After blowing a 35-21 third quarter lead, Urban Meyer called timeout as the Wolverines lined up for a two point conversion that would have given That Team Up North a one point lead with about thirty seconds left in the game. All that rested on that one play for the Buckeyes was a win over their rivals, the continuation of a 23 game winning streak, and a possible birth in the BCS championship game. As I sat in the opposite end zone on a now downright cold and overcast day in Ann Arbor, arms crossed, I could barely stand to watch what was about to unfold right before my eyes—a season preserved or destroyed on the strength of one play from the two yard line. When Tyvis Powell stepped in front of Bo Dever and intercepted Devin Gardner’s pass, the third of the stadium clad in scarlet and gray let out a collective cheer that was as much a sigh of relief as it was an expression of unbridled joy. For at least one more week, anyway, the Buckeyes would survive.
As I was riding home from Ann Arbor, I was informed that Alabama had gone up on Auburn 21-7, a win that was a foregone conclusion by those pundits who had already anointed Nick Saban a god amongst football men. This is the precise kind of game that Nick Saban never loses, not against his rivals with a trip to the BCS championship game on the line and a chance to defend his back-to-back BCS titles. I remember getting back to the house in a city not too far removed from the hot dog water beltway and checking my phone to see that with about ten minutes left in the game, Auburn was hanging in there, down by just a touchdown. Alabama was unable to put them away, and when Saban made the fatal decision to go for it on fourth and one from the Auburn thirteen yard line instead of kicking a field goal to go up by ten with under six minutes, it was fait accompli. And yet, what transpired thereafter will forever be etched in the collective consciousness of not just Auburn Tigers and Crimson Tide fans, but Buckeye fans and Missouri Tiger fans and even some Cowboys from Oklahoma—the most improbable of plays to seal a game since a band inadvertently entered the field of play too early in another rivalry that lives in ancient lore.
I don’t believe I ever saw a game won or lost on a walk off field goal touchdown return, and certainly not in a game that for all intense and purposes eliminated the two time champs from another shot at the title, and I am saddened by the fact that I probably never will see it again, at least not during the regular college football season. The sad reality is that, had this same set of events occurred just one year later, it simply wouldn’t have mattered in the same way, as both Alabama and Auburn would have been in the new fangled college football playoff, and even with a loss, Ohio State might still have been alive for a playoff spot with a win over Michigan State in the B1G Ten Championship game.
Few things in the history of mankind have been as unpopular as the BCS. A big tax bill, a root canal, and even a frontal lobotomy are all preferable to the diehard college football fan who has already worked himself or herself up into a tizzy by week eight of the season because he or she fears his or her undefeated alma-matter will not be given a chance to play for a mythical national championship. Even yours truly had pet names for the BCS, such flattering monikers as the “Broken Cash Scheme” and the “Bowl Conspiracy Series.” When the system was seemingly rigged to stack the game with SEC teams in 2011, even this guy conceded that it was probably time to implement a new system.
During most of this season, as happens every season, fans began to attack the BCS with “what if scenarios” of undefeated teams being left out of the BCS game even before the first BCS standings are released. This was particularly a source of angst for Ohio State fans, who let out a weekly collective groan as the new flavor of the week was anointed the next team to jump an undefeated Buckeye team that was ranked # 2 preseason (we are all very impressed with the Ducks win over Nichols State), despite registering mostly double digit wins, albeit playing in what had already been accepted as a conference on par with the likes of the Sun Belt or the Colonial and a schedule comprised of the “little sisters of the poor.” This season most assuredly was going to be the biggest BCS dumpster fire in the history of this soul torturing behemoth of sports injustice, and fans clamoring for a sixteen team playoff were already licking their chops.
But a funny thing happened along the way. Louisville lost. Stanford lost. So did Miami. Then Oregon. Baylor was decimated by Oklahoma State on national television. Next up was Alabama, and then finally Ohio State, until what was left was essentially a fairly undisputed BCS championship game (and we can’t do something un-American like leave a SEC team out anyway). Despite the fact that only one time in the history of the BCS has an undefeated team been left out of the equation (Auburn in 2004, and the BCS got it right that year as Auburn’s SOS was not that strong), this little fact has not stopped fans from attacking the BCS with such venom that finally those in charge relented and banned those evil computers with their wretched formulas to BCS purgatory (no more anti-BCS books for you Dan Wetzel!!) for a completely and utterly unnecessary four team playoff, at least this year anway.
All of this in pursuit of finding Big Foot.
There is a reason that fans of college football “CFB hereafter” have referred to the college football championship as mythical. With 125 teams playing radically different schedules in 10 different conferences (plus some independent teams), you will never be able to accurately compare teams in order to whittle the number of teams down to the best 4, 8, or 16 teams in the sport without subjectively assessing the purported strengths and weaknesses of various teams–and unless you are invited to the dance, you can’t be crowned the bell of the ball. Call it what you want–a playoff, the BCS, or a decision by the committee, but in the end, some guys and gals you have never met (who probably have favorite teams and consequently a truck load of bias) will be deciding who will play in your artificially orchestrated playoff tournament. And yes, this is true even if the college football playoff becomes bloated to encompass 16 teams.
Fans love the playoff format, saying such things like “Division 1-A football is the only major sport that doesn’t settle its championship on the field.” But there is a reason for this, as Division 1-A football (now divided into two subdivisions, FBS and FCS, with the FCS using a full blown playoff) is unlike any other sports league. At the present, there are 125 FBS teams (moving to 129 next year). These teams play the bulk of their games (8 or 9) against mostly regional foes in now loosely affiliated conferences. The other 3 to 4 nonconference games are scheduled via contract by the various schools athletic directors, mostly based on economic concerns, sometimes as much as ten years in advance when there is no realistic way to assess the strength of said opponents by the time the game is actually played. What other league functions like that?
It also comes down to math to some degree or another which underscores the point that realistically, there is no other way to do scheduling in CFB assuming that there is not a major reorganization of the sport . Other leagues, like the NFL, have smaller numbers of teams that play a substantial portion of the leagues opponents, allowing reasonable conclusions to be drawn between the teams based upon on-the-field results to separate playoff qualifiers when record alone does not dictate a particular result. This can never occur in a CFB world with 125 teams and counting. To illustrate this point, let’s compare recent champions from the NFL to that of two-time defending BCS champ Alabama. In 2013, Alabama played a total of 11 different FBS opponents (the FCS team is not being counted) of the 125 total, or 8.8% of the total FBS teams. Now, let’s compare them to the defending Baltimore Ravens schedule, who play 13 of 32 NFL teams this season (duplicate division games are not counted twice), or 40.6% of the total teams in their league. This means that when one has to assess whether the Ravens should earn a playoff spot over another team, some reliable conclusions could be drawn from the data of similar opponents (and though the NFL uses records rather than voters to pick teams, head to head and conference records indeed are used as tie breakers for wild card spots).
This analysis can be taken one step further if one tries to find common opponents between contenders in any given year in CFB, for the commonly referred to but illusory transitive property (e.g. Baylor is better than Ohio State because they beat the Buffalo Bulls by a larger point spread than the Buckeyes did). This is illusory because football is a matchup sport, discussed more fully below, because some coaches intentionally do not run up the score likely to avoid injuries, and/or because a sample size of one game is too small to draw definitive conclusions. But if you ignore these variables for a moment, at least ostensibly, common opponents would allow one to draw some conclusions when assessing the various strengths and weaknesses of particular teams. Without some standardizing of CFB schedules, there will be little to no shared data that would allow one to draw conclusions to separate one team from another. As the 2013 CFB season wound down and pundits compared the resumes of Alabama, Oregon, Stanford, Baylor, Florida State, and Ohio State, there were only 3 instances where these teams played the same opponents (Ohio State and Baylor both played MAC team Buffalo, Ohio State, Oregon, and Stanford all played California). This is simply too few data points for a valid basis of comparison.
And of course, this comparison effect is exacerbated by the fact that NFL teams, by and large, have similar talent to each other, which means that the results of the games can be standardized to some degree for comparison purposes. This is the case because, all NFL teams draft from the same pool of players, the upper 1% or so of the best college players, and draft sequentially in inverse order to in effect spread the talent evenly (yes, I know it is not even for the Oakland Raiders because they use their draft picks on guys in the first round that you never heard of or on a fat guy who didn’t want to play football, but at least theoretically, it’s even). Simply put, it doesn’t work this way in college football. Teams recruit players, and there is nothing even-steven about it, as a disproportionate share of blue chip players tend to play for tradition rich schools that are presently winning. Why does this matter? Because at the heart of these team comparisons will ultimately be the need to determine if an Alabama win over Georgia Southern is worth more than an Ohio State win over the Buffalo Bulls. The only way one can even begin to answer this question definitively would be if Georiga Southern and Buffalo actually played, an event that statistically will not occur most seasons as demonstrated by the percentages above. As such, comparisons like this involve complicated strength of schedule formulas that can never adequately fill the void created by the absence of head to head results. Regardless, there is no way around the fact that such determintations will be the staple of conversation when comparing the relative strengths and weaknesses of contenders’ various schedules. And unfortunately, this will not change with a playoff format, meaning that the subjectivity that so inflamed the anti-BCS crowd will still remain as a key component of the process.
It’s at this point that the playoff opponent would undoubtedly say, “but that’s precisely why we need a playoff tournament, since it really is impossible to determine who is better than who unless they play.”
For the second straight season, the underdog Cardinal knocked off the Ducks essentially ending their BCS title dreams each of the last two seasons.
This ignores the second problem with the playoff format– that few teams play each other within the playoff bracket, meaning that who you play can be almost as important as how you play. This is because football is the ultimate matchup sport, where a team’s chance of winning depends in part on being able to expose another team’s weakness. If you don’t believe that matchups matter, just ask the Oregon Ducks, who were generally regarded as the best team in the Pac 12 based on recruiting rankings the last few seasons. The problem for the spread option Ducks is that they simply could not beat a more physical Stanford team, who has beaten Oregon each of the last two seasons, keeping them from playing in the title game in 2012 and from the Pac 12 Championship game in 2013. So even if you could determine “the best teams” and put them in a playoff, determining whether one team should be an eight seed over a nine seed would implicate the exact same problems as deciding who gets in and who doesn’t. Seeding the teams from a group with few common opponents would once again require a subjective analysis of whether one team’s 11-1 season was better than another’s. So the problem of using incomplete data to include and exclude teams is compounded by repeating the same process in order to seed teams for a playoff tournament. This is an exercise in futility.
Finally, the last problem with the playoff format for football is that it ignores the very nature of the game itself as a war of attrition. Football is a brutal sport, with each passing week taking players off the field, replaced with lesser players from the depth chart. While injuries may in part be due to things like conditioning, there is no denying that some injuries are simply a fluke based on random occurrences like a player’s spike being caught in the turf or a player rolling up on the back of an offensive lineman’s legs while he is blocking. It is not unusual for teams to have the largest number of injuries at the end of the season (or for players to be playing with a growing number of nagging injuries limiting their performance). And while this is beyond obvious, a team’s chances of winning against a quality opponent is diminished if several of its best players are either not playing or are playing significantly hurt. And yet, under the playoff model, this is precisely when a team is most likely to be playing without key players, meaning the results of playoff games can be skewed by the injury factor. And though it has often been a source of criticism, the period of time between the conference championship game and the title game, typically about four weeks, affords the two title teams an opportunity to get injured players back, maximizing the chance that teams play at or near full strength. This is precisely the opposite of the playoff bracket model, such as the one in the NFL, that requires you to play two or three games in consecutive weeks that often is the most physical football of the season because of the stakes associated with playoff games. So while players may be getting healthy in preparation for the BCS title game, the NFL playoff model is exposing its players to an increased risk of injury.
The problem with the playoff model is that it transfers the weight from regular season games to the playoff tournament, as each regular season game only matters such that your record is “good enough” to qualify for the post season tournament (so that an undefeated team and a one or a two loss team enter the tournament on equal footing—a loss removes each from the tournament). Thus, the function of the playoff tournament is to attach the most weight to games at the end of the season when a team is most likely to be its most injured (there also remains the possibility that injuries are disproportional, meaning that the healthy teams are bestowed an advantage based at least in part of fluke occurrences). This is particularly important for student athletes in that the NCAA heavily regulates the number of practice hours for its players, meaning that second string guys generally are not given sufficient practice reps to be prepared in the event they are called on late in the season.
And this point underscores the logic for the poll-based Division 1-A FBS model since its inception, codified by the phrase “every week is a playoff.” The basis of that logic is that selecting the two best teams after equally weighing all thirteen games (the college football season is 12 games for all teams, but most conferences use a conference championship game), means that, college football’s selection of a champion is based on a larger sample size. Any scientist will tell you that results are more likely to be accurate when they are based on a higher volume of data.
So let’s compare this to the NFL playoff model which is going to underscore the basic difference between the College FBS model and the NFL model—aptly characterized as an under-inclusion versus over-inclusion issue.
The NFL is comprised of two conferences of sixteen teams broken down into four divisions of four teams. The winner of each division, irrespective of record, earns a playoff spot, and each conference then gives two teams wild card spots, based primarily on record. Just based on this model, one can immediately see that the NFL is not putting a premium on record for entry into its playoff tournament, which has resulted in many instances where less deserving teams have been given a playoff birth at the expense of “better” teams. For example, in 2008, the Arizona Cardinals were in the playoffs with a 9-7 record while the 11-5 Patriots were excluded, a result that was completely unjustifiable. In 2010, the 7-9 Seahawks were given a playoff spot while both the Tampa Bay Buccaneers and New Giants were kept out despite 10-6 records. Such a set up is not a serious endeavor to crown the best team the champion, but rather, an effort to create an artificial tournament so that a playoff bracket can be filled out to provide the consumer with additional entertainment.
Now let’s take a look at the results in the NFL by looking at how often the best team record wise in the regular season actually won the title. Here are the results the last ten years (the number in parenthesis is the team’s seed by record of the 32 teams* irrespective of whether the team was awarded a playoff spot):
Year Team Record Superbowl winner
2003 New England 14-2 New England (1)
2004 Pittsburgh 15-1 New England (2)
2005 Indianapolis 14-2 Pittsburgh (7)
2006 San Diego 14-2 Indianapolis (4)
2007 New England 16-0 N.Y. Giants (7)
2008 Tennessee 13-3 Pittsburgh (2)
2009 Indianapolis 14-2 New Orleans (2)
2010 New England 14-2 Green Bay (10)
2011 Green Bay 15-1 N.Y. Giants (10)
2012 Denver/Atlanta 13-3 Baltimore (9)
Looking at these numbers, only once in the last ten years has the team with the best record actually won the Super bowl (with the “best” team only playing in the game 3 times). But looking at the seeding numbers suggests that the results have actually been even more ludicrous—twice the tenth seeded team won it, once the ninth seeded team won it, two times the seventh seeded team won it, and once the fourth seeded team won it (the other three times, the possible second seeded team won it).
In 2010, Seattle was the first team with a losing record to win their division. While two 10-6 teams were left out of the NFC playoffs that year, the Seahawks hosted and won a playoff game.
This underscores another myth about the playoffs that most fanatics are unwilling to acknowledge—since the talent level amongst the “best” teams is much closer than most are willing to believe, a winner take all one game format provides insufficient data to draw accurate assessments regarding relative team strength. While the NFL playoff system may generate close fun games to watch, and includes more franchises, at least based on historical data, it has been a poor method for determining the champion, assuming one believes the regular season record which encompasses the largest body of data is of any consequence at all. The NFL model is the classic example of over-inclusion because, at least based on season records, the NFL is clearly including teams that are inferior in order to simply fill out a playoff bracket. When one considers how often underdogs win games based on things not related to talent (e.g. Injuries, blown calls, odd bounce of a football), determining a champion by overweighing end of the season games is a poor method at best and an outright fraud at the worst. Say what you want about the BCS, but only one time since its inception has a team with two losses been even permitted to play in the title game—the 2007-08 LSU Tigers. At least based on won and loss record, it is hard to argue that the BCS champion has ever been undeserving. This pales in stark contrast to a system that somehow tried to sell us in 2007-08 that the 12-7 New York football Giants were the champs over an 18-1 Patriots team because of one over-hyped Sunday gridiron contest (and of course, you were supposed to just ignore that during the last week of the season, the Patriots beat the Giants).
So after all of the huffing and puffing about the BCS and the possibility that a deserving undefeated team would be shut out of the BCS title game this season, it turns out that all of the hysteria was all sound and fury signifying nothing. If you have followed my argument above, then you know that I am not a fan of including teams in a playoff tournament that are not qualified based on their season record simply to fill out a bracket and air contrived matchups, because anything can happen in a one game scenario.
Here is the final 2013 BCS standings, with the computer score in parenthesis:
1. Florida State, 13-0 (.996); 2. Auburn, 12-1 (.964); 3. Alabama, 11-1 (.906); 4. Michigan State, 12-1 (.860); 5. Stanford, 11-2 (.819); 6. Baylor, 11-1 (.772); 7. Ohio State, 12-1 (.771); 8. Missouri, 11-2 (.726); 9. South Carolina, 11-2 (.715); 10. Oregon, 10-2 (.581).
In looking at these standings, and some additional numbers which I will discuss in a minute, it is pretty clear there is no real controversy this season. As a matter of course, I am eliminating all two loss teams, because in my model, losses actually do matter. By that same reasoning, Florida State, as the only undefeated team from a major conference, is clearly in. That only leaves a reasonable discussion as to whether the second spot should go to Auburn, Alabama, Michigan State, Baylor, or Ohio State.
Ohio State and Alabama are the easiest to eliminate, as they each played one of the one loss teams ahead of them and lost, so no way the Buckeyes or Crimson Tide would get in over Michigan State or Auburn. This leads to whether arguments could be made that either Michigan State or Baylor deserve the second spot over Auburn. For this, I use advanced adjusted statistics from football outsiders.com which uses the FEI index, with the full composite rankings linked below.
Here is their description of their methodology, a topic that is too complicated to delve into here, but I encourage you, if curious, to spend some time at their site:
The Fremeau Efficiency Index (FEI) considers each of the nearly 20,000 possessions every season in major college football. All drives are filtered to eliminate first-half clock-kills and end-of-game garbage drives and scores. A scoring rate analysis of the remaining possessions then determines the baseline possession efficiency expectations against which each team is measured. A team is rewarded for playing well against good teams, win or lose, and is punished more severely for playing poorly against bad teams than it is rewarded for playing well against bad teams
Once adjusting for opponents strengths, Baylor came in at # 7 and Michigan State at # 9 in overall FEI (rankings encompass offense, defense, and special teams). Auburn, according to the overall FEI rankings, came in at # 4. From a simple resume test to include best win-best loss, Auburn beat Alabama (FEI # 3) while their sole loss came to LSU (FEI # 16). Michigan State’s best win was over Ohio State (FEI # 7), with its sole loss coming against Notre Dame (FEI # 19). Though their numbers are relatively close to Auburn’s, both are worse. Finally, for Baylor, while their loss was a quality one against Oklahoma State (FEI # 6), their best win was against Oklahoma (FEI # 20). From these numbers, it’s hard to argue that Auburn doesn’t slightly edge out Michigan State or Baylor for the BCS title game. All of this is to say, there would have been absolutely no need for a four team playoff this year. The utilization of one would have functioned to strip the regular season of its excitement that post Thanksgiving weekend in order to include two teams whose resumes already suggested they did not deserve a shot at the title.
I guess I am saddened because I one day see CFB moving to an eight or sixteen team playoff, not because this would somehow be a better way to determine the “best” team, but because conference commissioners, athletic directors, and television executives will realize that it is lucrative to do so.
Once upon a time, CFB simply acknowledged the reality of the mythical title, and teams battled for regional supremacy. Growing up in the Midwest, all that really mattered for Ohio State fans was: 1) beating Michigan; 2) beating Michigan badly, and oh yeah, somewhere thereafter; 3) winning the conference and going to the Rose Bowl. Where you were voted in the polls was simply the proverbial cherry on top of the sundae.
A strange thing happened in Ann Arbor this year—thanks to the Conference Championship Game model that at times can be a farce, as discussed here, it really didn’t matter whether Ohio State beat Michigan or not, as they had already earned the right to play in the conference title game against Michigan State the following week, and the only talk that dominated the national landscape was whether TTUN could eliminate Ohio State from the BCS Championship game. Expand the playoffs to eight or sixteen teams, and this effect will be exacerbated, as an undefeated Ohio State team that might make the playoffs might decide that resting banged up players for future “playoff games” to come may make more sense than trying to beat their rival. For a sport that defined itself by rivalry games and regional bragging rights, the selling of its tradition for a few illusory championship games is yet one more way big boy division one football has sold its tradition in pursuit of the almighty dollar. If the NCAA was willing to take an already bloated but wildly successful NCAA basketball tournament of 64 and expand it to 68 (for silly play-in games involving teams that have no shot or claim to be in such a tournament), it would seem to be only a matter of time till they cave again to the torch and pitchfork masses and television executives and expand the CFB playoffs to either eight or sixteen teams.
* As a note, the NFL does not seed the 12 teams collectively from the AFC and the NFC. For my rankings, I simply went by record to include the 12 best records, irrespective of whether they were actually included in the playoffs. If two teams had the same record but one was a division winner, I ranked them higher, because the NFL seeds division winners that way. If two teams were either both division winners or neither were, I gave the benefit of the doubt to the Super Bowl champ and seeded them higher in an effort to be conservative in my seeding numbers.