College Football Playoff: Should the Committee Pick the “Best” or the “Most Deserving” Teams?

While the College Football Playoff is an improvement over the old BCS system it, unsurprisingly, has not done away with controversy or debate regarding the identities of the top teams in the country. The College Football Playoff Selection Committee released their first ranking today, and the predictable result was a lot of back and forth on television and online regarding what the committee got right and what it got wrong. The problem is not the fact of the arguing — quite nakedly, engendering “debate” is the sole purpose of releasing ineffectual rankings rather than waiting to release the ones that matter — it’s that the arguments usually involve people talking past each other.

There’s nothing wrong with this opinion; indeed, it’s probably even in some circumstances correct: some people do think that a team being undefeated means it’s good enough to beat any team that isn’t. But that’s probably a minority view, and if someone is ranking the undefeated team over the one or two loss team it’s probably not because they necessarily think the undefeated team would beat the other team it’s that, being undefeated, they deserve being in the playoff more than a team with a loss.

Most people don’t always express their arguments in these terms — “best team” versus “most deserving” — but that’s essentially all it comes down to. This tension is why the playoff, despite being an improvement over the BCS, is not a panacea, either. I wrote about all of this for Grantland right after the end of the BCS era, and, two years into the College Footbal Playoff experiment, I continue to stand by every word:

The larger issue is figuring out how we should determine a sport’s “champion.” The wildly unpopular BCS was one method, while the new College Football Playoff will be another, but I’m referring to something more fundamental: What criteria should we use to determine who gets the title?

One answer is that the champion should be the season’s “best team,” possibly defined as the best overall team or the team we think would be favored to beat every other team on a neutral field. Another answer is the “most deserving team,” loosely defined as the team that produced the best overall season. These two things are not always the same. It’s perfectly possible for the best team — i.e., the most formidable — to lose a close game or even two on a bad kick or a fluke play, while another team runs the table by winning close games.

In theory, the now obsolete BCS was designed to create a championship game by blending these two approaches: The coaches’ poll would reward the teams that had put together the best seasons, while the computers would crunch numbers to objectively measure the strongest teams. In practice, however, the BCS was incoherent and flawed. If the computers spit out data the voters didn’t like, the computers were changed, and the coaches’ poll has long been riddled with inexplicable results.

And so, the BCS is dead. In the last few years, many fans and pundits allowed the word “playoff” to take on something of a talismanic quality. Replacing the BCS with a playoff system would surely cure the evils of the BCS, they thought, and quite possibly “save the sport” by “settling things on the field.”

Here’s the problem: A playoff does not even attempt to crown either the best or most deserving team. The very purpose of a playoff or tournament is the exact opposite: No matter a team’s talent or apparent destiny, everything can be undone on a single day by a single bounce of the ball. (Admittedly, that’s actually the allure of a playoff, hence why they call it March Madness.) Yet we’ve become so accustomed to playoffs that it’s difficult for us to think of any other way of selecting a champion. (Playoff-think is such a dominant paradigm that Neil Paine of FiveThirtyEight proposed mitigating some of the arbitrary tendencies of the NFL playoffs by giving points to teams that had better seasons than their opponents before the games even start.)

The primary advantage of a playoff is certainty, and after years of endless BCS debate — which followed decades of debate under the earlier bowl systems — certainty has real allure. But in most sports that have playoffs, like the NFL or the NBA, the criteria for getting to the playoffs is basically objective. Most playoff spots are decided based on win/loss records, with certain mechanical tiebreakers in place and known in advance. It’s not that the playoff crowns the best or most deserving team — just ask the 10-6 New York Giants that knocked off the undefeated New England Patriots in Super Bowl XLII. It’s that the loser has nothing to complain about: Everyone knows the rules.

Yet the new College Football Playoff lacks the very thing that makes playoffs in other sports so palatable, namely a semblance of objective certainty. While the defective BCS formula should have been interred long ago, it has been replaced by a Council of Platonic Guardians. The College Football Playoff selection committee will meet confidentially, then announce the identities of the playoff participants by edict. That’s not exactly what I’d call “settling it on the field.”

Most fans realize the new system is flawed, but figure it’ll be an improvement over the BCS since we’ll be talking about no. 4 versus no. 5 rather than no. 2 versus no. 3. Maybe so. But that logic works even better for no. 8 versus no. 9, and better still for no. 16 versus no. 17. And while an eight- or 16-team college football tournament sounds genuinely amazing, it’s naive to think that wouldn’t have a real effect on how the regular season actually works. It also makes me wonder what happened to trying to crown the best or most deserving team as champion, rather than the team that happened to win a single-elimination tournament.

I’m not at all sad to see the BCS go, but I’m not sure any playoff, let alone this particular playoff, will solve much in a world of conference realignment and more than 100 FBS teams scattered across the country.

  • Corey

    I’m not convinced that we need an 8 or 16 team playoff. In trying to identify the “best” team (according to whatever criteria), there is frequently significant uncertainty that Teams #1 and #2 are actually any better than team #3, or in some years even team #4. Including four teams is necessary to reduce the chance that a team which has a “legitimate” claim to be the “best” team (e.g. 2000 MIami and Washington, 2003 USC, 2004 Auburn, 2008 Texas, 2010 TCU, 2011 Ok St.) is not left out. The farther down the rankings you go, the less compelling that justification becomes. In many years there is little discernible difference between #1/#2 and #3, but that’s less true with #5, and really less true with #9. The likelihood that the actual best team will be ranked below #4 is not very high, probably not high enough to be worth the additional randomness you get from adding an extra playoff round.

  • Llynus Caldwell

    Most soccer leagues get around it by having two parallel titles, one is a round robin where teams get 3 points for a win, 1 for a draw, 0 for a loss, and the team with the most points at the end of the season wins. Whilst they also have a tournament bracket featuring every team.

    Best of both worlds, everything’s settled on the field, it identifies the most deserving team (i.e. the team that has been consistently the best over the whole season), as well as the team who’s performed when the pressure is on in the tournament bracket.

    You can’t really apply the same system to college ball though. I think the closest you could get to “settling it on the field”, with the practical difficulties of college football would be some type of modified Elo ranking (the one they use in chess, international soccer, and esports). Each team gets a ranking for their estimated strength based on their results, gain points for a win, lose points for a loss, the amount of points is determined by the difference by the two teams ranking before the game. At the end of the season, put the best 4 teams in a playoff.

    It would still encourage teams to schedule quality opponents, as a win versus a low ranked opponent isn’t worth that much, and a loss will mean a big hit to your ranking.

  • meh130

    What was good about the BCS was the formula. It was transparent, respected the polls, and attempted to correct for human poll bias with the computer polls. Could it be better? Sure, it should have included additional human polls (AP, Legends, etc.) into the human poll component. Also, every D1-FBS coach should have a vote in the coaches poll. If they choose not to vote in a given week, just average those who do vote.

    What is both good and bad about the Committee is it is a check on the human polls (large differences will probably influence the AP and Coaches polls towards the Committee’s rankings), but it is a very small sample compared to the human polls.

    I would prefer a transparent formula to a secretive committee. I would also prefer a formula which gave additional credit to teams who play a 13-game schedule over a 12-game schedule, and gave additional credit to teams who won a conference championship game.

    Expanding the playoff to six or eight teams would also be a benefit, but such expansion should provide a stronger chance for a conference champion to be seeded. If a six-team playoff is done, 13-game schedule teams should get the byes over 12-game schedule teams, regardless of rankings.

    Ideally I would like to see an eight team playoff with the first round two weeks after championship Saturday played on the Top-4’s home fields. Losers would be bowl eligible and still go to bowl games. Winners would proceed to the semi-finals. An eight team playoff would mean a 16-game season for many teams, which may be too many games. Reducing the regular season back to 11 games might be required. But an eight team playoff might also see more non conference championship game teams in the mix. Again, I think the home field advantage should go to teams who played more games.

  • stan

    The BCS was an enormous improvement over what it replaced. No, it wasn’t a perfect, but it was a playoff (2 team) and it began with the recognition that the human polls were badly flawed. The biggest mistake made by the BCS was changing the formula to placate the really, really stupid complainers. And they did that because they were afraid to tell the writers and the coaches that they did a really bad job.

    The most illogical arguments made about sports ever were the vast majority of the stupid arguments made against the BCS. It was flawed, sure. But most of the idiots slamming it were usually using arguments that were were just painfully stupid.

  • stan

    btw — you might want to point out the problem caused by the fact that teams change over the course of the season. Best? When? November? Over the course of the whole season?

  • gomer_rs

    Though last year both Baylor and TCU were left out with pretty good arguments that they should have been considered for “the best” team.

  • gomer_rs

    Instead of the fancy Elo math, why not use the committee but use it differently,
    6 highest ranked conference champions and next two highest ranked teams. Effectively if you win a power 5 conference you’re in, the highest ranked non-power 5 is in, and two other teams are in, giving a shot to ND and BYU.

    Week after league championships the 4 highest seeded teams host the 4 lowest seeded teams. Then the 4 winners to the bowls as they are now.

  • Kevin Rusch

    Of course, “need [a playoff]” is relative. Is it any better for determining the best team? Not necessarily. It sure would be fun to watch, however. Which, in all honesty, is what we’re here for. It’s not like many of us care that much whether Pennsylvania or Ohio produce the best academic football players, because that ship sailed a very long time ago.

  • Kevin Rusch

    This is totally agreeable, *except* that an 8-team playoff just means a 13-game season for 8 teams, a 14-game season for 4, and a 14-game season for 2 of them if you play a 12-game season. Out of 100+ teams, that’s pretty acceptable, I’m sure those teams wouldn’t mind the games, exposure, and revenue, and again, it’s not like they’re really college students anyway.

  • Pingback: Manchester United U21s vs. Liverpool U21s: Date, Time, Live Stream and TV Info | MizzTigers Today()

  • Pingback: Bowl Predictions 2015: CFP Projections with Conference Championships on Horizon | Sporty()

  • Pingback: Bowl Predictions 2015: CFP Projections with Conference Championships on Horizon - #1 Info Portal()