I'VE FINALLY DONE IT!
https://docs.google.com/spreadsheets/d/1gmwHfDwzEPuC4nWZPuVYdbP6pLxL4NT8...
This is a work I've put off repeatedly basically since conference expansion first began. Partially because of time constraints, partially because of losing interest, partially because "who would be team #16"? In addition, I've started and restarted it a number of times because of problems with team balance and how best to work "how do teams advance to the ACC Championship Game?"
Some thing I tried to achieve in this setup:
1. an NFL-style division setup where there are basically four smaller divisions within the conference, each containing four teams
2. each smaller division ("pods") must contain two "northern" teams and two "southern" teams
3. each pod must contain two teams that, either historically or in the BCS era and beyond, have been nationally relevant for competitive balance
4. pods are played in a rotating bases, similar NFL schedules, so you'll see every team in the conference at least once every three years and at least twice every six years.
5. some rivalries were maintained within the pods (UNC-UVA, Miami-Pitt) while others were ended solely because of a combination of competitive balance and other scheduling issues to make this work (VT-Miami, FSU-Clemson) with the hope of new rivalries being created.
6. each team received a permanent rival they will play every year. This is generally their most traditional rival within the conference. In addition, each team
7. each team also has a second rival that they will play twice in a three-year span, or four times in a six-year period. In years where you play your permanent rival's pod, you will play your second rival as well.
The second tab has a list of each of the four pods as well as each team's permanent rival and alternate rival.
Before I get into how the conference championship game is determined, I'll note that the 15th and 16th teams chosen by me were Notre Dame and West Virginia. This can be reworked for other teams, so if you don't like those two being picked? Go do your own damn work with different teams!!!!! ;)
CONFERENCE CHAMPIONSHIP SELECTION PROCESS
In this hypothetical world of 16-teams and 4 pods, the process is pretty simple. Each year, two pods play each other to form a single division. Essentially you'll have two 8-team divisions where half the membership changes each season. The team with the best conference record from those two divisions will then meet in the Championship game.
The tiebreaker process can be found on the first tab, modified slightly for the existence of the smaller pods as well. I've even tested it with a dreaded eight-team tie to come out with a single team without using outside metrics such as rankings!
Questions? Comments? See some errors I made? Feel free to poke away!
Some things I plan on adding later include individual tabs for all sixteen teams so you can see their home and away schedules each year and possibly the six-year schedule for each pod.
Comments
This is a lot to parse through with less than four hours to kickoff. #puttingitontomorrowstodolist
Thursday games, while fun, condense the pre-game activities a bit.
You've got your priorities in order. Good on ya.
Here's Bill Connelly's, who did this Pod setup for every conference two years ago.
My problems with Connelly's plan are:
1. he wants to scrap divisions entirely, which creates a nightmare for tie-breaking when you even fewer common opponents to try to break said tie with, at most 3-4 depending on 8 or 9 game schedule.
2. UNC is almost guaranteed to 4-0 with their protected rivalries most years. No other grouping is that "easy"
3. you're not getting Notre Dame without a good 16th team (revenue-wise) to give them reason to join
4. we've got Syracuse as one of our protected rivalries
The one thing I like about it is you're guaranteed to keep as many rivalries as possible going every year. If I didn't abhor the 9-game conference schedule because of its unbalanced nature, I'd try coming up with one that gives everyone three rivals you play every year. I may still when I have more time just to see what it looks like.
Yeah I actually don't love Connelly's pod idea, but I figured I would post it as it was relevant to your OP and looks at other conference possibilities if anyone was curious about other conferences.
Possible to split up 16 teams into four pods, having each pod rotate a full season playing one of the other pods. factor in one guaranteed tie into each other pod and you would have a 9 game conference schedule in what are two de facto divisions.
Over a six-year period a team would play all 15 other opponent in the ACC home and away, while playing there pod crossover rivals every year as well as all three teams in their respective pod. The change would be the remaining three depending on which pod you were playing as part of your de facto division.
Notre Dame can be brought in if we find a suitable existing rival who could make up that 16th slot. When I did this a while back I had Navy as the 16th teen as they already have an existing rivalry with Notre Dame. I would Factor Boston College and Miami as their other two major rivals.
I think Navy would have to come with Notre Dame, just because Notre Dame already has too many "permanent" rivalry games.
It's almost like they're a pain in the ass on purpose....
For you, and because it's dead at work for me today, I'll try to figure out a 9-game schedule *shudder* that replaces WVU with Navy though I'd have to figure out a new protected rival for us and Pitt with the loss of WVU.
Navy would be a good protected rival for VT if they were to join a 16-team ACC. We have a lot of alumni in their region and Tech's military ties would seem a natural fit to pair with a service academy.
I was thinking the same thing, actually. Just not sure who our third would be unless I kept it as Pitt because I'm not really in love with the idea of anyone else and (while manageable) I don't like the idea of:
Team A has teams B, C and D as their protected rivals
Team B has teams B, E and F as their protected rivals....
anywhere near as much as I like the idea of
Teams A, B, C and D are each other's protected rivals.
I think the pod method is the best approach to take. The biggest issue with the ACC is how tangled the web of rivalries are. The Carolinas all want to play each other but then UNC and Virginia will never give their games up. I can't envision a 16 team ACC that includes WVU where VT and Pitt aren't their rivals. We miss out on annual games with Miami and Georgia Tech.
I think I'm trying to poke that geography isn't the right split point for divisions. I think you try to put as much bad blood in a pod as possible.
I'm talking VT-UVA-WVU-Pitt. We would permanent cross over Miami and alt crossover GT.
ND-Miami-FSU-BC
UNC-Duke-WF-NC State
Syracuse-Louisville-Clemson-Georgia Tech
UNC-Duke-WF-State is an awful pod, much weaker than every other pod combination you listed. Currently, Clemson-GT-Louisville-Syracuse is right behind it because Clemson would just be handed three wins every year. Though I'm not sure that's different than the current state of things.
I was trying to find a way to get it so you played three rivals, each four times in a six-year stretch but I kept running into problems. The same was true when I extended to 9 conference games and just had teams playing their three protected rivalries every year, but I prefer 8. Ironically, those rivals were exactly the pods you just listed.
You could make the argument that the Coastal and Atlantic divisions are currently super unbalanced. Last year the Coastal had nobody (us, kinda) and the Atlantic was running with Clemson FSU and Louisville. When you are setting long term pods like this, things will change over 5, 10, and 20 year spans. Short term, yeah, Clemson is going to rock peoples shit. The NC pod is weak at the moment but they will get stronger because someone has to get good.
Can it? I mean look at the four teams in that pod historically and the last 20 years. UNC's had a few good years, same for NC State. But long term? Three of those schools care far more for basketball than football, NC State is always trying to play little brother to UNC and Duke when they should generally be better in football than they are, and can we be sure Duke will be able to sustain what success they've had once Cutcliffe is gone?
I wouldn't miss GT being an annual opponent. That didn't become a thing until we joined the ACC anyway, so it's not like there's a long seated rivalry. Honestly, the "rival" part of that matchup is probably only so apparent because both schools are "Tech".
Granted, you could say the same thing about us and Miami in the Big East days. At least in that rivalry, there were times where the VT-Miami game knocked the loser out of the national title conversation for the year.
Well this sure would make things interesting. So much hate. Oh yeah and FUCK PAT NARDUZZI
Pods would be interesting, and almost certainly a better approach than what we currently have, but I think I still like the Bitter Plan better. If you're unfamiliar, it's something that Andy Bitter once proposed (although I'm sure he wasn't the first, but that seems to be what most around here know it as), where you scrap the divisions altogether and simply assign three permanent rivals to each team. Each season, you get to play your three rivals and five of the remaining ten teams in an eight-game schedule (six of the remaining twelve in a 16-team conference, necessitating a 9-game conference slate). This way, every team has a home and away game with every other team in the league every four years , and many traditional rivalries can be maintained. You'll have more balanced scheduling and enable a better ACCCG matchup in those years when the two best teams happen to reside in the same division. As an added bonus, this method works with the current 14-team ACC as well as the seemingly-inevitable 16-team superconferences.
I've seen the Bitter plan and I dislike it because of one of the reasons I dislike Connolly's plan -- tie-breakers are a nightmare because you have, maybe, 3 opponents in-common with another team. This is especially true in a mulit-team tie where you'll have even fewer teams in common... and may not have even played each other.
At least with divisional and pod play you can go through 5-7 teams and say "Okay, you beat A B and C but lost to D and E. You beat A, B, C and E, and you beat A and B but lost to C, D and E."
Tiebreakers, especially of the three(or more)-way variety are almost always ugly. Your complaint about paucity of common opponents is going to apply to almost any structure you set up for a large conference and limited number of games on the schedule. Your proposal addresses that elegantly, but I disagree with the underlying assumption that this is the biggest issue that a new scheduling model needs to address. To me, not getting to play half of your conference-mates frequently enough, and failure to pit the two best teams against each other in the CCG are more pressing issues.
The Pac-12, Big Ten, SEC, Conference-USA, MAC, MWC, AAC, NFL, NBA, MLB and NHL would all like to say "hi" when it comes to not pitting the two best teams in their championship games. In fact, only the Big XII pits their "best teams" in the conference title game.
With my model, you play your entire conference every three years -- which is what the old 12-team ACC did in four (but twice in five because of how teams rotated). A player who is here all five years of eligibility will see all but four teams twice.
Scheduling needs to address:
Tie-breakers are very important to this and being able to resolve a three-way tie when you have six or seven teams in-common is a LOT easier than breaking a tie when you have at most three teams in common. Because what happens when three teams have the exact same conference record, don't play each other (possible under this model), and only have 2-3 teams in common for tie-breaking and have the same result against those teams (also possible)? I'd give you an example but that takes a lot more time than I have right now but you do see the problem here, yes?
As for "best teams in your championship game"
* of the last ten Super Bowls, only five have pitted 1 v 1. Three featured no 1-seeds at all and 8 of the 20 participants were the 2, 4 or 6-seed in their conference.
* if you pitted "best conference record" against each other and not just "best team in each division"
a. the SEC arguably failed to put its two best teams in the championship game five times in the last decade because
b. in its first seven conference title games, the Pac-12 has failed to place its two best teams three times, in 2011, 2012 and 2015. Ironically, all three of these games would have been Stanford vs Oregon had the "best" teams played.
c. likewise, in the first seven Big Ten conference title games, the top teams arguably failed to meet three times:
d. Finally, the ACC would have failed to place the best two teams four times in the last decade:
Even in the NBA, 1v1 has only happened once in the last decade (2016) and three NBA finals (2011, 2012 and 2018) didn't feature a #1 seed at all.
Not an issue if you can eliminate the need for divisions.
Yes, I can see the problem, I simply disagree with your perspective that it's more pressing than the problems of not playing every other team in the conference more often and failing to provide the best matchup in the CCG. There's room for reasonable people to disagree on these things.
The professional leagues are not comparable here, because they all have much larger playoff systems, and there's no way for the second-best team to get left out. Also, FBS conferences are more analogous to divisions in the pro leagues. Those leagues don't have or need division championship games/series. They all provide true round-robin play and reasonably balanced schedules for all teams in the division, so the team that finishes first in the standings can simply be declared the division champion.
As for the other D1 football conferences...wouldn't it be better if they did make sure that the two best teams squared off for the conference title? In your breakdown of the conference title games, you listed 15 times in 34 tries that the CCG failed to provide the "optimal" matchup. That's a 56% success rate. I think that illustrates my point that the divisional structure is doing more harm than good (in this particular metric, on whose importance we obviously disagree). The year when a 6-6 Georgia Tech advanced to the ACCCG (was it 2012?) is a perfect example.
And Georgia Tech only advanced that year because of ineligibility by both UNC and Miami.
And, no, divisions exist for a reason - because the individual conferences are too large to have round-robin play. Just like the professional leagues have divisions because they're too large for round-robin play.
If you really want to go down this rabbit hole, then just do away with conferences entirely and go to a swiss format. 64 teams, play 8 weeks where you play a team with the same amount of points after week 1 where teams earn 3 points for a win, 1 for a loss. Then after week 8 you eliminate all but the top-16 teams by points and play a single-elimination tournament. Heck, include all 129 teams. Play 10-11 weeks, cut to the top 32.
Conferences and divisions as a concept are fine; how the leagues have decided to handle their scheduling is not. And is it really that big of a deal to you that teams only play each other once every three years in my proposal? I mean I get not liking the current "once every six years" thing the SEC and ACC have going at the moment but once every three years is about where we were when the ACC was a 12-team league. In fact, VT played every Atlantic team twice from 2005-2011 and had already played NC State, Maryland and Wake a third time when the conference expanded to 14 teams so if you think "twice in six years" is too infrequent, then I feel you're being unreasonable since even in the NFL it takes 4 years to play each team in the league once in the regular season.
I like once every three years better than once every six. It's definitely an improvement. But every other year would be better yet. In the course of a 4-year college career, you could, in theory, experience every stadium in the entire conference.
I'd argue that I'm not being unreasonable at all in preferring to see all our conference-mates as often as possible. College football is not the NFL, nor should it be. And hell, in the NFL you get to play every team in your division twice a year. I liked it better when every conference could play a true round-robin schedule or something close to it, like when leagues had 10 members and each team only had to skip one opponent on a rotating basis every year. The familiarity, the rivalries and the traditions are what makes college football, in my opinion, more enjoyable than ShieldBall, despite being an inferior product in strictly football terms. By discarding those things that are unique and appealing, I fear that you only hasten the inevitable demise of the sport we all love.
And even in the NFL you play each team in your conference at least once every 3 years, and each team in the other conference exactly once every 4 years unless you play in the Super Bowl.
"Champion" is not the same as "best team".
The two biggest hurdles I've seen with using geography to redraw ACC divisions are having four schools in North Carolina, and working to avoid dumping the old Big East schools together.
If we do North/South, the North would literally be the old Big East (except Miami) plus UVA. We would need to pull one of the NC schools into the North to balance it out. I figure Wake is the easiest to split off. Then the South would basically be the old ACC plus Miami.
I did consider an East/West split, which sounds strange, but kind of fits on a map.
East: Boston College, Miami, NC State, UVA, Syracuse, Duke, UNC
West: Louisville, Georgia Tech, FSU, Clemson, VT, Pitt, Wake Forest
Problem there is that is splits up too many preferred rivals, and one of my goals when playing "what if" with divisions was to eliminate the permanent crossover.
I did this back in 2005 time frame when we moved to ACC, 4 teams in each pod that rotates and you would have 1 every year cross over and 1 4 out of 6 year cross over.
This was back with Maryland in conference which means the #1 team to add is wvu. I had all 5 power conferences done and the ACC was basically the BE and ACC combined.
Now I would still add WVU in a heartbeat, ND makes the most sense for #16. But they make a lot more sense in the big 10. Cincinatti, UCF, USF, Temple would all fit I the ACC geography. So pick one. I would like to make a deal with the SEC to give them GT for USCe.
A few years ago I came up with a system of "rotating divisions" that's sort of a twist of the Pod system. In it, each of the 4 pods were split into pairs, and separated into opposite divisions. Each year, you would play each team in your division and each team in your pod, for a total of 9 games. The pairs rotated each season in such a way that ensured that you played each team at least twice (home and away) over any given 4-year span. The only downside was that the divisions changed dramatically every season.
I can try to dig it up if anyone would be interested.
It looks like you probably over-thought your tie-breaker rules. Since everyone in the two-pod division had to play each other, the only tiebreaker you need in a two-way tie is head-to-head result, isn't it? Now that we've had OT in college football for the last 20 years or so, ties don't happen any more. I'd also suggest doing away with any reference to in-pod records or win percentages in your tiebreakers. That gives an advantage to a team in a weaker pod. Since you have a true round-robin within the "division", it would make sense to use that greater sample size in your tiebreaker.
Pretty sure the tiebreakers are essentially taken from the actual current ACC tiebreakers. Which makes sense, in the unlikely case you have a weather cancellation and cannot make the game up, such as for a game played on Thanksgiving weekend.
It's a modified version of the existing ACC tiebreakers along with how the NFL handles tiebreakers for division champions and wild card.
I honestly thought of making it "break the tie in the pod first until only one team remains in each pod, then break then break those ties" but opted to make the entire larger division count for more and then if a tie couldn't be resolved to break down to the smaller pods. This is an extreme example, but in the dreaded 8-way tie in a division that's Pods B and C:
Miami -- beats NC State, FSU, Wake and Notre Dame (4-4 overall, 3-4 division, 1-2 pod)
Louisville -- beats Miami, Pitt, Syracuse and WVU (4-4 overall, 4-3 division, 2-1 pod)
NC State -- beats Louisville, FSU, Wake and UNC (4-4 overall, 3-4 division, 1-2 pod)
Pitt -- beats Miami, NC State, Syracuse and WVU (4-4 overall, 4-3 division, 2-1 pod)
FSU -- beats Louisville, Pitt, Wake and BC (4-4 overall, 3-4 division, 1-2 pod)
Syracuse -- beats Miami, NC State, FSU and Wake (4-4 overall, 4-3 division, 2-1 pod)
Wake -- beats Louisville, Pitt, Syracuse and Duke (4-4 overall, 3-4 division, 1-2 pod)
WVU -- beats Miami, NC State, FSU and Wake (4-4 overall, 4-3 division, 2-1 pod)
1. Head-to-head, everyone is either 4-3 or 3-4. The 3-4s are eliminated (Miami, FSU, Wake, NC State)
2. Division winning percentage of the remaining teams is the same (4-3)
3. You now have to break the 3-4 ties first:
Miami -- beats NC State, FSU, Wake and Notre Dame (4-4 overall, 3-4 division, 1-2 pod)
Louisville -- beats Miami, Pitt, Syracuse and WVU (4-4 overall, 4-3 division, 2-1 pod)
NC State -- beats Louisville, FSU, Wake and UNC (4-4 overall, 3-4 division, 1-2 pod)
Pitt -- beats Miami, NC State, Syracuse and WVU (4-4 overall, 4-3 division, 2-1 pod)
FSU -- beats Louisville, Pitt, Wake and BC (4-4 overall, 3-4 division, 1-2 pod)
Syracuse -- beats Miami, NC State, FSU and WVU(4-4 overall, 4-3 division, 2-1 pod)
Wake -- beats Louisville, Pitt, Syracuse and Duke (4-4 overall, 3-4 division, 1-2 pod)
WVU -- beats Miami, NC State, FSU and Wake (4-4 overall, 4-3 division, 2-1 pod)
1. Head-to-head, everyone is either 4-3 or 3-4. The 3-4s are eliminated (Miami, FSU, Wake, NC State)
2. Division winning percentage of the remaining teams is the same (4-3)
3. You now have to break the 3-4 ties first:
-Miami, State, FSU, Wake
--H2H same
--Division same
--can't do this step because breaking ties is why you're here to begin with
--Revert to pods:
---Miami beat NC State. Miami 3rd in pod, NC State 4th in pod
----Both Louisville and Pitt beat Miami, Pitt beat State. Pitt 1st, Louisville 2nd
---FSU beat Wake. FSU 3rd in pod, Wake 4th in pod
----Both Syracuse and WVU beat FSU. Syracuse lost to WF. WVU 1st, Syracuse 2nd
The pod tie-breaker may not be necessary except in the event a game gets cancelled for whatever reason, but better to have it and not need it than need it and not have it.
With the other method, my thought was:
* H2H first
* Record within division
* Break down to pods, try to break tie until there's one team in each pod remaining
* If you can't break the pod tie before going to outside rankings or commissioner draw, then go through the entire division to try to break the tie first.
The main thing I like about your system is it basically shakes up the divisions every year. Keeps things more interesting. It's absolutely dumb that we've somehow managed to avoid Louisville up to this point. How can we even say we're in the same conference if we seemingly never play certain teams?
BTW, I know I've disagreed with some of the alternatives but I do appreciate all of the commentary everyone's given so far - especially those I've disagreed with because it at least lets me think about it more and ask myself "Is there a better way that meets minimal criteria?"!
I love this approach but one thing I'd suggest is why not put protected rivals in the pods? Like UVA VT, UNC Duke, Wake State, Miami FSU, GT Clemson, BC L'Ville, Cuse Pitt, ND WVU. That way you could protect two rivalries yearly and one every 3/4 years. For us: VT UVA (pod), VT Miami (yearly), VT WVU (2/3 years). For Carolina: Duke (pod), State (yearly), UVA (2/3 years). Etc etc.
If you look a little closer, I did do that in a way. UVA has annual games against UNC (same pod) and VT (annual cross-over rival) and a second rival in Pitt (2/3 years).
There are a few reasons why VT and Miami aren't in the same pod:
1. Originally, my pods looked like this:
ND - GT - UNC - UVA
BC - Clemson - NC State - WVU
Miami - Syracuse - Duke - VT
FSU - Louisville - Wake - Pitt
The reason for the change-up was due to difficulties making scheduling work when everyone's permanent protected rival (the cross-division one) is in a different pod from each other. In the end, it just became easier to put everyone's cross-division rivals in the same pod.
2. I wanted to keep some competitive balance within and between the pods.
Historically speaking, the top 8 teams in this version of the ACC are (in no particular order):
FSU, Miami, Notre Dame, WVU, Pitt, Syracuse, Clemson and Georgia Tech
But based on the past 20 years or so, one could argue that the top 8 are:
FSU, Miami, Notre Dame, WVU, VT, GT, Clemson and Louisville
In the end, I went with this version of "top 8" and put no more than two of these teams in any one pod. I then tried to protect as many rivalries as I could while trying not to create the super top-heavy rivalry of Notre Dame - Miami - FSU - Clemson while keeping UNC - Duke - NC State - Wake.
Here were the main rivalries I wanted to keep in-tact through pod play and permanent protected rivalries:
The Holy War
The Backyard Brawl
Catholics vs Convicts
Miami-FSU
Clemson-Georgia Tech (127 miles apart)
UNC-Duke
UNC-NC State
Duke-NC State
Duke-Wake
NC State-Wake
VT-UVA
VT-WVU
Some rivalries I wanted to keep and ultimately didn't because of both competitive balance and scheduling issues were:
VT - Miami
NC State - Clemson
FSU - Clemson
Pitt-Syracuse
Notre Dame-Pitt
In addition, some rivalries that I either:
a. didn't care about ending and was glad to see gone
b. didn't know were a thing and unintentionally ended or
c. didn't know were a thing and managed to keep are:
FSU-UVA (didn't know)
BC-VT (didn't care)
VT-GT (didn't care)
WVU-Syracuse (didn't know)
BC-Syracuse (didn't know)
BC - Clemson (didn't know)
3. I also had pods very similar to what you listed when I originally started this process a number of years ago. However, I didn't like the way some of them were more top-heavy than others and the general balance between the pods. Especially when, at the time, I was thinking of a way to make the winner of each pod have a semi-final match-up during "rivalry week" where the winners met in the ACCCG. I also felt balancing travel costs between the pods is still important and also tried to keep recruiting in mind by giving as many teams biennial trips to the northeast, the Carolinas, Virginia and Florida as possible.
I still like the idea of pods being geographically based, I just don't know of any good ways to do it and keep some semblance of competitive balance.