10/24/18 Update: This post has been featured/republished (and made a whole lot prettier) on AthleticDirectorU.com: check it out.
This is the abridged version of my proposal to improve non-conference scheduling in FBS college football. If you don’t think it’s too fucking stupid and also have too much time on your hands, check out the ~9000-word piece here.
Shoutout to SB Nation’s Bill Connelly for planting the seed this idea grew into.
The Problem
—
It’s no secret that college football has problems. Like, a–shit–ton–of–problems. But far down that list of issues (and paling in comparisons of gravity) is one problem that can be tweaked to increase the sport’s competitiveness, parity, and entertainment value:
Out-of-conference college football games are scheduled by the schools themselves, and not by a central body.
This leads to top-tier FBS schools scheduling weaker ones in order to host lucrative home games, get the six wins needed for bowl game eligibility, and (hopefully) not suffer a loss knocking them out of the College Football Playoff conversation early. Weaker programs, especially those in FCS, volunteer for these drubbings in return for a fat paycheck and the exposure that comes with playing a big-name school.
For fans, the result is a non-conference season largely made up of complete mismatches, punctuated by a few breadcrumbs of Top 25-30ish matchups (several of which take place at sterile neutral sites).
Many of these more high-profile out-of-conference games are also scheduled five-to-ten years in advance, making it a complete guessing game as to whether it will still be an evenly-matched contest by the time it actually arrives.
This system (or lackthereof) not only robs fans of seeing exciting non-conference matchups on a more consistent basis, but the wild variation in difficulty muddies the CFP playoff conversation into unresolvable “Team X ain’t played nobody” debates. Further complicating matters is although a majority of schools play four non-conference games, others play three or even five[note]The ‘Hawaii Exemption‘ allows teams playing Hawaii (and the Rainbow Warriors themselves) to schedule a fifth non-conference game that season[/note].
The Solution
—
Instead, why not have a central authority create out-of-conference schedules[note]As is done for virtually every other sport and also college football’s conference season[/note] as balanced as possible? And while you’re at it, why not make a spectacle out of the process the way the World Cup does?
Entrust non-conference scheduling to a draw system. This would produce randomized schedules all roughly the same degree of difficulty, just five-to-six months in advance of the season.
How It Would Work
At the conclusion of the season, use some set of rankings[note]Preferably one like Bill Connelly’s S&P+ Ratings that takes into account recruiting rankings, talent returning/leaving, adjustments for opponent, and per-play-based statistics. However, since we’re dividing FBS into four sizable groups of teams, I imagine all ranking systems would produce mostly similar results.[/note] to divide the 130[note]For simplicity’s sake, I am counting currently-transitioning Liberty as FBS[/note] FBS teams into three tiers of 28 and one of 46. Teams ranked 1-28 go into Tier 1, 29-56 in Tier 2, 57-84 in Tier 3, 85-130 in Tier 4. The previous season’s CFP playoff teams are seeded 1-4, but from there the teams are sorted in descending order purely by ranking.
Members of each tier will play randomly-assigned opponents from a designated tier during a designated week. For example, T1 teams will play a team from T4 in Week 1, from T2 in Week 2, T3 in Week 3, and go against a fellow T1 school in Week 4. Due to the number of teams, not all T4s will play an opponent from each tier, but they will all play one week against an FCS school in order to preserve the flow of money between the two subdivisions.
Schedules would be determined via an NFL Draft-meets-major soccer tournament draw-type event. Hosting duties could be given to a different school or conference each year. Hold it someplace warm on an early spring weekend and the event would be a pilgrimage for fans, have little TV competition from other sports, all while injecting new life into a usually-slower period in the college football news cycle.
For this to work, schools would all have to play the same number of non-conference games (four is preferable, though it could also work with three). There are many other caveats, hangups, and challenges that would arise from instituting such a system, but none I feel would be insurmountable. I list some of these hurdles below, but also go into greater detail in my longer write-up.
Results From One Sample Test, Compared To Actual 2018 Non-Conference Schedules
—
To test this idea, I used Bill Connelly’s S&P+ 2018 preseason rankings to divide the teams into the four tiers and then a random number generator to determine schedules. That draw (with fun tier-title sponsors) can be seen here or below:
• POPS Avg.: Perceived OpPonent Strength Average. Just a shorter way of saying the average S&P+ ranking of a team’s non-conference opponents.
• Real 2018 Avg: Average S&P+ opponent rank for teams’ real 2018 noncon schedules. NOTE: Most teams play four non-conference games, however those in three conferences play only three while three other teams play five due to the Hawaii Exemption.
• Diff.: Difference in difficulty (according to S&P+ rankings) in a team’s real 2018 non-conference schedule compared to the one my simulated draw produced: (S&P+ opponent average from real 2018 noncon schedule – the same average taken from simulated schedule draw). A negative number indicates the draw-simulated schedule would be easier than their actual 2018 noncon schedule.
To see if there’d actually be a big difference between the draw-produced schedules and the ones teams actually play, I took the average ranking of those randomly drawn opponents and compared them to teams’ actual 2018 non-conference schedules. I call this number POPS—Perceived OpPonent Strength[note]Really just an easier way to say average opponent S&P+ ranking[/note]. For FCS teams, I assigned them a POPS rating of 192[note]The median number of FCS teams is 62 (rounded down), which I added to the number of FBS teams (130)[/note].
For the actual 2018 college football season, FBS teams as a whole [note]Not factoring in the six Independent schools, who don’t have conference/non-conference seasons[/note] will play non-conference schedules with a POPS average of 93.56. From most difficult to least, the range varies from 39.50 (Northern Illinois) to 139.33 (Oregon), nearly a 100-point difference in difficulty.
In my simulation, schedules for Tiers 1-3 (I left out T4 since their schedule makeup is slightly different) increased in difficulty to a 58.79 POPS average, nearly 35 points higher in difficulty than the real 2018 schedules. More importantly, the range in most-difficult to least-difficult schedule was reduced to just 21.5 points, a 78% decrease!
Even in Tier 4, with their mandatory scheduling of FCS teams, the POPS average (99.62) was basically the same as the actual 2018 national average (95.05) and the difficulty difference was just 34.5[note]Unless there is some workaround I am missing, one school (Charlotte in my example) has to play two FCS opponents which misrepresents this figure[/note].
Beyond the Selection Saturday show, another marquee event would be born out of such a system: Week 4, which would have the same amount[note]Pending inevitable ranking fluctuations during the first four weeks of the season[/note] of Top 25 matchups (12-13) in one weekend as there would be in the first four weeks of a normal college football season. My sample draw produced a more conservative slate (only two top 10 matchups), but one that might still go down as the most loaded in history (assigned days are my vision):
Thursday
Utah (28) vs. Virginia Tech (21)
Oklahoma St. (19) vs. Mississippi St. (14)
Friday
Texas A&M (24) vs. Wisconsin (12)
TCU (22) vs. Florida St. (18)
Saturday
Stanford (20) vs. Michigan (10)
Ole Miss (25) vs. Notre Dame (8)
Penn St. (9) vs. Oklahoma (3)
Texas (27) vs. Miami FL (13)
Oregon (23) vs. Alabama (1)
USC (15) vs. Clemson (4)
UCF (17) vs. Georgia (2)
Auburn (7) vs. Ohio State (5)
Boise St. (26) vs. LSU (16)
Michigan St. (11) vs. Washington (6)
In the current system, the worst thing about scheduling a big OoC opponent is that an early-season loss could spoil your team’s CFP chances due to other teams, that didn’t schedule a marquee OoC game, running the table. But if everyone is playing somebody, what’s to worry about? And other than the networks and conferences, who would have an impossible time figuring where to place all these games, how could you not want a three-day stretch like that?
Pre-Mortem: Potential Hurdles and Roadblocks
—
Beyond just convincing the powers-that-be that change and experimentation won’t necessarily affect their bottom lines, I came up with a list of logistical hurdles that would emerge by installing a non-conference draw scheduling system:
- How to determine home/away while maintaining SOS balance and still give schools enough home games to make up their budgets
- Abandonment of non-conference rivalries
- How to fill the Independents’ schedules
- What is the incentive to climb from one tier to the next, other than “status”?
- Takes away the deliberate scheduling of games in certain markets for recruiting purposes
- Potential for a CFP contender to draw four G5 teams while a rival contender draws four P5s
- Could FCS and lower-tier FBS schools still make enough to cover their budgets?
- Unlucky and expensive draws, transportation-wise
- Due to T4s playing easier schedules and T1-3s facing more difficult ones, bowl games might be loaded with less ‘attractive’ teams
- Can schools be forced to opt into such a system? Who enforces it, conferences or the NCAA?
- How do you get basically every school to break the current cycle of 10-year in advance contract deals (which probably all have expensive buy-out/exit clauses associated with them)?
In the longer version of my pitch, I elaborate on these issues and offer some potential solutions (or for some, a cry for help).
Conclusion
—
Scheduling will never be perfectly balanced in a sport with a 12-game regular season. No matter how they are determined, you can only play the schedule in front of you, then pray it looks good on paper in December.
But accepting there has never and will never be a level playing field in college football doesn’t mean we shouldn’t look for things we can tweak to give the sport a little more parity (and make it even more exciting).
There are obvious plot holes here—some small; others more glaring (such as the Independent issue). But despite its reputation, I believe college football is more open to change than we give it credit for: in the past five years we’ve seen a new post-season system installed essentially because enough people complained about the old one[note]After figuring out how to to preserve the current money/power structure, of course[/note] and we’re currently seeing a 100+-year old facet of the game slowly die before our eyes. Change in the sport isn’t impossible, it just requires the right impetus.
College football games and seasons are largely influenced by chance and chaos: pointy balls take funny bounces and headcase 18-to-22-year-olds often do headcasey things in big moments. This is a big reason why we watch. But trying to put together a playoff-worthy schedule 5-10 years in advance (which is basically throwing darts while standing on a merry-go-round, blindfolded), is the wrong kind of madness. Why not correct this problem by entrusting it to a system worthy of such a spectacle-laden sport, that still produces the kind of controlled chaos we love, and most importantly, is infinitely fairer.[note]Well, at least on paper[/note].
If you have answers, fiery critique, or ideas on how this idea could be built upon, I’d love to deep dive with you; leave a comment or send me an email[note]One thing I’d be really interested in doing is taking the final S&P+ rankings after the 2018 season ends and retroactively applying them to my simulated draw to see how POPS matched up to reality[/note].
—