Frontlines of the Nondigital – The Debate Over College Football and the BCS
Preface: So what’s the biggest game design challenge in the world? It’s deciding who is champion in NCAA Division 1 College Football. We all know of this sport, but does everyone realize in the 60+ year modern history that there HAS NEVER BEEN A CHAMPION CROWNED?
So to make a long, complicated story/problem short: college football relies on opinion and mathematical formulas to dictate who plays who in the BCS (Bowl Championship Series… not to mistaken for a playoff). I repeat: there is NO playoff. After the bowls, the NCAA does NOT recognize a champion. The Associated Press does (sports writers) and they cumulatively vote for who they think is the best team and that team earns a faux championship. The mathematical equations that are used to determine the bowl matchups are transparent but completely convoluted so no one really pays attention. In fact each week when the new BCS computer rankings are released, it’s a big surprise to see who comes out on top. These rankings take “strength of schedule” (a broken formula that I will get into later), opinion polls, and about 5 other computer based polls into account to make the BCS Meta-rankings. And the funny thing: all college football fans agree the system is broken but it seems to never get fixed. I want to argue two things: one, that an intrinsic part of the sport of college football is the debate to who is best; and two, that a better non-playoff system can be molded by game designers like us to effectively declare a champion.
Below are a series of email debates sent back and forth between my friend Josh (an Ohio State fan) and myself (a Kentucky fan). I feel like this is an appropriate place to begin the conversation. Feel free to just read the bold for the real game design topics.
From Josh:
Thought you’d find this interesting, since you love to say OSU hasn’t played anyone yet. Note that our SOS right now is stronger than UK’s. Why? Mainly because the Big 10 isn’t doing nearly as badly as the media portrayed conference to be, and the SEC isn’t doing quite as well. Granted, UK’s SOS should be stronger by the end of the year after playing through the SEC, but I guess we’ll see. It’s only the end of the season that counts. Last year Ohio State was blown up because they beat the #2 team twice–Michigan and at Texas. But neither of those teams turned out to be that great, and Florida did. So to me the polls don’t mean jack right now. If we can get our quarterback’s head on straight, I would love to play LSU in a Big 10-SEC rematch. Right now, we have the easiest road there if we can win at Michigan. Will be exciting to see how things shape up.
My Response:
To really have this argument, you have to get your hands dirty with the math behind strength of schedule. By name, it sounds like a prefect way to break down each team’s true value by having a number to rank how good the competition they’ve faced has been. But the name is misleading.
http://www.nationalchamps.net/NCAA/BCS/strength_of_schedule_explain.htm
The formula is entirely based on opponent’s record, not ranking. Playing a 5-0 Kansas team would give you a great strength of schedule. Playing a 4-2 Florida team (arguably the 2nd or 3rd best team in the country) would not be as good.
Furthermore, since conference teams are beating up on conference teams, then every conference has a cumulative record of .500 (for a team to win, another has to lose). Make sense? And so the only variable for all teams to not have a .500 (or identical) strength of schedule are non-conference games (very few in the grand scheme) and slight variations in conf. schedules (Ohio State not playing Wisconsin in a given year). With so few variables, the entire formula entirely DEPENDS ON THESE SLIGHT VARIATIONS to weigh the overall strength, which again I remind you is just based on record, not rankings. With a PhD in psych, I’m sure you understand the problem with having a limited test pool (the non-conf games).
This is just another reason why the BCS formula is broken. Anyone can look at the Big Ten vs. the SEC and tell you which one is tougher to win in… sadly this formula can’t always figure it out. I would also add that all BCS conf teams should really just schedule bad non-conf games to try to remain undefeated, as that ultimately holds more weight than strength of schedule.
Josh Again:
Fair analysis, but you are the only one I know who actually believes that Florida is even arguably the 2nd or 3rd best team in the country. They lost at home to Auburn, who is not very good. They should have lost last week. Florida should not even be in this discussion. Florida lost most of its team from last year and has already lost two games. If you’re going to be great, you need to at least win the games you’re supposed to win. Losing at home to Auburn eliminated them from all competition.
This still follows your faulty thinking that the SEC is so superior that the best 2 or 3 teams in the conference are among the top 5 in the nation. I just don’t see it. We’ll see in bowl season how this shakes out. Last year Arkansas made it all the way to the SEC championship game, being one of the 3 best teams in the league. What did they do in the bowls? Lose to Wisconsin, who was probably the 3rd best team in the Big 10. Top to bottom, SEC is indeed the toughest conference. But the cream of the SEC is not necessarily better than the cream of any other conference. This is certainly looking like one of those years.
For an extreme example, look at Florida State through the 90’s. They played in the ACC which was complete garbage then (a little less so now). They never lost a conference game, and they were pretty good when it came to playing other teams in bowls. Not great, but they won it all some years and lost others. Point is just because they come from a weaker conference does not mean they didn’t belong in the championship picture. But what was wrong about them was that there were some years when they did indeed make it to the championship simply because they never lost (via an easy schedule) and were hyped in the polls by the media. And it showed. In ’98 and ’00 they lost to big underdogs (UT and OK) because they probably shouldn’t have been there.
So you’re right, the system is broken. But the system is also broken by simply ranking teams this early. You’ve been fooled into thinking Florida would be a good team because they were ranked so high and came off of a championship. But their play has quite clearly rejected that. USC belongs in the same category as Florida. Maybe those are both supremely talented teams, but they’re not playing like it. So they don’t get to be mentioned in the picture.
It’s always the same in college football. Over-hyped teams end up losing because the underdog wants it more. It happens every year. When was the last time the favorite won in the national championship game? I can’t even remember. I guess when USC played Oklahoma in ’04 was the last time, and the time before that had to be ’01 when Miami won. It doesn’t happen that much. Here’s what I can guarantee you, the hype has gotten to the head of the SEC teams, and they won’t win anything this year. Just wait.
My Response:
a) Auburn is actually a pretty good team. Probably better than UK and certainly top 20.
b) Based on games that I’ve seen, the best teams are: LSU, Florida, Oklahoma. The next tier would include Ohio St., USC, South Florida and all three may be as good if not better than the previous 3, but I have question marks for each because of cup cakes on each’s schedule.
c) The fact that we are having this argument shows how broken college football is. In fact, I would argue part of the allure of the game is actually the fact that it is the only sport where no champion is declared and all is left to debate and opinion (the AP number 1 is an opinion based). No math formula is going to mathematically fix that. Without a tourney we can argue all day long.
d) If all of the teams on UK’s schedule played all of the teams on OSU’s schedule, UK’s would win more. That’s the simple answer on comparing team schedules. This is why Pat Forde argued that UK should move to the Big Ten a few months ago: easier schedules in both football and basketball, and therefore easier to dominate.
Josh’s Final Response:
I agree, me writing that last e-mail made me realize how inane this argument is. The teams should play each other plain and simple. It makes no more sense that LSU might not get to be in the championship because they lose one of their many tough games than the fact the OSU might not get to play in the championship because people have deemed their schedule too weak and thus haven’t “proven enough,” whatever that means.
Last year’s championship game should have been the best argument yet for a playoff. When Florida only gets in because no one wants to see Michigan play OSU again (even though most sport writers thought they were the top 2 teams), it becomes a serious problem when Florida shows that they not only belong, they actually are better.
Honestly, right now I’d be afraid to play USF. They’re just really tough and play great as a team, even if they don’t have hot shot recruits. I believe Okiedokle is once again overrated and poorly coached, and I think they’ll lose again. I think Florida should be really good but they have some issues. They are one to watch later in the season when they get better.
While I would love to play LSU this year if we can go undefeated, I’ll be pulling for UK this weekend and might even tailgate for the game.
My Final Response:
On my subway ride I was still thinking about how college football is a “broken” game that we all love to death. In which case my game design teacher would remind me that if people enjoy it, then it’s not really broken. But the idea that the meta game of college football doesn’t crown an official champion is a really odd occurrence (or lack of one), especially considering how much time and money is on the line. Yet oddly, time and money are the two reasons waved in our face as to why a playoff isn’t feasible.
So with a playoff out of the question, can you think of a better system than the BCS? I think we can. It needs to be transparent and simple, a la the NASCAR cup chase. I think that’s a start…
To be continued, but ideas how to improve the BCS “formula” and comments to the situation in a game design context are encouraged!
The Nextel Cup chase format is an interesting idea. I guess a start would be that you’d be required to start scheduling major opponents from the outset. If you can only play 11 games, and the only teams you play are your conference and 3 creampuffs, how does anyone know who’s better? In 06 Ohio State played and won at Texas, who was #2 at the time, so they at least had a fair claim to being deserving of a spot in the championship. This year they don’t have that signature game, and it might hurt them.
I might like to see something like each Big 10, Big 12, and Pac 10 team all be forced to play teams from the other conferences twice each season, while the SEC, Big East, and ACC play each other the same. Then you’d have an East vs. West champion out of a point race in the final game. The non-conference games, which could be scheduled whenever (pre-season or during conference season) would then serve as some sort of barometer to measure a conference’s strength. Since teams make their own schedules, they’d be forced schedule 2 major non-conference games but could choose which ones.
Perhaps in order to make good teams play other good teams , subsets of conferences can be linked to each other (like divisions or subdivisions) and you’d be required to schedule each team at some point in a given interval. So for example, Ohio State might be linked to the Texas subdivision of the Big 12, featuring UT, A&M, TTech, and Baylor. We’d be forced to play all of those teams in a 6 year period, for example, so we don’t just go and schedule Baylor every year. I think if you left the schools up to their own devices, the powers would gravitate toward the cupcakes for financial reasons. Baylor would love to repeatedly play OSU because they would fill their stadium a lot more than if they had to play Purdue. OSU wouldn’t make any more ticket money off Texas than they would off Baylor, so why not play the crappy team as a warm up.
That’s just the beginning of an idea, don’t know where it would go from there. But my major thesis here is that if you’re going to keep an 11 game schedule, you have to make the season (1) more important and (2) strengths of schedules more clear. The latter could be achieved by making the conferences play each other in a systematic way.
I don’t think you can dictate any team’s schedule. There’s too much money involved with that aspect and therefore not feasible. I like your idea about conferences being forced to schedule one another (in a sort of national league vs. american league style a la baseball). I think this would make certain formulas more reliable. But I just don’t see such an idea being implemented if a playoff cannot be.
Here’s my idea: form a committee! seriously, pick 64 teams just like how they do in college basketball. Do this at the half way point mid-October. This committee will pick the 16 best teams, then second tier 16, then the third tier 16, and the final tier 16. For every team that Ohio St. beats in the top tier, Ohio St. gets 5 points. For teams in 2nd tier, 4 points… etc. For beating a team not in any of the four tiers, Ohio St. would just get 1 point. For each road win, Ohio St. gets a bonus point. Only the last 11 games Ohio St. plays in a given season are considered. After each conference crowns their champions, take the 2 champions with the highest point totals for the National Championship.
The beauty about this system is that it’s transparent, simple, and teams would understand how games matter… unlike with the confusing formulas the BCS currently has where a win might actually hurt your team.
I’m going to run last year’s season with this method and post the results.
I’m guessing that college Football is in this state because of some legacy from a time when people believed that college sports was about healthy athleticism and sportsmanship, rather than money and status? That said, isn’t the situation perpetuated by those very things? As long as a real champion is never declared, then different teams (albeit a small number) can continue to argue that they’re the best, or that if they only had a little more money from their schools they could more obviously prove that they’re the best. I guess the point that I’m trying to make is that this ‘problem’ seems designed into the system, and is now maintained for new, if different, reasons.
Maybe I’m completely mistaken. After all I don’t claim to know much about sports. It does make me wonder though, how did college Basketball avoid this mess?
I don’t think your idea is bad Charley, but you still have the problem of teams playing 80% of their games inside of their conference, and no real way of knowing how strong or weak the conference is. There are just so many teams out there that people who make these decisions too frequently are people who have never seen these teams play.
Basketball has a ton of overlap between conferences so we have a better idea of how these conferences stand up. The entire Big 10 plays the entire ACC every year. Great teams play each other a lot in the early season tourneys. Football teams do not decide their schedules except for the non-conference games. The conference decides their in-conference schedules. So why can’ the non-conference schedules be handed over to the NCAA to make up? Then you’d have a kind of NFL thing, where the whole schedule is pre-determined in order to achieve balance between the teams and continuity among schedules. I realize a lot of money is to be made, but so what. If you determine a team’s non-conference schedule 3 years ahead of time then it shouldn’t be an issue.
Here’s how you could set it up:
-2 games scheduled of any teams in the country (to protect rivalries, to let teams play in state opponents, smaller schools, whatever)
-2 games from major conferences determined by the NCAA. Big Ten must play one Pac-10 and one Big-12 school each year. SEC must play one ACC and one Big East school each year. There would then be a sample of like 20 games from each league (Big-Pac-10-12[West league] vs. ACC-SEC-BigEast[East league]) to better measure strenght of schedules. Then we could no longer say “The Big 10 sucks because they look slow.” We could back that up with results.
-2 games from your conference but not in your division. All conferences must split into divisions and play a championship game
-5 games from the teams in your conference divisions (each division has 6 teams, so you play everyone).
-1 conference championship game if you win your division
—-So in sum you only have an 11 or 12 game schedule—-
From there, if we keep the bowl system, the conference champion with the most points (based on wins and real strenght of schedule) plays the team with the second most points in the from the other league of conferences, and vice versa. So East 1 plays West 2, and vice versa. These teams play at the home of the top seed on like December 1, just like they might do in 1-AA playoffs. Or you could determine a “regional final” location just like in basketball. Then there would be a month off so plenty of time to schedule travel for fans to go bowls. The regional winners plays in the championship bowl, the losers get assigned to bowls with the other conference champions and other teams in bowls. So all you’ve really done is added one week, to be played right after the conference championships. And only 4 teams will play in those games. And you’ve added a much stronger mechanism to determine strenght of schedule.
Logistically college basketball can have a 64 team tournament over the course of 3 weeks. It would take football a month and a half to have that many rounds and the powers that be argue that’s too many games and too much time. The odd thing is that Division 2 and 3 actually HAVE TOURNAMENTS. So why those leagues of smaller schools can logistically do it and Division 1 cannot, no one knows. It seems to be tied into the money of sponsorships for bowl games… i.e. the Tostidos Fiesta Bowl.