A Look Back at the Coalition/Alliance/BCS Era: Would a Playoff Have Helped?

It being a Saturday, I think it might be appropriate for a college football article...

As some of you may know, I have been something of a, oh let's say, vocal opponent of the BCS for many years. I have stated in the past that the BCS (or a BCS-like system for years prior to 1998) has produced "true" champions -- meaning champions with no "what ifs" attached -- at an appallingly low rate. I pointed out that under the BCS we continue to have ludicrous "split champions", even though the entire point of the BCS in the first place was to eliminate such situations forever (actually, that's not totally true; the entire point of the BCS is to make money for university athletic departments, but that's another story for another article). Anyway, I felt that the BCS had done a generally lousy job of determining the National Champ in one of the country's most popular sports, and, furthermore, that it was unacceptable for college football to continue to have elements of its championship determined off of the playing surface, similar to freak-show fringe sports like boxing and figure skating.

In light of such a horrible record, I reasoned, it was time for a change. Give us the "plus one," a Final Four of sorts for college football. To quote myself from 2004, it would go a little something like this:

"As for [BCS] alternatives, a 64 (!??!) team playoff is not very practical. Remember, only 6 weeks is half the length of the entire season, plus a 64-team playoff would include over half the teams in Division 1A! 16- and 8-team playoffs are better ideas, but still foolish: teams with 2 and even 3 losses would be given a chance to play their way to the national championship, a concept that would render the regular season meaningless. Case in point: everyone loves the NCAA Basketball Tournament, but it rarely produces as champion the best team throughout the season; perhaps the best example of this is Villanova in 1985. The Wildcats lost 10 regular-season games and barely made the Tournament, but they got hot for 6 games and won the National Title. I think it would be as unfair as the BCS' arbitrary 2-team format to allow undeserving teams to play their way to a title simply because they got on a roll for a few playoff games."

"It's fairly clear, then, that the solution to the BCS is to simply expand it from 2 to 4 teams. Use the existing ratings system to determine the Final 4, and then pit #1 vs. #4, and #2 vs. #3. The winners would then face off in a true National Title Game. The 3 games used for this playoff would be existing BCS bowls, with the hierarchy rotating each season (much like what is done currently). For example, if the Rose Bowl was the Championship Game one year, it would be left out of the playoff the next year (selecting teams its normal way), and then be one of the semifinals the next year. The other bowls would go on just as planned, and we would finally have an undisputed champion. Check the results, and you'll find that this system rectifies every major complaint about the outcome of each of the last 8 seasons. No more split titles. No more controversy. It's about time, college football."

Makes sense, huh? Four is generally the maximum number of teams that deserve to be National Champs in any given year anyway (2-loss teams typically don't deserve it, 2007's mess notwithstanding), and we'd finally have a champ crowned on the field. Plus, it would only add one extra game to the schedule, which seems to be a big consideration for the university presidents (ironically, they aren't bothered at all by the month-long mid-semester NCAA Basketball Tournament. But like I said, another story for another article). The plus-one is the most viable playoff out there -- basically, if we get to see an expanded playoff of any type in the next 10-20 years, it's going to be this one. It wouldn't satisfy everybody (what playoff would?), but this proposal is at least way better than the current system, right?

Well, maybe not. In this article, I'm going to look back at the 1992-2007 seasons and see exactly how well a Final Four or "plus one" type of playoff would have fared. The rules are simple: seed based on the final pre-bowl AP poll (available here), and have #1 face #4 in one semifinal & #2 face #3 in the other; then, the winners square off for the championship. If a matchup occurred in real life, we (duh!) go with the real-life outcome. If not, I'll do my best to determine who would be favored by Vegas if they had to come up with a line for the game. I'll also explain why I think a certain team would be favored, although your opinion certainly may differ. At the end, we can look back and see how many times the Coalition, Alliance, and BCS got the correct champion at the end of the year. And I'll say that, while the BCS is incredibly flawed and deserves to be maligned as such, it really hasn't done that much worse than a Final Four would have.

2007

Let's just get this out of the way early: 2007 would have been a fiasco for any playoff system, not just the BCS or the plus-one. Eleven (12 if you count BYU) viable teams with 0-2 losses heading into the bowls? While the temptation is to use 2007 as justification for the ever-popular 16-team bracket, it remains to be seen whether 2007 was an aberration or the start of a new trend. All I know is that under the "old rules" of college football, 16 is way too many teams to throw into a playoff -- you're eventually going to end up with a 3- or 4-loss champion, which would totally invalidate the regular season. And that's one of the few things the BCS has going for it: it maintains the sanctity of the regular season more than any other sport.

At any rate, this was the pre-bowl poll. Putting aside who should and shouldn't have been in the Top 4, I think Ohio St. would have been narrowly favored over Georgia by virtue of their pretty superior point differential (18.6/g vs. 12.5/g), fairly equal schedule difficulty (for reference, I'm going to be using Andy Dolphin's schedule strength metric here because it's easy to use, goes back to 1992, and it's no worse than any other computer ranking), and the fact that Ohio St. had 1 fewer loss. LSU would be favored over Oklahoma because of LSU's schedule strength and the fact that OU has a reputation for, um, choking in postseason play. In the final, we get the real-life LSU-OSU matchup, with Louisiana St. winning 38-24.
Verdict: BCS got it right

2006

The pre-bowl AP poll gives us the Buckeyes and the Tigers in one semifinal and the Gators and Wolverines in the other. Ohio St. had 2 fewer losses and equal point differential/schedule strength, so they would definitely have been favored over LSU. Similarly, Florida's point differential was better than Michigan's (against arguably a tougher slate) and Michigan was coming off a tough loss to Ohio St., so you have to think the Gators would have been expected to win that matchup. Again, we get the real-life final, which Florida won quite handily, 41-14.
Verdict: BCS got it right

2005

In 2005 you had 2 clear-cut teams that have separated themselves from the rest of the pack, so the argument can definitely be made that the BCS would have been more favorable than a Final 4. Still, in our alternate universe you'd see #1 USC favored over #4 Ohio St. because of 2 fewer losses and a massive difference in point differential (26.2 vs. 17.4), and #2 Texas favored over #3 Penn St. due to Texas' even bigger advantage in scoring differential (33.8 to 17.4). In the final, Vince Young and Texas prevail over USC in one of the classic college football games of all time.
Verdict: BCS got it right

2004

2004 was another difficult season for most playoff systems, as 8 teams went into the bowls with either zero or one loss. The AP would have seeded it this way: #1 USC vs. #4 Cal in one semi, and a battle of undefeateds, #2 Oklahoma vs. #3 Auburn in the other. USC would be favored over Cal because of their superior resume and the fact that the Trojans had already defeated the Golden Bears 23-17 earlier in the season. Oklahoma-Auburn would have been a tough call (made tougher by Oklahoma's weak showing vs. LSU in the semifinals the season before), but I think oddsmakers would have looked at OU's tougher schedule and favored the Sooners going into the final. And either way, neither Oklahoma nor Auburn would have been favored against USC, who destroyed OU 55-19 in the real-life title game.
Verdict: BCS got it right

2003

In 2003, our playoff system would have sorted out the three 1-loss teams who stood atop the AP poll going into the bowls. In one semifinal, USC's superior credentials would have had them favored against Michigan, while the other semifinal matchup actually happened in real life, with LSU shutting down Oklahoma 21-14. In the final, you have to think Louisiana St. would be favored to beat the Trojans, given their stronger defense, better point differential, and tougher schedule strength. So just like in real life, LSU is crowned champion, albeit via a different route.
Verdict: BCS kinda got it right

2002

2002 saw two undefeated teams and two one-loss teams, an ideal situation for both the BCS and plus-one systems. #1 Miami would definitely be favored over #4 UGA, since they had not lost in 34 games, while #2 Ohio State would have been narrowly favored over #3 Iowa because of their tough D, undefeated record, and slightly harder schedule. In the final, the Buckeyes outlasted Miami 31-24 in a real-life double-OT thriller.
Verdict: BCS got it right

2001

In one of the BCS' biggest controversies ever, Oregon/Colorado got passed over for #2 in favor of a Nebraska team that didn't even win its conference championship and was fresh off of a brutal 62-36 whipping by the Buffaloes. The plus-one fixes that misdeed, with Nebraska seeded behind both Colorado and Oregon going into the semis. In the first, we get to see Miami trash the Huskers just like in real life, while in the other we get to see the Ducks trash the Buffs... just like in real life. In the much-sought-after final between Miami and Oregon, Miami is heavily favored because they're a vastly superior team with a much better point differential against a much tougher schedule. So even with the boneheaded pick of Nebraska, the BCS selected the right champion after all.
Verdict: BCS kinda got it right

2000

The 2000 season featured just one undefeated team (Oklahoma) and a number of 1-loss squads, leading to the first big controversy surrounding the selection of the BCS' #2-ranked team -- Florida State made it into the title game despite having lost to Miami, who was ranked 2nd by both human polls. Luckily, in our universe there was no such controversy, as the only teams to get left out (Oregon State and Virginia Tech) were inferior in the schedule department and combined for just one win over a Top 25 team. In the first semifinal, #1 Oklahoma would be heavily favored over #4 Washington because of a much better point differential against a harder schedule, while #2 Miami would be a narrow favorite over #3 FSU because the 'Canes had already beaten Florida St. earlier in the season. In the final, I feel that the two teams would be very evenly-matched, but that Oklahoma would be narrowly favored as an undefeated team coming out of the tough Big XII.
Verdict: BCS kinda got it right

1999

With 2 clear-cut undefeated teams atop the polls, there wasn't a whole lot of BCS controversy brewing -- and therefore no need for a different system -- in 1999. Nonetheless, in our first semifinal, #1 Florida State would be expected to beat #4 Wisconsin because of a 2-loss disparity and a gap in schedule strength favoring the 'Noles. In the second semi, Michael Vick and #2 Virginia Tech would be narrowly favored over #3 Nebraska because of their record and superior point differential (despite playing a slightly weaker schedule). And that sets up the real-life title game, which was won pretty convincingly by FSU.
Verdict: BCS got it right

1998

As the inaugural BCS campaign, 1998 was a controversial season all around. Perhaps it's fitting, then, that our Final Four would see its first controversial result as well. In one semifinal, #1 Tennessee would go up against #4 Kansas St., and an objective look at the numbers would have to lead Vegas to favor K-State despite Tennessee's undefeated record -- the Wildcats had a significantly better point differential against a tougher schedule, playing in a much more demanding conference. They were simply the better team that year. Similarly, in the other game between #2 FSU and #3 Ohio State, the Buckeyes would be favored because of their superior margin of victory and far stronger conference alignment. In the final between Ohio State and Kansas State, K-State would be favored to win again for the simple fact that they were a better team than the Buckeyes.

Then again, this hypothetical argument that K-State would have been favored over both Tennessee and Ohio State (while probably true, given the information the bookmakers would have had to work with) kind of gets deflated when you consider that the Wildcats went on to lose the Alamo Bowl to Purdue 37-34, despite being 13-point favorites. The counter-argument is that KSU was so miffed about being passed over by the BCS that they came out flat against the Boilermakers, but the fact remains that we can't definitively say the BCS got it wrong in '98.
Verdict: BCS got it wrong... maybe

1997

In 1997 we saw a split title, which in many ways provided the impetus to create the BCS in the first place. We can assume that the real BCS would have given us the sought-after matchup of Michigan and Nebraska in its championship, had it existed in 1997. What will the Final Four decide? In one semifinal shocker, #1 Michigan would not have been favored against #4 Florida State, as the Seminoles' superior schedule strength and point differential would have suggested that they were in fact a stronger team than the undefeated Wolverines. In the other semifinal, #2 Nebraska destroyed Peyton Manning's Vols in real life, 42-17. In the final, Florida State and Nebraska would have been closely matched, but the Huskers would be favored because they were the more dominant team all year long, and they were coming off a dismantling of the #3 team in the country.
Verdict: Alliance kinda got it right

1996

1996 was an interesting season, as 6 teams had either zero or one loss going into the bowls, making the BCS and the Final Four equally controversial for having to set a cutoff point. But at least the plus-one would exclude just BYU and Virginia Tech, neither of whom played a very strong schedule. We can't even begin to assume who the BCS would have taken as their 1-2, but we know the Alliance chose Florida State-Florida as its championship game, creating a rematch of a game that took place just weeks earlier (FSU won 24-21). The Final Four would consist of the 4 clear-cut best teams in the country: Florida St., Arizona St., Florida, and Ohio St. In the first semifinal, I would expect Ohio State to be favored over the #1-ranked Seminoles despite their undefeated record -- remember, in the real-life Sugar Bowl, Florida was favored over FSU despite having lost to them a month earlier. Besides, OSU had a superior point differential and a tougher schedule. In the other semi, Florida would be favored against Arizona St. because of a more dominating season against an equally-strong schedule. The final would essentially be a push, but I think the oddsmakers would give the Buckeyes the benefit of a doubt after ending FSU's undefeated season in the semis.
Verdict: Alliance got it wrong

1995

The Alliance worked out perfectly in 1995, as there were only two undefeated teams going into the bowls; a Final Four system would have had to choose between one-loss Northwestern, Ohio State, and Tennessee for the the final 2 slots in the playoff. You could make a case that the AP's decision to exclude Tennessee is wrong, but you have to place the cutoff somewhere, so we end up with #1 Nebraska vs. #4 Ohio St. in one semifinal and #2 Florida vs. #3 Northwestern in the other. I don't think there's any way that Nebraska and Florida aren't favored in those games, and we all know how badly the Cornhuskers decimated Florida in the title game, so this was a clear case of the Alliance doing what it was intended to do: identify the 2 best teams in the country, and match them up in a championship game.
Verdict: Alliance got it right

1994

1994 would have been perfect for the BCS -- two undefeated teams and a bunch of teams with one loss and/or one tie -- but under the old Bowl Coalition (which did not include the Rose Bowl conferences, the Big Ten and Pac-10), we had to settle for the two best teams in the country playing lesser competition in different bowls. Under our Final Four system, though, fans would get to see their dream matchup of undefeateds... if both could advance to the final, that is. In the first semifinal, there's no scenario under which #1 Nebraska would not be favored over #4 Colorado, a team they were clearly superior to and had beaten 24-7 during the season. It's also hard to imagine #2 Penn St. not being heavily favored over #3 Miami, given their vastly superior average margin of victory and their tougher SOS. In the final, I believe that Penn St. would be favored over Nebraska, as PSU had the better point differential and played a harder schedule, in a tougher conference.
Verdict: Coalition got it wrong

1993

At first glance, 1993 looks like a mess, with 6 teams entering the bowls at zero or one loss (and 3 more with one loss & one tie). However, the picture gets clearer when you consider that Auburn was ineligible for the postseason, and that Texas A&M's schedule was not as strong as that of FSU, Nebraska, or Notre Dame. So our Final Four pits #1 Florida St. against #4 Notre Dame in a rematch of the Game of the Century, and #2 Nebraska against #3 West Virginia in a battle of undefeated teams. Despite Notre Dame's victory earlier at South Bend, Florida St. would definitely be favored over the Irish, while Nebraska would just as certainly be favored to beat the Mountaineers. And in the final (which happens to be the real-life title game as well), FSU outlasted Nebraska 18-16 when Byron Bennett hooked a potential game-winning FG wide left.
Verdict: Coalition got it right

1992

The 1992 season looked like a quagmire in the making in November, but Washington lost twice in the season's final 3 weeks, leaving only three undefeated teams and one with a single loss -- a perfect setup for the Final Four system. #1 Miami, who garnered all but 1 first-place vote and was in the midst of a 29-game winning streak, would definitely be favored over #4 Texas A&M in the first semifinal, and I have to think #3 Florida St. would be favored to beat #2 Alabama -- not only did Florida St. have a superior point differential and schedule strength, they whipped Florida 45-24 (albeit at home) the week before the Tide beat the Gators just 28-21 at a neutral site, and the Seminoles only lost to Miami by 3 in Coral Gables (while Alabama was an 8-point underdog going into the real-life Sugar Bowl vs. the Canes). In the final, Florida St. would have been favored to snap Miami's win streak on a neutral field because of their superior average victory margin, SOS, and the fact that their lone loss came by just 3 points, on the road at Miami.

However, what transpired in the real-life bowls greatly diminishes an argument for anyone but Alabama in '92. While the bookmakers would probably have favored the Seminoles over the Crimson Tide in a hypothetical matchup, they also favored Miami pretty heavily (-8) over Alabama in the real-life Sugar Bowl, which the Tide went on to dominate by a score of 34-13. FSU might have been favored over 'Bama, but there's no guarantee they'd win -- meaning we can't say for sure that the Coalition was wrong here.
Verdict: Coalition got it wrong... maybe

Conclusions

Although the BCS has been (fairly or unfairly) criticized by just about everyone on the planet at some point over the past decade, a look at the results shows that it has actually not fared any worse than a plus-one system would have in selecting true college football champions. In its ten years of existence, the BCS has picked the same 1-2 matchup that would likely have occurred under a Final Four-type playoff six times, and has ended up with a different champion than the plus-one just once -- 1998, its inaugural season -- and that's actually debatable, since K-State went down in the real-life Alamo Bowl (although to be fair, Ohio State would have been favored over Tennessee as well). Just one "incorrect" champion in 10 years is actually not a bad track record... if you buy into the idea that all's well that ends well, that is.

Looking back further into the past, we can see that while a 4-team playoff would have been a definite improvement over the Bowl Coalition/Alliance system, a 2-team BCS-like system would have been acceptable as well in every season but 1996 (and perhaps 1992, but that's very debatable). All told, there are only 2-3 instances in the last 16 years where a 4-team playoff would have been a clear upgrade over the BCS. And before you say, "what about an 8-team playoff or a 16-team playoff?", I'll repeat what I said earlier: A) It's never going to happen (the closest we'll see in our lifetimes is the plus-one/Final Four, and even that's unlikely); and B) Why would we want it to happen? 2007 notwithstanding, all an 8- or 16-team playoff would accomplish is introducing into the championship mix a number of teams whose regular season performances did not earn them the right to play for a championship. And, frankly, diminishing the importance of the regular season like that may actually be a bigger travesty than determining the champion with computers & polls instead of with games on the field.

So we're left with either the 2-team BCS or a "Final Four"-style plus-one. And despite the rampant criticism directed toward the former (plenty of which has come from yours truly), a 4-team playoff would not improve matters that much. I'm not saying we should give up and resign ourselves to never having a DI-A College Football Playoff, but I am saying that what we have now isn't actually that bad. Given that we're most likely never going to have a better system, and that if we do, it will simply be a Final Four-type playoff as laid out above, we should probably just be happy with what we have. The BCS is far from perfect, but throughout its existence it has done the job essentially the same as a 4-team playoff would have. Besides, maybe all of those BCS controversies add flavor and excitement to the college football season, ultimately giving it more meaning. Not buying it? Okay, fine. But when all the alternatives either represent only a minimal improvement or are impossible, perhaps the best thing to do is to enjoy what we have today. Like the old Stephen Stills song says, "if you can’t be with the one you love, love the one you’re with."

---

Filed under: College Football, Classic Posts