Building a College Football Super League, Part I: Measuring the Value

I: INTRODUCTION

Earlier this past month, the Pac-12 effectively imploded as a result of its inability to secure a sufficiently attractive media rights deal in the aftermath of USC and UCLA’s departure to the Big Ten, with three schools departing to the Big 12 and two to the Big Ten in a single day. This does mitigate one of the problems with the USC and UCLA move by giving them West Coast partners to play and making them less isolated in that way, but it makes most of the others worse. There are still no schools in the Big Ten between Nebraska and Los Angeles, and we’re now looking at conferences 18, 20, or even more schools in size, with no real steps to make them feel like actual conferences and no real guarantee of a clear conference champion with every conference moving to go without divisions and simply send their top two teams to the conference championship game – only a scheduling and TV rights alliance promising the ability of teams to play the most valuable schools in the conference in football fairly regularly and to earn TV rights fees partly based off of the association with those schools. With an effectively national conference and no divisions, the Big Ten is making its Midwest and Eastern teams in non-revenue sports make the long trek west, or vice versa, all the more often, and to travel decent distances within the West.

It all had the effect of turning my thoughts back towards my wondering what might have been if college football’s biggest schools had decided to leave the auspices of the NCAA and formed a “super league” – whether a relatively modest secession of just the most valuable schools, or a full-fledged split of FBS itself taking most of the Power 5 with it as the Knight Commission proposed – and my desire to create a more robust means of determining what schools might join such a league than I engaged in at the start of the year. The dawning start of the college football season in earnest is as good a time as any to present my more rigorous findings. This post will document my attempt to find a formula to measure the viability of teams for a super league; the next post will attempt to actually identify the teams, and a future post may explore how this affects the conferences for the schools left behind and in other sports. 

II. GENERAL PRINCIPLES

• Think like a university president (or a businessman), not like a sports fan. Since Chicago-based blogger Frank the Tank introduced this mantra in his Big Ten Expansion Index post it has come to dominate the discourse around conference realignment. This asks fans to put aside such concerns as “wouldn’t it be cool if these two teams met regularly?” or “wouldn’t it be cool if this school joined this conference?” or “shouldn’t this school that’s winning a bunch of games this year be included in the super league?” and instead consider what factors the actual decisionmakers are using to weigh their decisions, which mostly means the extent to which any given school can make them more money. In past conference realignment decisions on the Power Five level, that has mostly meant adding schools that can bring more revenue to the conference than the share of the revenue the conference is already getting that they would take away from the existing schools. In a super league, all the most valuable schools would be joining at the start, so any new school can’t help but shrink the existing schools’ share of the pie. University presidents may be up for adding more schools than are strictly necessary, perhaps to take away any claim the remaining NCAA schools have to being on par with the super league and so raise the profile of the league itself, entice schools that may be reticent to join (that is, Notre Dame) to join and grow the pie further, mollify concerns of fans over breaking up rivalries, avoid legal action from the schools, or just because they want to associate with those schools, but all else being equal they’d prefer if the league was as small as possible. The only real alternative they might accept to the 24-32 team league I proposed at the start of the year is a full-fledged FBS secession of 48 (or more) teams.But then, they aren’t the only decision-makers here. If a super league is being formed it’s probably because a handful of billionaires are calling the shots and writing big checks to draw schools away from the existing system. That could mean they’re treating the league as a toy to play with and inviting schools for idiosyncratic, inscrutable reasons, such as bringing in their own alma mater, but it also means they have an interest in running the league like a business. As such, they might be more amenable to adding a school if it makes the league more money overall, or at least more money than it would cost. That may sound like much the same thing as the presidents’ concern – making more money than you have to pay them – but depending on the structure of the league it may not be that simple. For one thing, adding more schools doesn’t just mean adding more regular-season inventory to sell to distributors, it means adding more bowls that can punch above their weight in value and that cities can compete to stage, increasing the value of the league more than the value of the school itself might suggest.
• This is a football-only league. I set this rule in my previous post and it holds now; I’m assuming the schools in question would be able to remain in the NCAA for all other sports. That means only the value that schools bring in football will be considered here.
• The basis of television revenue in college sports is shifting, and the way it works is different for a national league than a regional one anyway. In previous rounds of realignment, the overarching factor in determining the impact of geography on the value of a school tended to be how they would help the conference’s network, especially in the Big Ten and, to a lesser extent, the Pac-12. Schools were picked largely on the basis of whether they could bring the Big Ten Network into new states and new markets, or could increase viewership of conference games in certain markets. That’s not a factor with a national league, especially one aiming to be the undisputed highest level of college football. To the extent geography matters at all, it matters only for determining how many fans a school has or might ultimately have.More to the point, with the cable bundle collapsing the source of revenue for a super league is shifting. No longer can schools and conferences count on affiliate fees collected from almost every household in a given area. Funding from national networks, especially national broadcast networks, remains valuable, but going forward filling the role of regional and to a lesser extent even national cable networks will be streaming services. As the collapse of the Pac-12 shows, university leaders aren’t quite ready to accept streaming services as the main distribution mechanism for the majority of their games, but nonetheless it still stands to become an important part of the equation. As mentioned elsewhere, streaming services care first and foremost about the overall value of a school, and especially how many fans each school has that they can entice to choose to sign up for a streaming service. That means there are likely fewer confounding factors in figuring out how much revenue a school may potentially bring; the value they have now, on their own, is pretty close to the value they’ll have in a super league.

III. FACTORS TO CONSIDER, AND NOT TO CONSIDER

These are largely based on Frank the Tank’s Big Ten and Big 12 Expansion Indices as well as Nate Silver’s own Big Ten expansion index effort at FiveThirtyEight (before he left that site earlier this year), with an emphasis on the Big 12 index. Body text (not headers) in bold describe actual components of the system, which will be brought together in section IV.

a) The One Thing That Matters Above All Else

• Money. As with my original attempt at forming a super league, the basis of all schools’ values will be the revenue of their football program over the course of the four years prior to the pandemic, as reported by the Department of Education’s Equity in Athletics Program data. The only FBS programs this doesn’t include are the service academies, which are Group of Five schools and are less likely to make the jump to a super league in any case unless a substantial number of G5 schools are already coming with them.

b) Things That Don’t Matter, Or Are Already Reflected in the Revenue

(I’m lumping these two together because there are things that fall in both categories to varying extents.)

• Long-term on-field success. The most valuable programs in college football have built a legacy of excellence over generations, not just in the last few years or within the tenure of a single coach. If a school that’s been irrelevant for decades suddenly becomes one of the best programs in the sport when a new coach (perhaps one whose name rhymes with Sabo Dwinney) takes over, what’s stopping them from lapsing back into irrelevance once that coach leaves? And don’t even get presidents started over schools that didn’t even enter FBS until the 90s or later and expect to be treated like royalty after a few good seasons (talking to you, Boise State and UCF… wait). On the other hand, one coach can singlehandedly change the culture of a program even after he leaves; while Nebraska has yet to reclaim the heights it reached under Tom Osborne, Penn State has remained at least close to college football’s elite even after Joe Paterno’s ignominious departure, and the same could be said, to a lesser extent, of Virginia Tech after Frank Beamer or Florida State after Bobby Bowden. The most storied programs tend to be at least on the periphery of bowl eligibility year in and year out and attract the best coaches through the value of their brand name, maintaining their quality through sheer momentum. But a legacy of greatness doesn’t mean much if the school isn’t profiting from it, and that’s not just a crass commercial consideration; if a school’s best days are behind it and its fanbase has dwindled to nothing, that’s a sign that the legacy and brand itself has faded. Generally, if a school has a strong, active football brand, that’ll reflect in the revenue it gets from fans, sponsorships, boosters, and TV contracts.
• TV ratings. While a game that gets more viewers means more advertising revenue, that’s historically been less important to cable networks than the ability to collect subscriber fees, and the difference is arguably more slanted on streaming services – and in any case, the effect of raw viewership on a school’s prospects is very indirect, as schools don’t get paid any more or less if they attract more viewers in a given season. More to the point, figuring out the contribution of a particular school to a particular game’s rating is a surprisingly tricky process, and certainly isn’t as simple as averaging out the viewership of every game involving the school. It’s not just that there are two teams playing in a game that are both contributing to its viewership (meaning the value of a school that plays a high-value school a lot can be inflated). There’s also the matter of the stakes of the game, whether the schools are in the CFP mix, are struggling to be bowl-eligible, or simply tanking down the stretch of a terrible season – not to mention that bowl games themselves, which are often significantly more watched than regular season games (even the lame ones that seem to be oversaturating the landscape), factor into the conversation as well, with some bowl games (especially the New Year’s Six, with the Rose Bowl and playoff semifinals standing out within that group) providing more value than others. A game going up against the Game of the Century is likely to be deflated as well. Games on broadcast networks or ESPN will be higher rated than games on ESPN2 or FS1, which in turn will be higher rated than games on ESPNU, FS2, BTN, or the like. Add it all up and there’s no reason to trust simplistic metrics like the “four million club” as a measure of a school’s inherent TV value. Whatever TV value schools are realizing is baked into their revenue in any case.
• Attendance. While attendance does reflect more directly on a school’s own fan base (though the success of the program and attractiveness of the opponent are also factors), the school also profits most directly from it.
• Non-football sports. As before, this is a football-only league, so anything having to do with basketball and other non-football sports aren’t under consideration here.
• Politics. I once might have put this in the category of “a factor, but not quantifiable”. Certainly a major factor in the collapse of Larry Scott’s “Pac-16” model was the fact it was a Pac-16; all Scott really wanted was Texas and Oklahoma, but he felt he had to grab all of the Big 12 schools in their respective states to get their legislatures to sign off on it, except Baylor, and political considerations weighed heavily on the Big 12’s consideration of Houston even before Texas left the conference, even though on paper the Big 12 didn’t need to shore up its Texas presence. Politics weighed heavily on which schools joined the Big 12 to begin with when the Southwest Conference was collapsing; if they had alums in sufficiently high places in the Texas legislature, Houston, Rice, and SMU could have ended up charter Big 12 members. It’s not just a Big 12 thing; though Virginia Tech didn’t join the ACC until the 2000s (in large part due to the influence of politicians), it and Virginia have generally been thought to be inseparable if anyone tries to raid the ACC, and the same for that conference’s North Carolina schools. (Silver calling North Carolina a “no-brainer” for Big Ten realignment, ahead of any non-Notre Dame school, overlooks that, at minimum, their storied basketball rival Duke, considered a “strategic reach” by Silver, would have to come with them, and probably also NC State which he puts in his bottom category. If adding UNC becomes another Pac-16 adventure, that drives the opportunity cost of a UNC expansion way up.) But then Texas and Oklahoma left for the SEC without bringing anyone else with them, and the regents at UC-Berkeley proved utterly powerless to do anything about UCLA’s defection to the Big Ten. If enough money is involved, politicians can do little but sit and sulk.
• Institutional fit. It is worth considering what left Baylor on the sidelines circa 2010, namely, its religious conservative viewpoints that didn’t sit well with the Pac-12 presidents running mostly liberal schools in liberal states; until recently BYU’s stance on LGBT issues and the like was similarly a turn-off to the Big 12 even with its more conservative fanbases and institutions. Bringing in schools with stances some people find morally repugnant could end up backfiring if it leads to a left-wing boycott, and some presidents might not want to associate with them either. Then again, the mere existence of a super league is likely to turn off some people, and the Big 12 just took on four Pac-12 schools (admittedly mostly in its more conservative states) to join a league with both BYU and Baylor in it. By necessity, there’s going to be strange bedfellows in this league; just keep the most egregious cases away from each other when subdividing the league, if possible. In a broader sense, “institutional fit” generally refers to the profile of the schools that make up each conference; the Power 5 conferences, except for (increasingly) the Big 12, are largely made up of public flagship universities in each state, with private schools, with a handful of exceptions, relegated to Group of Five status (especially in the American), and the Big Ten, ACC, and (until 2024) Pac-12 have generally placed a lot of emphasis on high academic credentials, more so than the Big 12 or SEC. But if Notre Dame decided to give up independence every conference would try to add them in a heartbeat, and every power conference has at least one private school in it. (The Big Ten has Northwestern, the ACC Duke and some others, the Pac-12 Stanford and USC, the SEC Vanderbilt, the Big 12 Baylor.) This may come into play more once I try to figure out conference alignments in non-football sports, but for now, again, strange bedfellows.
• Recruiting. A major factor in realignment in the past few decades has been trying to give each conference a presence in the states with the most potential recruits, especially California, Texas, and Florida. Indeed, this was a major factor in the Big Ten taking the otherwise-questionable additions of Maryland and Rutgers. But a coast-to-coast super league that starts with the biggest brand names and most valuable programs is probably going to have a presence in the best recruiting states anyway, with the likes of USC, Texas, Florida, Georgia, and Ohio State, and even in the states where it doesn’t have a direct presence, if it tries to lay claim to being the undisputed top tier of all of college football it’ll have its pick of the best recruits no matter where they go. What may still swing some recruiting decisions, though, could be the ability of recruits’ families to see their child/sibling in person. More on this later.
• TV markets and other geographic considerations – for the most part. As mentioned earlier, our super league has coast-to-coast ambitions and hopes to be the undisputed top tier of college football, such that even fans in places without super league teams can’t help but follow it, so trying to co-opt or muscle out the competition shouldn’t be a factor, and trying to colonize specific markets to spread a regional TV network isn’t part of the equation either. Since this is a football-only league, travel considerations are less of a burden (just ask the Big Ten), and we should be able to subdivide the league to keep it to a minimum. That said, there are some areas where geography matters; for example, if being part of a super league can help a program itself by being the only super league school for hundreds of miles around, in terms of attracting fans and recruits. Again, more on this later.

c) Things That Affect Revenue Such that They’re Factored In in Unexpected Ways

• Short-term football success. By most accounts, USC is by far the most valuable football program in the pre-2024 Pac-12, so what are they doing sitting behind Washington and Oregon for revenue? What are they doing sitting behind Utah, of all schools? The obvious answer is that Washington, Oregon, and Utah were all quite successful during this period, while USC was mostly in the wilderness. Short-term success can actually be a drawback to your realignment hopes; the true test of a program isn’t whether you’re currently successful on the field, but whether you can turn out fans week in and week out even when you stink. As such, deduct 2.5% times the average number of wins the school received over the course of the four-year period from each school’s revenue, to arrive at an estimate of wins-adjusted revenue. This does not entirely factor out on-field success – in particular, attending bowl games can result in more than your fair share of revenue accruing to your school – but it should better approach the inherent value of each school.
• Revenue from the conference. Besides the surprising Pac-12 schools ahead of USC, also ranking ahead of the Trojans are Arkansas and South Carolina – in fact they’re in the top 25 overall, ahead of schools like Miami and Oklahoma State. Indeed, the only schools outside the SEC and Big Ten ahead of them are Texas, Notre Dame, Oklahoma, Washington, Florida State, and Oregon. It’s hard to argue that Arkansas and South Carolina deserve to rank that high among the country’s biggest, most valuable schools; more likely, they’re benefitting from getting revenue from one of the two most valuable conferences in college sports, not to mention the single most successful one. Factoring out the contribution each school is getting from their conference is tricky, not least because it isn’t limited to TV revenue; specifically, revenue from bowls, including the College Football Playoff, accrues to the conference as well and is distributed to all members. Conversely, it should be obvious that the least wealthy team in a power conference doesn’t completely lack value of its own, so you can’t just take it as the baseline on which all other programs build. The key is to recognize that, just like in pro sports, revenue sharing to make the less valuable teams richer means the most valuable teams aren’t recognizing as much of their value as they should. The degree of revenue sharing, though, can vary from conference to conference. Thus, to adjust each school’s revenue for the value of the conference:
  • Take the standard deviation of the wins-adjusted revenue for all schools in the conference as a percentage of the average revenue.
  • Take the difference between each school’s wins-adjusted revenue and the average wins-adjusted revenue in the conference.
  • Increase that difference by 50% minus the percentage identified in the first step, then re-apply it. If the standard deviation is more than half the average (i.e. it’s the Big 12), make no adjustments.
  • To take an example, suppose State makes $100 million in a conference where average revenue is $75 million and the standard deviation is $6 million, 8% of the average. 42% of 25 million is 10.5 million, so conference-adjusted revenue is $110.5 million. A school with the same revenue, but where conference revenue is $60 million and the standard deviation is $9 million, 15% of the average, would see the difference hiked by 35%, and their total go up to $114 million.

Even these two factors together don’t really solve the problems they’re intended to, but this was about as aggressive as I was willing to get without introducing new counterintuitive outcomes. (For example, Ohio State was very successful during this period while Nebraska struggled, yet Nebraska’s revenue wasn’t that far behind Ohio State; as a result, even the adjustments used here give Nebraska a higher adjusted revenue than Ohio State, and I almost considered calculating the conference adjustment before the wins adjustment as a result, as the conference adjustment has the effect of spacing schools out more, but I decided that ultimately didn’t matter much given how I’m going to be starting the process.)

d) Something that Matters to Revenue But Isn’t Currently Reflected In It

• Demographic changes. The goal for any sports league looking to add or move teams, whether a professional league, a traditional college conference, or a college-sports super league, is to place teams in places that can cultivate a fanbase for years, even decades, to come, but the current distribution of population throughout the country isn’t static, nor is it expected to remain static. If you want to know why the Big Ten has been so aggressive at expanding into areas well outside its traditional Midwestern footprint, with Maryland and Rutgers a decade ago and raiding the Pac-12 now, look at the population projections: while the Midwest is expected to continue the “Rust Belt” process of economic and demographic stagnation, states throughout the South and West have been booming and are expected to continue to boom in population. So the Big Ten felt it had to expand into the places where, increasingly, the people (including its own former fans and graduates that flock to the South to retire) are, or at least into places that still have relatively thriving economies and large populations (the DC area and the outskirts of New York City), if it didn’t want to lose too much ground, especially to the SEC. Of course, this process has become a flashpoint in the more general area of politics, but we don’t need to get into that.(I will say that the parallel trend of the increasing urbanization of the US population is also one that college conference administrators have been factoring into their realignment decisions, and chasing both trends is itself something that likely chafes in the rural areas and mid-size cities that have historically been the backbone of college football fandom. The weekend after the Pac-12’s D-Day, it was briefly reported that the Big 12 was engaged in preliminary discussions with Oregon State and… San Diego State. If you ignore the tendency for “old money” to stick together in college sports and keep existing power conference teams within the power conference club, it makes sense based on the criteria that the decisionmakers look at: SDSU has been one of the most successful non-power conference schools in recent years in both football and basketball, and stands to be not only the dominant college in a sizable market in a growing state, but potentially one of the premier focal points for sports in general in the area. But besides the optics of snubbing a school that’s been associated with most of its Pac-12 brethren for over a century for one that didn’t even move up to Division I until 1969, Washington State, as a rural land-grant “X State” university, better fits the profile of what I think of from the mid- to lower tiers of the power conferences than a local urban campus of the larger state university system, and is more isolated from other FBS programs than either Oregon State or SDSU. As a city-slicker lefty I’m not normally sympathetic to “land area” arguments, but it’s easy to see how this might reinforce the sense of college football as another institution that’s abandoning its rural roots to chase urban audiences that might not be as immersed in the sport, and college football might want to do well to look at how that’s worked out for NASCAR, which has been rebuilding its presence in its Southern base in recent years.)

  The main point to make here is that the people making the decisions for the super league will want to know how big a fanbase a school might have 20 years from now, not just right now. While the size of the school’s fanbase is not the only factor in determining revenue even with the adjustments made in the last section, it should still be the primary one at this point. Therefore, the conference-adjusted per-win revenue is further adjusted by one-half of the projected rate of population growth or loss for its home state by 2040, based on numbers from the University of Virginia’s Demographic Research Group. If the population of a state is expected to grow by 35%, the adjusted per-win revenue is adjusted by 17.5%. I’m only adjusting the revenue by half the growth rate to reflect inertia from already having a big brand and the possibility for some schools, especially those in stagnant or declining states (as they often tend to export their graduates elsewhere), to have fans outside their home state (such as Big Ten fans in New York City).

e) Things That Matter Separately From Revenue

• Academics. This is the main category that university presidents care about beyond raw revenue. The degree to which they care varies; for the Big Ten, whose members form the Big Ten Academic Alliance (formerly the Committee on Institutional Cooperation), if you’re not a member of the Association of American Universities, you’d better have a truly sterling academic reputation (or at least be Notre Dame), whereas it’s more of a nice to have for the SEC and Big 12. Since the existing conferences will remain mostly intact in non-football sports, more academically-oriented schools don’t need to worry about mingling with glorified commuter and safety schools more than necessary, but they’ll still want to make sure super league schools have some academic credentials, if only to further the image that this is still an academic enterprise and not a glorified “NFL lite”. A school’s academic score is 350 minus their 2022 National Universities ranking by US News and World Report. Schools that are AAU members will either have their score fixed to 400 or have a 100-point bonus applied to their USNWR-derived score, whichever is lower. Finally, since most major-conference FBS schools are public flagship universities and the private schools tend to have sterling academic reputations, any private school will have a 25-point penalty applied, with an additional 15-point penalty for religious schools – but this penalty will be applied before the AAU bonus, so private schools can still score a perfect 400.
• Rivalries. The number-one complaint fans have about conference realignment is that it breaks up long-standing, bitter rivalries and replaces them with matchups that look cool on paper but have no history or bad blood behind them. Sure, having Texas and Michigan play regularly, or Alabama and Notre Dame, sounds like a juicy matchup, certainly to johnny-come-lately urban fans like me who only started following the sport in the BCS era and think about these sort of “wouldn’t it be cool” matchups, but the hardcore fans of the sport that sustained it for over a century before ESPN came along couldn’t care less about them, and they’d care far less about, say, Baylor and NC State. This is a big reason why a super league wouldn’t be a single 16-team conference; the perception that it was just a means to facilitate “kewl” matchups like Penn State-Georgia would be the kiss of death for its legitimacy, and even for a larger league it would be vital not to break up too many long-standing rivalries. The purpose of a super league, at least as perceived by fans, should be to nourish and support college football, even save it from itself, not further tear it apart in pursuit of the almighty dollar.What, then, defines a rivalry? It’s tempting to say it’s a game with an associated trophy, but to my knowledge Michigan-Ohio State, the undisputed king of college football rivalries, doesn’t have one. Similarly, proximity does not a rivalry make – you need an established history between two schools, you can’t just create that sort of bad blood from scratch – and in fact it’s entirely possible for rivals to be thousands of miles apart in very different parts of the country, as with Notre Dame’s longstanding rivalries with Stanford and USC. Perhaps the best way to define a rivalry, then, is to do it as Nate Silver does, as simply how many times two schools have played each other.

  But it doesn’t mean a whole lot that the University of Morgania has played West Morgania Tech every year since 1867 if the Flames have whooped the Techsters’ asses every single time. On a smaller scale, Oklahoma State may look forward to Bedlam with Oklahoma every year, but the Sooners don’t care nearly as much about it as they do the annual Red River Showdown with Texas, even though they’ve been played almost the same number of times. (For that matter, Oregon and Oregon State have played twelve more times than Oregon and Washington, but the latter is clearly the Ducks’ bigger rival.) So more than just sheer longevity, what also matters is how evenly matched the series has been over the years – how many times each school has made the other swear to get them back next year, how many times the trophy, if there is one, has changed hands. This is a similar challenge to how I assess games for the Sunday Night Football Flex Schedule Watch, where I consider how good a game is in terms of the records of the teams involved, but also how lopsided it is. There (and in my Games that Should Be Nationally Televised list earlier this summer), I generally rank games based on the record of the worse team until the last few weeks of the season where playoff implications become paramount (or in the case of the list, juicy storylines regardless of record). Similarly, the initial score given to a rivalry is the number of games won by the school that has won fewer games in the rivalry, with ties counted as one-half win.
  
  Finally, there is some reason to consider the recent history of a rivalry in this analysis; if a rivalry used to be competitive but now one school regularly kicks the other’s ass year after year, it’s easy to see how the heat of the rivalry might have dimmed. LSU and Tulane used to be each others’ fiercest rivals, but as Tulane de-emphasized athletics and eventually left the SEC in the 60s, the “Battle for the Rag” eventually became a yearly LSU beatdown of Tulane until the series stopped being played in 2009. Conversely, UCF and USF have only played 14 times with USF winning six, but the “War on I-4” has already become a fierce enough rivalry to be a genuine cause for concern when UCF joined the Big 12 while USF was never seriously in the running. But at the same time, you don’t want to penalize a rivalry too much for being broken up by conference realignment. Therefore, the score of a rivalry considering just the last 30 years (1993-2022) is added to the score based on the rivalry’s all-time history. In addition, each game in the 20 years prior to that (1973-1992) won by the school that has won fewer games since adds an additional point, with 1/20 (.05) deducted for each year prior to 1993 the game in question took place in.

  For example, say State has an 18-12 lead over Tech over the last 30 years. State won in 1991 and 1992 as well, taking a 20-12 lead, but Tech won the three years prior to that. State won in 1987, but that broke an eight-year Tech winning streak. The rivalry would, first, get a 12-point bonus for the 30-year results. The last time Tech won before 1993 was three years beforehand, so that would be worth 17/20 of a point, or .85. The other two Tech wins would add .8 and .75. State’s win wouldn’t add to the total since they’re already leading and are now up 21-15, but then you go back over Tech’s wins: .65, .6, .55, .5, .45, .4. Once you’re looking at the series from 1981 onwards, the schools are tied 21-21, so the next two Tech wins don’t add to the total, but the last State win before the winning streak does since they’d now be trailing, so it adds .25 to the total. At this point, you’re adding 5.8 to the bonus for a total bonus of 17.8, before looking to see if State won any of the four previous games. (You could have convinced me to triple each rivalry’s score for the last 20 years with the preceding 30 dropping off as before, since as it stands rivalry scores are still primarily influenced by the all-time record.)

  Now for the actual effect of rivalries on the ranking. Only each school’s four highest-ranking rivalries are considered, plus any rivalry that is in an existing league school’s top four. The four rivalries’ scores are added together, with each score being positive, adding to a school’s total score, if the rivalry is with an existing league school, and negative, subtracting from the total score, if it’s not. In other words, it’s beneficial to have a strong rivalry with a school that’s already been selected, because selecting you will keep the rivalry together. But if your strongest rivalry is with a school that hasn’t been selected, now that becomes a drawback because the league now feels obligated to take the rival.

  This means that the rivalry score has to be continually re-calculated after each school is added, meaning the rankings aren’t static and I can’t post a single list of the highest-ranking schools at the end of this post and be done with it. But it’s not the only dynamic component of the rankings…

f) Something That Matters in Ways That are Hard to Predict

Actual and potential fans and recruits. Dispensing with the bullet points because there’s only one thing in this category and, even though it’s only a small part of the calculation (especially for schools that are already blue bloods), it’s going to get even more complicated than the rivalry factor. We sort of covered this when we talked about demographic shifts, but this category is really about the potential size of the fanbase resulting from membership in the super league itself.

As mentioned before, a school’s existing following should be reflected in the revenue the school receives, so counting the size of the fanbase separately seems like it should be double-counting. But the following a school has now doesn’t necessarily correlate with the following it could have in the context of the super league. If two schools are in relatively close proximity but only one gets picked for the super league, the gap between the schools is going to widen. Obviously for one school to be picked for the super league while the other isn’t the gap has to be pretty wide already, but the idea is that, if the super league really does claim its place as the undisputed pinnacle of college football, only the absolute most devoted fans of the other school won’t at least follow the super league school on top of that. Conversely, it’s possible for one school to utterly dominate an area, even the area immediately surrounding another school, but if the other school gets picked for the super league suddenly they might have a shot at something approaching parity. Ohio State completely dominates the state of Ohio, even in Cincinnati, but the University of Cincinnati could start to claw out a space around itself now that it’s in the Big 12.

More important is the idea of untapped potential. Some schools may have a lot going for them that could allow them to punch above their current weight if they took full advantage of them – presence close to a major city, for example, or a large alumni base that could have their fandom awakened if the school was more competent or visible in football. Some schools may be relatively isolated from other super league schools, allowing them to claim fans for miles around, including potential fans of non-league schools, and if they’re located in a recruiting hotbed, that could make them punch above their weight in attracting recruits because of the ability of families to watch them close by. (If a school is the only super league school in its state, the effect of this is all the more compounded.) In short, unlike some analyses, I don’t think the question of where a school is located is entirely irrelevant in a streaming era, because it affects more than just the existing fanbase but also how many fans a school might have in the future.

Of course, this is something of a double-edged sword. The existing super league members aren’t particularly interested in providing welfare to underperforming schools because they might get better later. At some point, if a school has all the advantages, you start to wonder why they aren’t taking advantage of them already. And things that can benefit the school don’t necessarily benefit the league; a school can benefit from people who currently root for the state’s 800-pound gorilla switching to them, but from the league’s perspective existing fans are just moving their loyalties around. Many of these factors are mostly relevant for Group of Five schools that might be able to grow into Power 5-caliber programs, auditioning for conferences that don’t think they have any Power 5 brethren they can poach; schools already in Power 5 conferences, which are most of the ones we’re looking at, are presumably already tapped out. In the end, while this component does reflect market size, it mostly reflects the market size for the surrounding areas that currently root for schools that aren’t in the league.

This component consists of a number of different factors. Most of them are calculated in very simplified form because this is such a minor component and I don’t want to spend months on end on it, especially when I also intended to work on the Web 3.0 series this past month.

• Calculate each state’s percentage of the total FBS recruits per year as calculated by Football Study Hall, and take the average of that and the state’s share of the total population. Take the result and apply the rate of expected population growth or loss for the state by 2040. (Note that there’s some margin of error for states below Hawaii where FSH didn’t report specific recruit numbers, so I have to reconstruct them from the per-capita numbers.) This is an attempt to calculate the total share of football fans in each state, using the number of recruits as a proxy for the popularity of youth football. Of course, the popularity of playing football doesn’t necessarily correlate with the popularity of following football; there are fans of the NFL pretty much anywhere, including in places that might not have a big high school football tradition (though college football often appeals to different audiences than the NFL, audiences that might be more likely to invest their loyalty into high school football as well), while conversely, Florida punches above its already-considerable weight in terms of football recruits for other reasons than just the popularity of football there. Coupled with the potential for teams in the super league to attract new fans, this leads to factoring in the total population of each state for purposes of estimating the total potential fanbase.
• If the school’s state does not currently have a team in the super league, and the school is the highest-ranked public school in the state prior to this step, take the average of the estimated 2040 football fanbase of the state as calculated above, and the state’s share of FBS recruits pro-rated to 2040 population growth. Football fandom can change sharply across state lines, so it’s assumed that a team in the super league will take the lion’s share of the fans in its home state, at least if it’s the only school in the state. The public-school criterion is where I attempt to factor in the general preference of decision-makers for public flagship universities, but it’s not perfect. Clemson is ranked higher than South Carolina in the USNWR rankings, but South Carolina is generally considered the state’s flagship, so I needed to use something other than academic rankings to identify a flagship (not all states may explicitly identify a single flagship university). But in Idaho, the University of Idaho would probably be that state’s flagship and is higher ranked than Boise State, but by this criterion Boise would still get credit as the state’s flagship. But I think this is appropriate since this step is about which school the state’s population will rally around, all else being equal. (That being said, even this isn’t perfect, as we’ll see.)
• For each state that does have a team in the super league, allocate the state’s 2040 football fanbase to each county in proportion to their current population. Then determine the dominant football program for each county in the United States based on the New York Times’ 2014 map of college football fandom. (The NYT map is actually based on ZIP codes, but I’m not going through over 30,000 of them. Going county-by-county will be enough work as it is.) For each county where the dominant football program is not an existing super league school, take the ratio of the as-the-crow-flies distance from thr rough center of the county to the stadium of the school under consideration, to the distance from the center to the stadium of the nearest super league school. (By “rough center” I mean the “internal point” of each county in the data tables on the US Census Bureau’s TIGERweb site.) If the dominant program is not the school under consideration either, add the ratio of the score of the dominant program to that of the program under consideration. Cut the result in half, subtract from 50%, and if the result is still greater than zero, apply to the county’s allocated 2040 football fanbase. Note: Clemson and South Carolina are considered to be co-dominant in every county claimed by South Carolina alone in 2014, mirroring the adjustment made by Silver to the NYT map. Also note that “dominant” is a relative term; I’m basically taking the school I estimate to have a plurality of fans in each county, oftentimes fudging things to give schools the benefit of the doubt where the race is close.
• Allocate each state’s share of FBS recruits, adjusted to 2040 population growth, to each county according to present population. For each county where the dominant football program is not an existing super league school, take the ratio of the distances calculated above, cut it in half, subtract from 50%, then add one-quarter of the ratio of the distance to the school under consideration to that to the dominant school in the county (if different), and apply to the county’s allocated share of 2040 football recruits. Add to the result calculated above. This attempts to calculate how many recruits each school might bring in once it’s a member of the super league; the addition of the ratio of the distance to the dominant school attempts to figure out how many recruits the school might be able to steal from the dominant school.
• Add each school’s results for each state and county together to determine a single index of potential for that school.

IV. PUTTING IT ALL TOGETHER

Based on the Big 12 Expansion Index, and to a lesser extent Nate Silver’s calculations, the importance of each component to the calculations should be, approximately:

• Adjusted revenue: 80 percent
• Academics: 10 percent
• Rivalries: 6 percent (but a target of about one-twelfth of adjusted revenue is acceptable)
• Potential: 4 percent

As it happens, while only a handful of schools have adjusted revenue over $80 million a decent number have revenues in the $70 million range, so assigning a million dollars for each percentage point should be a decent benchmark. Thus, the independent components are added to adjusted revenue as follows:

• The value of academics in millions of dollars is the academic score divided by 40.
• The value of rivalries in millions of dollars is the total rivalry score divided by 40.
• The value of potential in millions of dollars is the index of potential times 100.

This spreadsheet shows the resulting permanent value of each school – their adjusted revenue and academic score, but not their score in rivalries or potential that depend on which schools are already in the super league, which I hope to have for you by Tuesday at the latest.