Defining a “National Champion”

December 29, 2007

The first part to defining a “national champion” is easy: national refers to this nation, the USA, and therefore only US schools are eligible. Or, at least until the NCAA admits Canadian schools. But for the moment, it’s about US universities only.

The second part is tricky, because “champion” can be defined in several different ways. Here are a few I could find:

  • Dictionary.com: “anything that takes first place in competition”
  • Wikipedia: “one who has repeatedly come out first among contestants in challenges (especially the winner of a tournament or other competition) or other test”
  • Popular Use: The best team of a particular sports league in a given year, as in, “The Bears were the 1985 NFL Champions.”

By Dictionary.com’s definition, every game has a champion because one team takes first place in the game while another takes a loss. Wikipedia’s definition is more exclusive since it requires repeated first place finishes, so that eliminates 1-loss FIU, for instance, from being a champion. However, while it’s not specific on how many first place finishes define a champion, it singles out tournament winners especially as being champions. The popular use definition is the most specific, and it is commonly the one used when talking about national champions for college football.

How Many National Champions?

The standard assumption is that there can’t be more than one national champion, which is why people say things like “In 2003, there was a split national title” rather than “In 2003, there were two national champions.”

It’s all semantics, but it’s an important distinction. The NCAA’s website provides “a year-by-year history of Division I-A football national champions as determined by the BCS championship game and… polling organizations.” Since the NCAA maintains that I-A football “do[es] not participate in the NCAA Division I Football Championship,” it doesn’t care how many champions are named every year.

The BCS does care about the number of champions however, stating that it “was established to determine the national champion for college football,” (emphasis mine), not a champion for college football. This fact proves that at the very least the 6 BCS auto-bid conferences and their members, Notre Dame, and the 4 BCS bowls care about having one and only one champion as those were the institutions that created the BCS.

Now, some like Kyle T. King of Dawg Sports argue that college football doesn’t necessarily need a champion, and that’s fine. However, I’ll eat my hat (and it’s pretty nasty; ask my girlfriend) if like-minded people aren’t few and far between. Way too many people invest way too much time in determining/arguing over who is the best team for many people to be of the belief that naming a champion is superfluous. Plus, universities, investors, and advertisers have invested billions of dollars in the BCS, a construct designed to determine one and only one champion. I will not spend any more time on the concept of college football not needing a champion because it’s a fringe view at best.

The existence of the BCS to determine a champion also proves that the powers that be of college football believe that the regular season is not enough to determine a champion. Due to imbalanced scheduling and differing strengths of conferences, I wholeheartedly agree. The regular season alone is insufficient for choosing a champion.

Determining Who Is Best

Does a champion by definition indicate the best team? Go back to the definitions for a sec, I’ll wait…

Notice how only the popular use definition includes the requirement of a team being “best?” All that is required by the established definitions is first place finishes, not being the best.

Some people define the “best” as having won the most games in the most impressive fashion. Other define the best team as having the best group of players. When you get down to it though, there’s a dizzying array of shades of gray when it comes to determining the best team over the course of a season.

What do you do with Oregon? The Ducks looked like world beaters with Dennis Dixon at quarterback, but as soon as he went down, the team lost it’s heart and went into the tank on both sides of the ball. Which is the real Oregon? And when you’re picking who’s best, will you consider the Ducks with Dixon and the Ducks without Dixon to be the same team when they clearly were not?

What do you do with Hawaii? No one finished first in games more often than the Warriors did. Perhaps the games were not as challenging as others’ games, but no team fits the “champion” definition more than Hawaii does.

What do you do with Georgia? Once Mark Richt was forced by injury to quit being stubborn and play Knowshown Moreno, the Bulldogs became one of the better offensive teams in the country. Around the same time, Richt loosened up and changed his attitude and as a result the defense played better too. Which Georgia team do you count when you’re picking who’s best, the Moreno-less, fire-less team that lost to South Carolina and was blown out by Tennessee, or the team that had Moreno playing and had proper motivation that finished the season strong? Or do you count them as the same team when they clearly were not?

How about the Boston College at Virginia Tech game? For 57 minutes, VT dominated and ended up leading 10-0 over a clearly overmatched BC team. Then, for some inexplicable reason, VT switched to a prevent defense that allowed Matt Ryan to throw two TD passes to give BC a 14-10 victory. Now, which team was better? Having watched most of the game, I can tell you that VT was the better team that day, with the Hokies playing better than the Eagles for 57 of the 60 minutes. However, BC’s 3 good minutes allowed it to finish in first place for the contest. Was the best team the champion? Simply put, no.

The point is, it’s nearly impossible to choose which team is best in a season because of how many variables there are involved. Injuries and differences in schedule strength especially make comparison difficult, not to mention strange outcomes like the BC-VT game that hide what’s really going on. Picking who’s “best” is a fool’s errand.

What’s all this mean?

If picking one “best” team is a fool’s errand, what should picking the two “best” teams for a national championship game be called? Well, I’ll leave that as an exercise for you, dear reader. Some people would say that picking the two best teams is called the BCS, but actually, it’s not.

You see, the BCS says in reference to all 5 of its games that it “has become a showcase for the sport, matching the best teams at the end of the season.” That statement is a tacit confession that it can’t determine the best team, only the best teams. Yeah, it’s semantics like before, but again, this is important.

College football’s current post season is not about finding the best team, only determining a champion. Any proposed playoff system would do the same exact same thing – determine a champion, not the best team – because it’s impossible to even precisely define what makes a team best.

And, don’t forget, a post season is required for determining a champion for the reasons stated prior, so judging teams solely on the regular season isn’t enough.

The more precise definition of “champion” requires repeatedly coming up first, especially in a tournament. Because of that fact, it stands to reason that a tournament, which requires repeated first place finishes, is a better champion-finding system than a one-shot national championship game it since that requires only one first place finish. That’s really all there is to it.

Advertisements

The National Championship Game, Part II

December 28, 2007

In the first post titled “The National Championship Game,” the most interesting stat (to me, anyway) I uncovered was that since 1992, only 12 unique teams have participated in the college football national championship game. Now, part of that is because from 1992-97, no Big Ten or Pac 10 teams participated in the Bowl Coalition or Bowl Alliance. If you go by who finished #1 or #2 in the AP Poll for 1992-97, that number increases to 15.

It still made me wonder whether this was on par with other sports. After all, a common complaint is that baseball lacks of competitive balance with major market teams outspending the smaller market teams. Plus, since 1988 only 6 different franchises have won NBA championships. I want to see how much variety some other sports get at the end of the season compared to college football. So, I looked at the participants in the finals of the MLB, NFL, NBA, and NHL, plus the championship game of the NCAA D-I tournament. In addition, I threw in I-AA football since it is somewhat similar to I-A. Any division lower than that is beyond my realm of familiarity.

How does the polls-and-bowls system stack up against other sports in variety of finalists?

I-A College Football

15 Unique Teams: Alabama, Miami, FSU, Nebraska, Penn State (AP #2 in 1994), Florida, Arizona State (AP #2 in 1996), Michigan (AP #2 in 1997), Tennessee, VT, Oklahoma, Ohio State, LSU, USC, Texas

Percentage: 15/32 = 46.88%

This is our baseline. Only 46.88% of the top 2 at the end of the season have been unique teams. College football is known for having marquee teams dominating, but as mentioned above other sports have the same perceptions. If we assume that college football’s top teams come and go at the same rate as other sports’ do, we can now determine whether the variety is lower or higher than other sports’ varieties.

NBA Finals

17 Unique Teams: Chicago, Portland, Phoenix, Houston, New York, Orlando, Seattle, Utah, San Antonio, LA Lakers, Indiana, Philadelphia, New Jersey, Detroit, Miami, Dallas, Cleveland.
Percentage: 17/32 = 53.13%

Jordan/Pippen. Olajuwon. Stockton/Malone. Duncan. Shaq/Kobe. Kidd. Of all the multiple appearances of teams in the finals, only Detroit didn’t have at least one superstar or superstar combo leading the way. With basketball, that’s almost all you need since only 5 guys from a team participate at a time. Sometimes, all you need is one transcendent star and a grab bag of guys who don’t care about their own shot to get to a Finals – think Iverson and 2001 Philly or LeBron James and 2007 Cleveland – though you’re not likely to win the title once you get there.

Div. I College Basketball

18 Unique Teams: Duke, Michigan (vacated), UNC, Arkansas, UCLA, Kentucky, Syracuse, Arizona, Utah, UConn, Michigan State, Florida, Maryland, Indiana, Kansas, Georgia Tech, Illinois, Ohio State

Percentage: 18/32 = 56.25%

Despite college basketball’s tournament being known for upsets and chaos, when it comes to the final game the cream rises to the top. Some of the lack of variety can be attributed to the fact mentioned above that with basketball, you can ride 2 or 3 outstanding players to the finals or a championship. After all, Michigan, Duke, Arkansas, Kentucky, and Florida all had teams in back-to-back championship games or at least two in three years.

Major League Baseball

17 Unique Teams: Toronto, Atlanta, Philadelphia, Cleveland, NY Yankees, Florida, San Diego, NY Mets, Arizona, Anaheim, San Francisco, Boston, St. Louis, Chicago White Sox, Houston, Detroit, Colorado

Percentage: 17/30 = 56.67%

Remember, there was no 1994 World Series due to the strike, so there’s only 30 possible teams since 1992.

Baseball has seen some remarkably consistent winners in the Atlanta Braves and New York Yankees. Some teams bought their championships and were really good for only one year (’97 Marlins, ’01 Diamondbacks), some got hot at the right time (’98 San Diego, ’07 Colorado), and others just made timely pitching acquisitions (’05 White Sox and Astros). In the end, only the Yankees’ hegemony over the AL and Braves’ hegemony over the NL (6 and 4 Series appearances, respectively) kept the number so low; besides them, no team has appeared more than twice in the World Series, with only Toronto doing it in back-to-back years.

National Football League

21 Unique Teams: Washington, Buffalo, Dallas, San Francisco, San Diego, Pittsburgh, Green Bay, New England, Denver, Atlanta, St. Louis, Tennessee, Baltimore, NY Giants, Tampa Bay, Oakland, Carolina, Philadelphia, Seattle, Indianapolis, Chicago

Percentage: 21/32 = 65.63%

The NFL is supposed to be built for parity, and you can see that reflected in the larger number of unique Super Bowl participants. This time period does include 4 visits by the Patriots and 3 each by the Cowboys and Bills, but only Favre’s Packers, Elway’s Broncos, Warner’s Rams, and Cowher’s Steelers made multiple visits. The league got what it wanted for the most part, though the current Colts and Patriots seem to be increasingly resistant to the parity virus.

National Hockey League

20 Unique Teams: Pittsburgh, Chicago, Montreal, Los Angeles, NY Rangers, Vancouver, New Jersey, Detroit, Colorado, Florida, Philadelphia, Washington, Dallas, Buffalo, Carolina, Anaheim, Tampa Bay, Calgary, Edmonton, Ottowa

Percentage: 20/30 = 66.67%

Remember, there was no 2005 Stanley Cup playoffs due to the lockout.

I must confess I’m not really a hockey fan; I attempted to get into it when the Lightning went to the finals a couple years ago, but I just couldn’t watch an entire game on TV. I think part of the problem was the neutral zone trap, a defense of which I have a tenuous grasp; basically it sucks the like out of the sport. It was one of the things fixed as an aftermath of the lockout, along with the institution of a salary cap. It’s too early to say what effect the salary cap will have on competitive balance, but no sport in the past 16 post seasons has produced more finalists than the NHL has. It’s mainly due to parity in the early ’90s; since 2000, only 6 of the 14 finalists have been newcomers, whereas 1992-99, 14 of the 16 finalists were newcomers (Detroit being the only returner).

Div. I-AA Football

13 Unique Teams: Marshall, Youngstown State, Boise State, Montana, McNeese State, UMass, Georgia Southern, Furman, Western Kentucky, Delaware, Colgate, James Madison, Northern Iowa

Percentage: 13/32 = 40.63%

Here, we see that I-AA football actually produced fewer finalists than I-A did. Looking at the results, I can tell you that it’s because of coaches  cultivating dynasties: Jim Donnan at Marshall, Jim Tressel at Youngstown State, Paul Johnson at Georgia Southern, and recently Jerry Moore at Appalachian State.

Why so few finalists in college football?

As stated above, the I-AA championship game can be dominated for years at a time by excellent coaches that are too good to stay at the I-AA level for long (such as Donnan, Tressel, and Johnson). Opponents of a playoff might try to point out the fact that from year to year, the best teams change more slowly in college football than other sports, so picking a top two should be easier than in the NFL for instance, where recent Super Bowl losers have been prone to falling off the map.

Consider this though. The I-AA football playoffs since 1992 have all consisted of 16 teams apiece, and while the number of teams seeded has changed, the top four seeds every year are enumerated. The teams are determined by a playoff committee, similar to the basketball tournament. Only 5 times in the past 16 years have two teams from the top four played each other, and only once (1996) have the top two seeds played each other. That fact shouldn’t come as a surprise when you remember that never have all four #1 seeds made the Final Four of the men’s basketball tournament.

The Coaches’ Poll, Harris Poll, and computer poll average of the BCS make up the de facto I-A football playoff committee, whose job is simply to pick out #1 and #2. Records and stats alone aren’t enough, because teams from different conferences play completely different schedules, and comparing conferences is difficult when so few of the elite of any of the conferences play each other.

Even the committees, who are made to be balanced and whose members watch more film than any pollsters do, can’t accurately guess the best four teams most years, much less the best two. Can the choice of the two teams that play for the championship really be made by people who may or may not watch film of more than 13 teams (the coaches), people who may or may not watch more than a game a weekend (the Harris Poll voters), and those who may or may not have hidden bias incomprehensible to common people (the computers)? Almost certainly not if you’re going for accuracy.


The National Championship Game

December 26, 2007

For the purposes of this essay, all years refer to the season, not the actual year the national title game was played. For example, the 1997 Sugar Bowl was the national championship game for the 1996 season, so it gets referred to as the 1996 national championship game.

There have been 15 “national championship” games, from the beginning of the Bowl Coalition in 1992 to the advent of the BCS National Championship Game last year. Three times in that span the game featured a #3 team due to the Big Ten and Pac 10 refusing to participate prior to the BCS in 1998: in 1994 it was #1 Nebraska vs. #3 Miami; in 1996 it was #1 FSU vs. #3 Florida; and in 1997 it was #2 Nebraska vs. #3 Tennessee.

Counting the upcoming game for this season, out of the 32 possible spots only 12 unique teams have participated in the national title game, in chronological order: Alabama, Miami, FSU, Nebraska, Florida, Tennessee, Virginia Tech, Oklahoma, Ohio State, LSU, USC, and Texas. Only 3 of those 12 teams have played in it once – Alabama, VT, and Texas – with LSU falling out of that category this year. The longest stretch of games featuring newcomers was 2002-05, with Ohio State, LSU, USC, and Texas playing, and for that matter winning, their first national title games in that period. Only 5 of the 16 games did not feature a team from the state of Florida: ’97, ’03, ’04, ’05, and ’07. That 4 of the 5 games have come in the last 5 years shouldn’t be a surprise given the decline of the marquee Florida teams under Ron Zook, Bobby Bowden, and Larry Coker.

The average score of the national title game is 36 – 18.47, for an average margin of victory of 17.53 points. Eight of the 15 games have been blowouts, meaning the margin of victory was greater than two scores (more than 16 points). We have never gone more that two years without a blowout, though there were three blowouts in a row in the Bowl Alliance years of 1995-97. Those were also the only instances of back-to-back blowouts. Only 6 of the 15 games had a final margin under 10 points.

The most points scored in a national title game was 62 by ’95 Nebraska, and the fewest was 2 by ’00 FSU.
Only 5 teams scored fewer than 30 points and won – ’93 FSU, ’94 Nebraska, ’98 Tennessee, ’00 Oklahoma, and ’03 LSU.

The BCS Era

The BCS ostensibly was created to pit #1 versus #2, but how good has it been at selecting those teams? In the first four years, #1 was a perfect 4-0, but since #1 is only 1-4. The average score is 34.22 – 18.89, with an average margin of victory of 15.33 points. Four of the nine games have been blowouts; four of the 5 non-blowouts had margins under 10 points. At least by that quick look, it seems that half the time it sets up a good game and half the time it doesn’t. It’s statistically a coin flip as to who will win, but trends suggest #2 LSU has an excellent chance against #1 Ohio State this year.

Only two conferences – the SEC (with Tennessee, LSU, and Florida) and the Big 12 (with Nebraska, Oklahoma, and Texas), have put two or more different teams in the BCS title game. Since Miami in 2001-02 was a Big East member, it counts as a Big East team rather than an ACC team; no current Big East team has made the title game.

Almost each year, a controversy has arisen. Almost each year, a tweak to the system has been made to correct that error. However, since the BCS is a reactive institution, it will almost certainly never produce a perfectly agreed upon year unless two BCS conference champions go undefeated. To wit:

  • 1998: The problem was #3 Kansas State not being picked for a BCS bowl. The “Kansas State Rule” was enacted giving an automatic bid to #3 (or #4 if #3 is an auto-qualifier).
  • 1999: The problem was undefeated Tulane and #6 Kansas State (again) being passed over for #8 Michigan.
  • 2000: The problem was having one undefeated team and three 10-1 teams. Two (Miami and Washington) had legitimate cases for the #2 spot, but it went to the third 10-1 team FSU, who had lost to Miami.
  • 2001: Nebraska ends up #2 despite being blown out by Colorado in the Big 12 title game and finishing #4 in the human polls. The #2 team in the human polls, one-loss Oregon, was 4th in the final regular season BCS poll behind two-loss Colorado. Oregon would blow out Colorado in the Fiesta Bowl.
  • 2002: No controversy in selecting Miami and Ohio State, the two lone undefeated teams. There was that pass interference call though.
  • 2003: Three one-loss teams from BCS leagues and no undefeated teams for the first time since 1996. USC was #1 in the human polls, but the BCS title game ended up with the other two one-loss teams, LSU and Oklahoma. OU had not won the Big 12, having lost in the Big 12 title game. Three non-BCS teams – Miami (OH), Boise State, and TCU – finished with one loss but none got a BCS bid. LSU and USC split the title, something that the BCS was supposed to prevent from happening.
  • 2004: In a fit of irony, after a year with no undefeated teams there were no less than 5 undefeated teams: USC, Oklahoma, Auburn, Utah, and Boise State, with the first two playing for the title. This was also the year of Mack Brown campaigning against Cal, but it did have some good with Utah being the first BCS Buster.
  • 2005: Like 2002, the system got lucky with only two undefeated teams – USC and Texas – who played a classic in the Rose Bowl. Oregon did get shafted again thanks to Notre Dame’s automatic bid if it finishes in the top 8 in the BCS standings, an increasingly bad idea.
  • 2006: Florida or Michigan? Michigan or Florida? And what about Wisconsin, Louisville, and Boise State? Florida’s trouncing of the Buckeyes led a lot of people to pontificate that the system “got it right.” However, that’s a smokescreen since the computers were evenly split between UF and UM, meaning it was the human polls that decided the game’s participants. The “system” that got it right was no different than the one in the old Bowl Coalition/Alliance days.
  • 2007: Huge upsets every week. No undefeated teams except Hawaii, a team that played a historically bad schedule. Ohio State is the only one-loss BCS conference champion, but its schedule was weaker than some others and it lost a lot of credibility in the desert the year before. The computers say LSU and VT should be in, but since the humans that make the system marginalized the computers a couple years back, it’s LSU and OSU. LSU destroyed VT in an early season game, but what about Oklahoma, Georgia, USC, West Virginia, and Hawaii?

Picking a #1 and #2 can sometimes be easy (2002, 2005), and sometimes extremely difficult (2000, 2003, 2004, 2006, 2007). The BCS is an inherently conflicted system – it includes computer polls to add an element of impartiality unswayed by tradition and regional bias, but after the human polls were overruled by the computer polls several times, the people running the system gave the humans nearly unsurmountable power. It’s a wonder the BCS still includes them on a regular basis; at this point they can only make an impact if the Coaches’ and Harris polls differ on who’s #2.

Until and unless we see a playoff, it’s unlikely we’ll see a controversy-free, true national championship game. And even then, there will still be plenty of whiners. Such is the existence of college football.


A Brief History of the Post-Season in America

December 18, 2007

I am going to be doing a haphazardly-published series on playoffs and college football. I would prefer to see a playoff decide a champion rather than polls,  for the record. This is the first in the series.

The longest-running post-season event in major American professional sports is baseball’s World Series. The first one was in 1903, when the National League and American League, then two completely separate entities, organized under the mantle of Major League Baseball. Each league’s champion played a best-of-9 series to determine the overall champion. The necessity for this playoff was the fact that AL and NL teams didn’t play each other during the regular season. After a dispute canceled the series in 1904, it returned in 1905 and would be played every year since except the strike-shortened 1994 season.

The next-oldest professional post-season event is the NHL Playoffs, as the league has had some sort of playoff determining a champion every year since its inception in 1917. The lone except is 1920, when the Ottawa Senators won both halves of the regular season and the league decided a playoff would be unnecessary. The league’s regular season system was strange up until that point; read the Wikipedia page linked to above for details.

After that, you have the NFL playoffs. The NFL was founded in 1920, but from its founding until 1932, no playoffs were held. From 1920 to 1923, the champion was selected by the owners voting at the annual owners meeting. From 1924 to 1932, the team with the highest winning percentage won the championship as the teams all played different numbers of games. In 1932, the Chicago Bears and Portsmouth Spartans tied for the lead in winning percentage, so a one game playoff was thrown together hastily to determine a champion.

Responding to fan interest in the game, the NFL split itself into two divisions (East and West) in 1933. From then on, playoff games were held if necessary as tiebreakers and then the east and west division winners played in a championship game. A consistent tournament to determine who got to play in the NFL title game was not held until 1967 when the league expanded to 16 teams. The first Super Bowl was played in 1967 as a championship game between the NFL and AFL winners, and it became the NFL championship game after the AFL/NFL merger in 1970.

The NBA playoffs have occurred every year since the precursor BAA league was founded in 1947. The league had east and west divisions from the start, and at least the top three teams from each division have appeared in the playoffs every year. Perhaps the relatively late founding of the NBA allowed it to observe the other leagues and set up a proper playoff tournament from the start.

The NCAA

The precursor to what we know as the NCAA was the Intercollegiate Athletic Association of the United States (IAAUS). It was founded by Teddy Roosevelt after his son broke his collarbone playing football at Harvard while running the offense known as the flying wedge. The idea was to have a governing body setting rules for collegiate sports to cut back on the injuries and yes, deaths, being experienced by college athletes. The organization took the name NCAA in 1910.

The NCAA at first was a a discussion group and rule-setting club until 1921, when the first NCAA championship was officially recognized: the National Collegiate Track and Field Championships won by Illinios. In the years since, it has come to sponsor 44 women’s, 41 men’s, and 3 coed championships.

The only sanctioned sport without a recognized champion is Division I-A football, a.k.a. the Football Bowl Subdivision. Only in the sport of football is a relevant distinction made between multiple parts of Division I.

Bowl Games

As we all know, I-A football uses a system of bowl games as its post-season fare. They were originally a method of attracting tourists for the areas in which they were played, and they were scheduled around the new year to give fans time to plan trips and travel to the site.

The first bowl game was the “Rose Bowl” of 1902. I put it in quotes because while it was put on by the  Tournament of Roses, it was called the “Tournament East-West Football Game.” It featured a dominant Michigan team versus a decent Stanford team, and it ended in the third quarter when Stanford quit while trailing 49-0. The Tournament of Roses was so scarred by the blowout, it wouldn’t sponsor a football game again until 1916. The game wouldn’t take on the name “Rose Bowl” until 1923 when the stadium known as the Rose Bowl was completed and hosted the game. Fun fact: it wasn’t actually a bowl stadium at the time, but a horseshoe stadium.

The Rose Bowl pitted a team from the Pacific Coast Conference (the predecessor to the Pac 10) and an eastern US team up until 1947, when the champions of what are now the Pac 10 and Big Ten became the annual contestants. It was the only major bowl until 1930, and the oldest surviving bowl games besides the Rose are the Sugar, Orange, and Sun Bowls, all founded in 1935. Besides those, the Cotton (1937), Gator (1946), and Florida Citrus (1947) are the only bowls that have been held consistently for more than 50 years. The first major bowl with a title sponsor was the (in)famous Poulan Weed-Eater Independence Bowl, operating under that name from 1990-1996.

Football Playoffs

Up until 1973, the NCAA had two divisions – the University Division, roughly football’s Division I, and the College Division, roughly football’s Divisions II and III. In 1973, the I-II-III system was set up, and Divisions II and III immediately began holding playoff tournaments for football. Division I did not, however, set up a playoff tournament thanks to the tradition of the bowls and polls.

In 1978, the NCAA partitioned Division I into three divisions: I-A for the principal football schools, I-AA for the lesser football schools, and I-AAA for the Division I schools that did not play football. Division I-AA from its inception has had some sort of playoff tournament, probably because none of its participating schools would be bowl material. This fact confirms that the real reason I-A has no playoffs is due to the bowls; every other excuse given (demands on players, the sanctity of the regular season, etc.) is secondary to the bowl games. The NCAA must have realized in the late ’70s that teams with no hope of making a bowl were playing meaningless seasons, so a separate division with playoffs included was created. No other reason for the existence of Division I subdivisions makes sense.

The Polls

The absence of an officially recognized champion of major college football naturally created a power vacuum of sorts that many organizations have been eager to fill in. The NCAA on its website keeps a record of every major poll service’s pick for national champion dating back to 1869. No polls existed at that time, but poll services such as Richard Billingsley, the National Championship Foundation, and Parke Davis have gone back and somehow come up with champs for all those years.

The two oldest surviving polls are the AP poll and the Coaches’ Poll, the latter initially being published by UPI before being taken over by the USA Today in 1991. The AP poll began in 1936, but it didn’t release a post-bowl season poll until 1965, and it wouldn’t do so on a consistent basis until 1968. The Coaches’ poll, for its part, began in 1950 and didn’t release post-bowl season polls until 1974.

Over time, mathematicians began taking cracks at making polls since human-based opinion polls can be influenced by bias, ignorance, and misinformation. The BCS has used a variety of them over its decade of existence, but the ones used today are Jeff Sagarin’s ELO-CHESS, Richard Billingsley, Anderson and Hester, Kenneth Massey, Peter Wolfe, and the Wes Colley Matrix. This group was chosen because they all do not rely on margin of victory.

One final human poll has come to prominence, the Harris Interactive Poll, after the AP pulled out of the BCS formula in 2005. The poll is made of former players, coaches, administrators, and current and former media members selected at random from a pool of candidates. Harris Interactive is a market research firm that specializes in opinion polls.

A National Title Game

For the most part, national champions for Division I/I-A football since 1950 are recognized to be the final #1 in the AP and Coaches’ Polls. That’s fine when they agree with each other, but what if they disagreed? You’d get two teams with equally legitimate claims at a title. How could one convince both
to vote for the same #1? Why, by having a national title game, of course.

The first attempt at creating a national title game was the formation of the Bowl Coalition. It consisted of the SEC, Big 8, SWC, ACC, and Big East partnering with the Orange, Sugar, Fiesta, and Cotton Bowls. The idea was that the site of the national title game would rotate among the four bowls, and it’d take the #1 and #2-ranked teams from the AP and play them against each other. This setup might require the breaking of tie-ins of conference champions to their traditional bowls, but the Coalition agreement made that possible. It lasted from 1992-94.

You may notice the absence of the Pac 10, Big Ten, and Rose Bowl. They did not participate in the Coalition, and they kept their traditional arrangements with each other. This resulted in 1994 of  #1 Nebraska playing #3 Miami in the “national title game” while #2 Penn State played in the Rose Bowl.

Following the formation of the Big 12, the Bowl Coalition was replaced by the Bowl Alliance. It consisted of the SEC, Big 12, ACC, and Big East along with the Orange, Sugar, and Fiesta Bowls. The purpose and goal was the same as the Coalition’s, but the absence of the Pac 10, Big Ten, and Rose Bowl created the same problem. Twice a #1 vs. #3 game was forced to occur in the so-called national title game. It lasted from 1995-97.

In 1998, the three stubborn laggards finally came aboard to form the Bowl Championship Series. The goal was the same – have #1 and #2 play each other – only this time it would use the AP poll, Coaches’ Poll, and an index of computer polls to determine #1 and #2. Initially, strength of schedule and losses were their own categories, and in 2002 a quality win category was included as well.

By 2002, the BCS purged all computer models that included margin of victory to discourage teams from running up the score. However, it’s impossible to keep the human element from considering it, and margin of victory definitely plays a part in the human-generated polls. In 2004, it was streamlined to include just the human and computer polls with no other categories. In 2005, the Harris Poll replaced the AP poll. In 2006, the system was tweaked to deemphasize the computers, and the result has been that the human polls control the BCS formula almost completely. Only a huge anomaly in the computer element could override a unanimous human selection. That situation creates a Catch-22, since such an anomaly would likely cause an outrage, probably leading to further deemphasizing of the computers.

A Brief Timeline of the Post-Season in America

1902: The Tournament East-West Football Game

1903: The first World Series

1916: First annual Rose Bowl game

1917: NHL formed; first NHL playoffs

1921: First officially recognized NCAA championship

1932: First NFL Championship Game

1935: First annual Orange Bowl, Sugar Bowl, and Sun Bowl

1936: First AP Football Poll

1937: First annual Cotton Bowl

1939: First NCAA men’s basketball tournament, consisted of 8 teams

1946: First annual Gator Bowl

1947: First annual Florida Citrus Bowl

1947: Advent of NBA precursor; first annual NBA basketball playoffs

1950: First Football Coaches’ Poll

1965: First post-bowl season AP poll

1967: First Super Bowl

1968: First annual post-bowl season AP poll

1971: First annual Fiesta Bowl

1973: NCAA creates Divisions I, II, III; first annual D-II and D-III football playoffs

1974: First annual post-bowl season Coaches’ Poll

1978: NCAA creates Div. I-AA; first annual I-AA football playoffs

1984: NBA playoffs expands to current 16-team format

1985: NCAA men’s basketball tournament expands to 64 teams

1990: NFL playoffs expands to current amount of 12 teams

1992: Bowl Coalition formed

1992: SEC expands to 12 teams, plays first ever football conference championship game

1993: NHL playoffs expand to current format

1994: MLB institutes the wild card; World Series canceled due to strike

1995: Bowl Alliance formed

1996: Big 12 formed; first Big 12 Championship Game

1998: BCS formed

2001: NCAA men’s basketball tournament adds 65th team, play-in game

2002: NFL reorganizes to 8 divisions, drops one wild card per conference to keep playoffs at 12

2003: Split national title between LSU and USC; BCS formula completely rewritten

2004: NASCAR implements its “Chase for the Cup” quasi-playoff system

2005: ACC expands to 12 teams; first ACC Championship Game

2005: AP Poll drops out of BCS formula, Harris Poll is formed to replace it


It’s Official: Tebow Has Won the Heisman

December 8, 2007

The votes are in, the announcement’s been made, and Tim Tebow is your 2007 Heisman Trophy winner. I honestly thought during the preseason that he had zero chance of winning the award; after all, he’s a sophomore, and there were other, older guys out there like John David Booty and Mike Hart who were poised to put up excellent numbers on high-profile teams.

As those teams and players fell one by one, Tebow rose to the top like the cream always does over the course of a college football season. His team may have faltered some, but that was obviously due to the defense, not offense. Then, he started setting records (first 20/20 guy, longest streak of throwing and rushing for a TD in the same game) and his case became more compelling. After he scored 7 TDs against South Carolina, I thought it could happen. After wiping the field with the same FSU defense that knocked Matt Ryan out of the Heisman race, I figured he had it in the bag.

If you want to see (or see again, as the case may be) video from the ceremony, hit up ESPN.com. For now, enjoy watching all 51 of his touchdowns and reveling in the fact that you live in historic days.

It’s great to be a Florida Gator.

As a final note, Florida now has more Heisman winners than both FSU and Miami, and honestly, should Gino Torretta even count? One more stiff arm statuette, and it’s as many as they have combined.

Only 4 schools have more Heisman winners than Florida does – Notre Dame, Ohio State, and USC each with 7, and Oklahoma with 4. Florida is tied for fifth with Army, Michigan, and Nebraska. When you compare Florida’s tradition with the rest of those schools (about 30 good years out of 101), that’s not too shabby at all.


Trophy Day

December 8, 2007

Three trophies for Tim Tebow the Cheat!

Tonight is the Heisman Trophy ceremony, and by all accounts Tim Tebow will win it. He’s already won the Davey O’Brien and the Maxwell Award, and for him to take home the Heisman would be a great ending to an amazing season. No one deserves it more than Tim does, and I fully expect him to make history tonight.

Unfortunately, I am not in a place with cable at the moment, so I will not be able to watch the ceremony tonight. I have it recording on my DVR back home, so I will get to see it when I get back to Florida. That’s fine though, since most of the show is fluff-ridden biopics of each candidate. I can live without yet another ESPN piece on Colt Brennan, considering there’s been one nearly every week on College GameDay.

Congrats, Tim. You’ve earned it.


SSOS Awards

December 7, 2007

Statistical Strength of Schedule (SSOS) has become a weekly feature of mine, and you can read the rationale and about how it’s calculated here.

I’ve got the final SSOS calculated, but I’m not done with the writeup and charts and all. In the meantime, enjoy these awards I just made up last night on an airplane. They’re based on the final numbers, which should be up sometime before Ragnarok.

The SSOS Champion: Best overall SSOS

WINNERS: Nebraska (team): 48.52 SSOS score; SEC (conference): 29.75 average rank

Huskers, even though you got torched constantly on defense, had a wildly inconsistent offense, and got your coach fired, at least you did it all against the nation’s toughest schedule.

The SEC showed just how tough it is by overcoming 10 games against I-AA opposition to win the conference battle comfortably over the Pac 10. No more whining about the SEC having weak out of conference opponents – the teams still graded out as having played the strongest schedules among the BCS conferences.

The SSOS Goat: Worst overall SSOS

WINNERS: Hawaii (team): 81.44 SSOS score; ACC (conference): 59.33 average rank

Hawaii, you’re a nice story and all with your BCS bid, but I hope you know it’s fraudulent with as easy of a schedule as you played. I know Michigan State pulled out of its game with you, but playing two teams below I-A will get you this award nearly every time. At least you play Florida next year.

ACC, by now you know that no one cares about your conference when FSU and Miami are having bad years. The attendance in Jacksonville a week ago proved that. However, your attempt to look better by playing the weakest overall schedule by far didn’t work because your teams really are that bad and that boring. Please try to play a real slate in the future, which means finding strength in your non-conference games because you sure won’t find it inside your conference.

Mr. Bland Award: For scheduling mediocrity

WINNERS: Wisconsin (team): ranked 60th; Big Ten (conference)

Wisconsin, you finished exactly in the middle. There were 59 teams ahead of you, and 59 teams behind you. That is the perfect embodiment of middle-of-the-road. It makes sense considering your conference.

Big Ten, you finished with all of your teams in the second and third quintiles. No one particularly exerted itself, but no one took it easy either. It’s an interesting strategy, albeit one that gets you ranked second-to-last among the BCS conferences. Ohio State dropping Youngstown State picking up USC certainly helps, but don’t let the Buckeyes’ ambition steer you away from your dream of blandness. It suits you well.

Go Getter Award: Largest gap between the conference’s first and second place

WINNER: Syracuse

Syracuse, you win this one for having the toughest schedule in your conference and for finishing with the biggest gap between you and the second place team (Pitt) at 16 spots. Way to put the rest of your conference to shame. Perhaps this is why Greg Robinson still has a job.

Deadweight Award: Worst schedules in each conference

WINNERS: Georgia Tech, Kansas, UConn, Northwestern, USC, Arkansas

If not for you all, your conference’s scheduling marks would look a lot better. I hope you’re happy. Readers, please note that there are two teams here that made BCS bowls. I’m just saying.

Anchor Award: Worst schedule for a team in a BCS conference

WINNER: Kansas (112 rank, 74.92 score)

Kansas, you’re the only team in the country that played in a BCS conference and still managed to have a schedule in the bottom 20%. That’s not easy to do. Sure, it just so happened you missed Oklahoma, Texas, and Texas Tech in your conference rotation, but few teams went to the bakery for bigger cupcakes than you did non-conference. Put it this way: throw out your numbers and the Big 12 has the toughest overall schedule for a conference; with them, it drops to third. Of course, that schedule is probably the main reason Kansas is in a BCS bowl, so the Jayhawks will probably make this a habit.

Deposed Nigerian Prince with an Email Account Award: Most fraudulent records

WINNERS: Boise State, Boston College, BYU, Hawaii, Kansas, UCF 

These are the teams who won at least 10 games with a schedule in the bottom two quintiles. Try to play some more notable teams in the future, will ya? Readers, please note that there are two teams here that made BCS bowls. I’m just saying.