a question about it. Admittedly, Va Tech is unlikely to be able to get a home and away with UCLA or Kansas, but is it really true that they can't get a home and away with Cincinnati, Xavier, Louisville, Kentucky, Penn State, etc, which are places that have good teams and appear to be close by? It's fine to have a few patsies, but are they really the only team in a power conference that can't get a decent OOC sked?

It depends largely on which team your talking about. They have a home an home with Penn State already (though I don't know whether or not its being continued). They haven't necessarily had trouble getting high major teams on their schedule, the problem is that it has mostly been the dregs, including Seton Hall, Mississippi St., Penn St (last year notwithstanding), etc. Getting a upper end team like Louisville or UK would be much tougher. Teams that are already successful with their non-conference scheduling have no reason to change it. This is part of the "rich get richer" theme. Louisville and UK are good enough year in and year out, that when they do schedule home and home series, they can do it with teams that they know will always be a quality game at the end of the year, not a "risk" like Va Tech, who won't always be a tourney team.

That being said, there are plenty of teams out there like Va Tech, who are good, but not always good, or high end mid majors. Cincy and Xavier are two good example. I'm thinking of teams like Vandy would also fit in there. You also have to be willing to travel far. A good example is when Wake did a home and home with Gonzaga recently. I think someone should try and put teams like Colorado and Utah on the schedule for a two game series. The problem is that it takes two to tango, and not all of these teams are looking to play a tough schedule *cough*Cincy*cough*. Last year, Va Tech scheduled 3 BCS teams, got another from a tourney, and another from the challenge, and played UNLV, a top non-BCS team in a tourney as well. Va Tech left a game off the schedule for some reason so that ended up being 6 out of 13 non-conference games. That's already a pretty good number. It wouldn't be too hard from a pure basketball standpoint to take a guarantee game and replace with a home and home/neutral site game against a peer in a major conference or a top end mid major, but they have their reasons for not doing it (I'm sure the extra home game is a big one).

Being in the mountains and having limited tradition hurts Va Tech's abilities to schedule. The teams who don't have that excuse are the ones like Maryland and Georgia Tech, who play in big cities, have tradition, and should be able to play better guarantee games than the ones they do. Maryland this year played NINE non-conference game against teams outside KenPom's top 200. There's no excuse for that.

pfrduke
08-01-2011, 02:09 PM
The teams who don't have that excuse are the ones like Maryland and Georgia Tech, who play in big cities, have tradition, and should be able to play better guarantee games than the ones they do. Maryland this year played NINE non-conference game against teams outside KenPom's top 200. There's no excuse for that.

This, as a whole, hurt the ACC last year more than usual. Lots of teams, rather than playing non-conference opponents ranked between 100-200, ended up playing teams that were sub-200. The league, as a whole, played 74 games against sub-200 teams (and another four against the teams that finished 194-197). There were 170 OOC games last year, meaning that about 45% of those games came against bottom feeders (embarrassingly enough, the conference lost about 10 of these games, with Wake and GT being the most egregious offenders). That's really poor scheduling, and something that doesn't depend on getting more marquee high-major opponents on the slate. In some instances, it completely negates the good opponents a team schedules - NC State played Georgetown and George Mason on a neutral site, went to Wisconsin and Syracuse, and hosted Arizona. Superficially, that looks like a great non-conference schedule. But 8 of their nine other non-conference games were against teams that ranked 203, 235, 247, 296, 309, 315, 319, and 322. As a result, NC State had, as a whole, a non-conference strength of schedule that ranked 232nd. Now, you can debate this method of measuring schedule strength - after all, I doubt there were 231 other schools that had 5 non-conference games as difficult as the 5 listed about for the Pack. But under the way SOS is calculated, there's no question that scheduling sub-200 opponents kills a team's non-conference SOS.

Here's the team-by-team breakdown of sub-200 opponents, in order of regular season finish. Note that with one exception (FSU), the top 6 teams scheduled fewer cupcakes than the bottom 6 teams - perhaps playing 9 games against the little sisters of the poor isn't necessarily the best way to prepare a team for ACC play.

UNC - 2
Duke - 3
FSU - 7 (and two more games against teams ranked 196 and 197)
Clemson - 5
VT - 6
BC - 3 (and two more games against teams ranked 194 and 195)
Maryland - 9
Virginia - 7
Miami - 6
Georgia Tech - 8 (they lost 3 of these, by the way)
NC State - 8
Wake Forest - 10 (they lost 4)

CDu
08-01-2011, 04:08 PM
These results are from the 2010-2011 regular season, where I looked at the out-of-conference games between all teams in the six BCS conferences. I did not weight the games by the supposed quality of the opposition. Nevertheless, the results show that the ACC teams played more games against other BCS teams than the other five BCS conferences.

A couple of notes: These were results I put together for another purpose and have not been double-checked. Also, I would not criticize the Big East, which appears at the bottom in terms of total games, because I believe their teams play an 18-game schedule in conference. But I can't see why or how the ACC would be at the bottom of any list.

sagegrouse

I took a look at the non-conference SOS from Pomeroy's rankings for 2010. Here is how the ACC teams rate:

Duke #13
Georgia Tech #62
North Carolina #77
Wake #89
NC State #154
FSU #187
Maryland #193
Clemson #197
UVa #218
Va Tech #307
BC #310
Miami #314

Considering that half of the teams are below the midpoint, I think that 16th ranking for the ACC seems reasonably possible.

There's obviously a fair amount of variability in any season's SOS, so I don't blame UNC, GT, or Wake. I might even give NC State a pass. But teams like Va Tech, BC, and Miami apparently scheduled some real dogs. FSU, Maryland, Clemson, and UVa didn't do the conference any favors, either.

sagegrouse
08-01-2011, 06:56 PM
It turns out that the conferences with the strongest strength of schedule are some of the weakest conferences in the land. Here is the ranking of conferences by out-of-conference strength of schedule, using Pomeroy's data for 2011:




Conf Ave SOS
SWAC 28.60
SC 97.67
BSth 106.40
ASun 122.18
MEAC 123.91
BW 137.11
P10 144.00
BSky 144.44
Horz 150.30
WCC 150.88
A10 162.64
MAC 164.42
Slnd 166.25
B10 167.27
MWC 170.78
CAA 175.42
SB 178.08
NEC 186.75
MAAC 186.90
BE 187.00
Sum 190.50
SEC 193.25
Ivy 193.25
AE 194.33
MVC 202.80
OVC 206.70
ACC 210.08
GWC 212.29
WAC 232.56
Pat 237.63
B12 246.92
CUSA 247.75




BY the way, the ACC is not even 16th. By this method it is 27th out of 32. The Big 12 is 31st. Among the power conferences, only the PAC-10 at #7 ranked in the top 10.

The correlation coefficient between the average Pomeroy ranking of a conference and the OOC SOS was -0.38, which means that the lower-ranked conferences play a tougher OOC schedule than the best conferences. One reason is that neighbors tend to play each other. Thus, when the ACC and the Southern Conference play a bunch of games against each other as neighbors, then it tends to produce what we see here.

I don't know what OOC SOS means for getting teams in the NCAA-T. Maybe there's a pattern; maybe there isn't. Here is the overall rating and the OOC SOS for the six power conferences:





Conf SOS Pom Rtg
P10 144.00 75.60
B10 167.27 36.09
BE 187.00 50.44
SEC 193.25 89.83
ACC 210.08 90.00
B12 246.92 61.00


Interestingly, last year the ACC was the weakest conference by average Pomeroy rating. The Big Ten was the strongest.

sagegrouse

Jderf
08-01-2011, 07:14 PM
It turns out that the conferences with the strongest strength of schedule are some of the weakest conferences in the land. Here is the ranking of conferences by out-of-conference strength of schedule, using Pomeroy's data for 2011:

[...]

The correlation coefficient between the average Pomeroy ranking of a conference and the OOC SOS was -0.38, which means that the lower-ranked conferences play a tougher OOC schedule than the best conferences. One reason is that neighbors tend to play each other. Thus, when the ACC and the Southern Conference play a bunch of games against each other as neighbors, then it tends to produce what we see here.

sagegrouse

Very interesting. Intuitively, this could make sense. In the aggregate, the strongest conferences should be able to beat down the teams they play against, thereby lowering their own SOS. The inverse would then be true for the weakest conferences. I'm not sure if I buy into that explanation, but the idea seems plausible.

Mtn.Devil.91.92.01.10.15
08-02-2011, 06:01 PM
It turns out that the conferences with the strongest strength of schedule are some of the weakest conferences in the land. Here is the ranking of conferences by out-of-conference strength of schedule, using Pomeroy's data for 2011:
....
Interestingly, last year the ACC was the weakest conference by average Pomeroy rating. The Big Ten was the strongest.

sagegrouse

Wow. Great numbers, sagegrouse. Who knew my hometown Big South (Go UNCA Bulldogs!) had such impressive schedules?

So, perhaps what we have learned is that aggregate strength of schedule doesn't matter too much where conferences are concerned? Clearly it's a useful tool for the NCAA in looking at team resumes for the tournament (who did those 19 wins come against, etc). Or maybe that's more of a "straight-faced test" as my friend calls it (can you pile up wins against A&Ts and directional schools and claim them with a straight face), and not so much raw data (ah, your SOS is 23.02, that's solid).

Whatever, I still blame Virginia Tech. Just because it's fun.