Memorable Quotations for Bulldog Nation

by T Kyle King on Mar 9, 2007 7:17 AM EST in Football

I go back and forth over which is the better recent observation.

On the one hand, I think it may be this:

These issues of scheduling 1-AA opponents and ridiculous numbers of home games deserve to be kicked upstairs to the NCAA, or would if the NCAA weren't so toothless and gun-shy. Member institutions' athletic directors have a vested interest in maximizing revenue at their institutions. At big schools, it means scheduling every home game possible. For what it's worth, Michigan clears a good $4 million from every home game. Appalachian State will also walk away with a guaranteed $400K, demonstrating that it's also in the best financial interests of a small school to take the paycheck and the pounding. It's a win-win from an AD's perspective, but the loser is college football. The NFL long ago recognized that selling the league as a whole was more important than any one team. Obviously, there are many ways in which the NFL model is incompatible with the college game, but the NCAA needs to recognize that real competition is integral to growing the brand. That isn't helped by 8-home-game seasons and 1-AA opponents. Simply capping the number of home games at seven would be a good start. In a better world, 1-A teams would be banned from putting 1-AA cannon fodder on the schedule, but - barring that - wins over 1-AA opponents shouldn't count for the BCS or towards bowl eligibility.

I also want a pony.

On the other hand, I also think it may be this:

[T]here ain't exactly a whole lot of difference between the RPI and math behind the BCS. . . .

Before you can "settle it on the field", as playoff advocates like to bray, you've got to pick who gets to show up in the first place. That's why I continue to argue that there are flaws in any playoff format that have to be addressed before anointing it as new and improved over the current state of affairs in D-1 football.

And the bigger you make that playoff field, the more you magnify the flaws.

Preach it, gentlemen!

Go 'Dawgs!

Tweet Comment 15 comments | 0 recs |

Do you like this story?

Comments

I'm totally with Senator Blutarsky

Playoff advocates love to talk about how George Mason's Final Four run last year was "proof" that playoffs are good, since even a bubble team could do that.

On the contrary, I think their tournament run was an argument AGAINST such playoffs, since their regular season of suspect quality was reduced to minimal importance. The regular season should always count a whoooole lot. God bless the bowl system.

by Jeff on Mar 9, 2007 10:02 AM EST reply actions

Gotta disagree with Blutarsky on that

"And the bigger you make that playoff field, the more you magnify the flaws." I disagree, but it takes some defining to get there.

When people say "settle it on the field", it is absolutely true that you need to decide who gets to show up in the first place.

But we have to step back first. Is the argument over the best way to crown a champion, or the best way to ensure that the champion of a particular sport is "the best team"? If the former, I'd argue that Sen. Blutarsky's comment is exactly backwards. If the latter, I'd argue that we're tilting at windmills.

"The Best Team" is impossible to discern with factual accuracy. It is, by definition, the result of opinion. A "Champion" need not be "The Best Team" (though one may have the opinion that a champion of a particular contest is effectively the best). A Champion is merely a team that wins the contest. Matt Leinart may have been right after the 2006 Rose Bowl - USC might have been the best team in the nation. But they weren't the Champions. The St. Louis Cardinals might not have been the best team in baseball last year, but they were World Series Champions. All they did was win the contest as it was set up - and which other teams involved had the opportunity to win just as freely as St. Louis. The last seeded Pittsburgh Steelers might not have been "the best team" in the NFL in 2005-6, but they were Super Bowl Champions.

I believe that at some point folks need to wrap their minds around the ideas that "the best team" and the Champion are not necessarily the same. After a contest (be it a playoff system, tournament, regular season, etc), we have a factual assertion as to who is the Champion - as defined by whatever set of rules are in place that each team has an equal chance at attaining. We might have an inference as to "the Best Team", but we do not know this as a fact. The Best of anything is necessarily opinion.

So what this directly leads us to is this: a playoff system is no better at determining who is "the best team" at the end of the year than any other system. A Bowl selection system is not worse or better than a playoff at determining who is "the best team". No sports have a way of determining "the best team" because such is purely the result of opinion.

Now, if the discussion is on determining a "champion", this is an entirely different argument. And "better" also has nothing to do with it. In determining a "champion", the question goes to "fairness" or "certainty", not necessarily "better".

Hopefully I haven't confused everyone yet (though I think I've confused myself)...

Let's compare systems of selecting teams to compete for the title of champion.

In college football, a formula of polls and computers selects 2 teams out of 119 to compete for the title. No objective qualifications are stated. Neither team necessarily has to win its conference, neither team necessarily has to remain undefeated. Arguably, a team with two losses could be selected to play for the title ahead of a team with no losses. In terms of "fairness" and "certainty", which I readily admit are areas where we deal in opinion, the issue of subjectivity in selection may raise concerns. At the beginning of each college football season, a large percentage of teams, before snapping a single ball, do not have an opportunity to compete for the title of champion. To precisely quantify the teams who, before the season starts, have no chance at the title is similarly impossible, since the rules are subjective. But it's possible that there are several teams capable and qualified who "should" have a chance at competing for a title of champion were excluded. Not competing for the title of "best team", but champion. There are dozens of schools which, if they won every single game they played, would not compete for a title.

Now look at pro football. Selection to the playoff system is completely objective - based on records. It might not be perfectly fair, since an excluded team in one particular conference may have a better record than an included team in the other conference, and the champion of a division may have a worse record than an excluded team. But going into the season, every team knows what the rules are for inclusion in the playoffs and every team has an equal opportunity of competing for the title. If you don't think equaly opportunity is akin to fairness, that's a different debate entirely. From a standpoint of "who is the best", nobody knows. But in terms of determining a champion, every team has the same opportunities, and 32 out of 32 teams are on equal footing - meaning we can be certain that "should" have a chance to compete for the title did.

Now consider college basketball. At the beginning of the season, out of the 336 teams in Division I, only 11 independent teams are left to rely on subjectivity in whether they are able to compete for a title. 325 teams know that if they were to win their conference tournament (and for some conferences, that requires winning enough in the regular season to qualify for the conference tournament), they would be able to compete for the title of champion. 96.7% of the teams know, objectively, what they need to do to win the title of champion. Only those 11 teams, 3.3% fit into a category of "no matter what they do, they still might not have a chance at winning the title."

Now, all of this is the preseason to postseason comparison. Now let's compare simply postseason methods.

In college football the BCS formula chooses 2 teams to contest the title. These 2 teams may have as good a claim as 0 others, 5 others, 10 others, or 15 others. Because of varying schedules, imperfect comparisons and small sample sizes, it is impossible to determine whether those 2 teams are the proper teams to compete for a title. In some years, the 2 selected might be the proper choices, while in other years, that might not be the case. Numerically, 117 out of 119 teams, 98.3% of the teams simply have no chance at competing for a title. Now, clearly not all of those 117 "should" be a proper choice for a team that can compete for a title. But there is a relatively high level of uncertainty that one or more of those 117 excluded teams couldn't beat the selected teams in a title game. Almost assuredly that is the case. Of the 117 excluded teams, several of them could have the ability to win a title if given the opportunity.

Now consider college basketball. The selection committee chooses 65 teams out of 336 based partially on objective factors and partially on subjective factors. Once the tournament field is selected, 80.7% of the teams in Division I have no chance to win the title. Out of those 271 excluded teams, there might be a couple with the ability to win the title of champion but without the opportunity. But because of the size of the field, the number of excluded potential champions is far smaller than in college football.

Again, because of subjectivity, it's impossible to quantify. But how many college football teams could have competed in the BCS title game in a given year, and depending on the day, could have claimed the title of champion? 5? 10? Had this year been USC vs. Florida, is it reasonable to think USC could have won? I think so. That means USC is an excluded potential champion. What about LSU vs. Ohio State? Could LSU have beaten Ohio State? Possibly. They too could be an excluded potential champion. Louisville? Boise State? Auburn beat Florida earlier in the season. What about them?

Now, the other issue is the one-off nature of college football's title crowning. There are probably several more "excluded potential champions" because all those excluded teams would have to do is win a single game.

Compare all of that to college basketball. Upon selection, once you disregard several automatic bids, approximately "the best" (subjective, I realize) 40-50 teams are included in the tournament. Each of those teams has an equal opportunity to claim a title - by winning 6 games (yes, seeding might make things more or less difficult, but the opportunity based solely on winning and losing is the same). The "excluded potential champion" starts at the bubble teams. This is a major distinction from college football. In college basketball, the "best" "excluded potential champion" is approximately considered around the 40th or 50th best team in the country before the tournament. In college football, the "best" "excluded potential champion" is considered the 3rd best team in the country. There is a large gap there. Further, the structure of the tournament makes the pool of "excluded potential champions" even smaller: consider, an "excluded potential champion" wouldn't have to win just one game (where they might get lucky), they'd have to win 6 to win the title. So an "excluded potential champion" would have to be (a) severly underestimated by a selection committee (40+ places ranked behind the top contenders, rather than 1 place ranked behind top contenders) and (b) able to win a series of games, not just one. Now, that's not to say college basketball has a perfect system. It's not. George Mason last year was one of the last teams admitted into the tournament and came closer than 61 other teams to winning the title. But in terms of objectivity in selection, college basketball is far moreso than college football. In terms of certainty that a "potential champion" is not excluded from having an opportunity to compete for the title, college basketball is far more certain.

Now, none of that has to do with ensuring that "the best team" wins the title. Though I think it's probably easy to get confuse by what I'm writing, because there's a lot of opinion getting mixed in here.

Here are the basics of what I think:

"The Best Team" does not necessarily equal "Champion.
There is no reasonable way of perfectly determining "The Best Team" in any sport, aside from a large number of games against each opponent with all externalities removed.
In terms of determining a "champion", each team having objective knowledge of what is necessary to earn such a title, and each team having an equal opportunity is the best system of determining that.
Where subjectivity enters the equation and teams are excluded from the opportunity to earn a title, including more teams (if not all) rather than fewer is a more certain way determining that the champion was not crowned so over a complete set of potential champions. (This is where I confuse myself because I'm trying to be precise...)

I should've written this over at my place rather than take up your space.

by LD on Mar 9, 2007 10:40 AM EST reply actions

The perfect is the enemy of the good.

ld, I'm impressed. Let me respond to your basics, if I may:

# "The Best Team" does not necessarily equal "Champion.

-- I agree that that is often the case, although it's exacerbated in single elimination playoff formats like March Madness.

# There is no reasonable way of perfectly determining "The Best Team" in any sport, aside from a large number of games against each opponent with all externalities removed.

-- Yep. Although I think this again implies how much single elimination tourneys work against that.

# In terms of determining a "champion", each team having objective knowledge of what is necessary to earn such a title, and each team having an equal opportunity is the best system of determining that.

-- Absolutely. It's one area where the pro format is superior to the format for either college football or basketball. Unfortunately in the realm of college football, until Notre Dame's situation is radically changed (i.e., it's forced to play in a conference and not given any special preferences to play in the BCS), this isn't likely to improve.

# Where subjectivity enters the equation and teams are excluded from the opportunity to earn a title, including more teams (if not all) rather than fewer is a more certain way determining that the champion was not crowned so over a complete set of potential champions.

-- Here's where we part ways. I don't care about "potential champions". The purpose - and the greatness - of college football's regular season is to weed out potential and leave us with a small group of teams that we can agree were the best for that season. The BCS tries to balance that with the goal of funnelling the two best teams (subjectively speaking, of course) into a direct match. It's not a perfect scenario, but it's not bad... although I don't doubt that it could be improved.

I'm not an anti-playoff purist as Kyle is. I just think that before we head down the playoff road, we need to make sure that we aren't destroying what makes the D-1 regular season so special. And one thing that I'm very confident in is in saying that the larger you make the post season, the smaller you make the regular season. No thanks.

by Senator Blutarsky on Mar 9, 2007 2:26 PM EST up reply actions

I think I can clarify my position a bit more...

Senator, I do think we're pretty close on most of the big issues, but I should probably be a little more open about my own playoff thoughts.

I'm not pro-playoff, I'm just anti the current system, sort of. I don't think a playoff would solve much of anything in college football because there are too many teams, not enough games, and there'd still be subjectivity involved in a college football playoff.

I'm simply pro-objectivity. I actually think that every single team in the nation should know, going into the season, what it takes to win a title. No style points, no reliance on strength of schedule that's beyond the control of teams. Simply put, if you win all your games - you are a National Champion. If that means multiple titles conferred, so be it. If that means Boise State can claim a title this year, I'm OK with that.

Here's where I have problems: "The purpose - and the greatness - of college football's regular season is to weed out potential and leave us with a small group of teams that we can agree were the best for that season." We can't all agree on a small group, especially if the maximum size of that group is 2. When teams play just 12 games, there is too small a smaple size. Further, with conference schedules and teams choosing to play regional opponents, the amount of knowledge we have about multiple teams could be woefully inadequate (for example, take this year's OSU team). With every additional game, we might learn more about teams, but the schedule prevents us from learning too much. Just last year, I think 8 teams could arguably have won a national title had they been put in the game and had the right matchup (Florida, Ohio State, USC, LSU, Boise State, Louisville, Michigan, Wisconsin and probably more). Had voters believed all along that certain of those teams "deserved" it, any of them could've ended up in a title game based upon human polling - and then won a title.

I don't think a playoff would solve the problem. I do think that setting strict rules that every team knows ahead of time might. Of my list above, USC and LSU lost twice. Both could be considered among the best teams in the country. If either of those teams played anyone else in the country 10 times, they'd probably win many more than their opponent. They're good enough to have been considered as a title contender if the rules are purely subjective. In fact, I'd argue that if we're choosing a title game matchup based upon subjectivity, both of those teams should be considered. That's why I think it matters that teams are "potential champions". I think it's possible that Louisville could have been as good as Florida, but we'll never know. Boise State didn't lose a game, so how can we know that they would have lost a game? Yet both of those teams were outside of the discussion.

Let me give another hypothetical. Say in College Football there are 6 teams we all agree on as elite. Some years there might be more, some years fewer. This past year I think around 8, the previous year probably 2-4, the year before that 6. Say in college basketball there are 18 teams that we all agree as elite (smaller rosters, more programs are the reasons for increasing the numbers). In college football, two/thirds of that elite subset are excluded from a title. In college basketball, 100% of that elite subset is included. Now, in college football, there's no "filler" (teams outside of that elite subset) that could cost an elite team a title, but the cost is that several deserving teams just don't get a chance. Meanwhile, in the college basketball tournament, there are 36-37 filler teams that might knock off one of the elite 18, but at least all of those 18 had a chance on the court.

Now, does that mean that college football should emulate basketball? I don't think so. Mainly, because I don't think that's possible. They can't play enough games for us to truly know who should be included and excluded from a tournament. That said, I think the current system in place excludes too often deserving teams capable of claiming a title, and it does so through subjectivity.

I'm not sure of the best system, but I know it'd remove subjectivity from the equation and all teams would have an equal chance at a title.

by LD on Mar 9, 2007 3:04 PM EST up reply actions

Well, maybe just a pinch of subjectivity...

ld, I like the idea of promoting as much objectivity in the quest for a champ as possible. I'm just not sure the best system would completely eliminate all traces of subjectivity.

Here's what I mean. If I could exercise Godlike powers over college football (hah! I can't even get my daughters to pay attention to me anymore), I'd limit D-1 teams to eight ten team conferences and make them play a nine game conference schedule and three games against other teams in the Big 80. The conference winners would then go to the playoffs and the bowls could have the pick of the rest.

Sounds nice and objective, right? There's one problem - the chance that a team that qualifies for the playoffs doesn't take a non-conference game too seriously. How to fix that? By making seedings and opening round pairings (which would be at home for the top four seeds) dependent upon the poll/BCS rankings. You tank any of those games and you risk a lower seeding in the playoffs against a tougher opponent on the road.

Of course, none of this is going to happen.

by Senator Blutarsky on Mar 9, 2007 4:19 PM EST up reply actions

I have problems with your problems.

"Here's where I have problems: "The purpose - and the greatness - of college football's regular season is to weed out potential and leave us with a small group of teams that we can agree were the best for that season." We can't all agree on a small group, especially if the maximum size of that group is 2. When teams play just 12 games, there is too small a smaple size. Further, with conference schedules and teams choosing to play regional opponents, the amount of knowledge we have about multiple teams could be woefully inadequate (for example, take this year's OSU team)..."

I agree that in many years, two is too small a group to judge the best. But in some years, like 2005, it's right. I'd rather see some flexibility built into the current BCS setup by letting three or four teams playoff for the MNC in years that that's needed than simply decree that eight or sixteen teams will head to the post season every year.

College football's greatest attribute is the meaningfulness of its regular season. I don't want to see it jeopardized for some bastardized version of March Madness.

by Senator Blutarsky on Mar 9, 2007 4:30 PM EST up reply actions

Exactly

The difficulty with inter-team and inter-conference comparisons in college football is that we have a fairly small sample size. Compounding this is the fact that we have a paucity of pairwise comparisons as most teams do not play each other. And even within a conference, not every team plays every conference member.

It is a simple statistical fact that we have less information about the relative performance of college football teams than about basketball teams, and that our analysis of football teams is therefore more prone to error.

It is not inaccurate, therefore, to say that those celebrating the sanctity of the regular college football season are privileging relatively ignorant assessments of inter-team performance.

by 34hawk on Mar 9, 2007 10:07 PM EST up reply actions

Is that important?

For the sake of argument, let's say I agree with you.

How does a single game elimination tourney improve that assessment?

by Senator Blutarsky on Mar 10, 2007 8:34 AM EST up reply actions

Size Matters

I was not trying to justify a single elimination tourney here. But if we have one, and if we want the best teams included, our lack of information regarding relative performance should lead us to favor a larger rather than a smaller tournament.

So I'm trying to agree with LD's thesis that you have it "exactly backwards" in claiming that "the bigger you make that playoff field, the more you magnify the flaws."

by 34hawk on Mar 10, 2007 3:49 PM EST up reply actions

Gotta disagree.

Implicit in my argument is the premise that in any given year there are a finite number of teams that emerge from the regular season worthy of playing for the MNC.

I hope I'm not guilty of putting words in his mouth, but I don't believe that ld disagrees with that. Rather, I believe his objection is that, because of the limitations of regular season scheduling, the BCS (and I, too, for that matter) is unduly restrictive in limiting that pool of eligible candidates to two schools. His point is that the BCS should cast a wider net to make sure that no deserving team is left out of the opportunity to play for the championship.

Assuming for the sake of argument that I accept all of that on its face, I still think my point is a valid one. Let's say that we use ld's universe of eight eligible teams this year as a departure point. What happens when you expand the playoffs beyond those eight? By definition, you've diluted the field with schools that aren't championship worthy. With sixteen, you've have an equal number of worthies and non-worthies. By the time you get to that field of 64 that I dread, ld's eight teams are outnumbered by a factor of 7 to 1! In this case, the more definitely isn't the merrier.

With a single elimination tourney, upsets are a given. The more non-worthies in the playoffs, the more likely that is to happen. And the more often a team that should be playing for the MNC is eliminated by a team that really shouldn't, the more the impact of the regular season is reduced. In my mind, that's a flaw.

by Senator Blutarsky on Mar 11, 2007 8:13 AM EDT up reply actions

We're making progress

Firs, two observations.

One, I'm not sure it's very accurate to assume that only a small handful of teams are championship worthy. How often, for example, do teams seeded third or lower when the Basketball championship? I think 3 and 4 seeds win the thing fairly regularly.

And two, even if we assume that only a small number are "championship" worthy, my point is that we don't know which teams these are, since we don't have much to go on.

But taking the 8 "champion worthy" teams case as a point of departure, I think the relevant comparison would be between:

The probability that a "worthy" team was defeated by an "unworthy" team that shouldn't have been included, versus
The probability that a "worthy" team was excluded because the list of invited teams was too small.

I am much more sympathetic to those teams that fall into category 2. This is partly because I'm not sure a team can be "unfairly" defeated by a lesser team. What does that mean? If a high seed can not defeat a low seed, it tells me that the high seed is overrated.

Certainly a team could be unlucky (say the starting QB gets the flu), but this can happen in the regular season as well as in the postseason. So why should we punish a team that loses "unfairly" in the regular season, while seeking to protect teams that lose "unfairly" in the post-season?

by 34hawk on Mar 11, 2007 3:48 PM EDT up reply actions

Or maybe the terms of the discussion have changed.

I notice that you have a post up in the diaries section here that contains a proposal for an 8 school, 10 game regular season playoff. How do you reconcile that with your statement above that "I'm not sure it's very accurate to assume that only a small handful of teams are championship worthy"?

As for your second point, when you write "I am much more sympathetic to those teams that fall into category 2", that's fine, but it seems that our discussion has evolved into one about personal preferences and the aesthetics of how you'd like a playoff to look. That's a long way from asserting as a matter of statistics that I have things "exactly backwards" when I write that the bigger the playoffs, the more harm is done to the regular season (among other problems).

By the way, I don't think your basketball analogy holds much water. Because of issues like depth, team size and early departure of top underclassman talent to the pros, the talent level among schools is likely to be much closer in basketball than it is in football from year to year...

by Senator Blutarsky on Mar 12, 2007 7:45 AM EDT up reply actions

Good catch

I've thought a lot about this issue since writing the diary you cite. I haven't necessarily changed my mind, but I've broadened the set of criteria that are relevant to my decision process. So it's less of a contradiction in need of resolution, and more of a reflection of things I learn from conversations like this one.

We're going to have to get really deep into the statistical weeds to sort through this discussion though, including a discussion of classical, or frequentist, versus Bayesian statistics.

On the surface, their is a statistical basis to your argument: the longer a playoff the less likely it is that any one team will win that playoff. Equally matched BCS Championship game teams have a 50% chance of winning the championship (in the same sense that a "fair" coin has a 50% chance of coming up heads. In a 64 team playoff, their probability of winning it all will approach 10% even under very generous assumptions about their probability of winning early matchups (i.e 92% likely to win first round, etc....).

You regard this as unfair to the best teams. I, on the other hand, regard the 50% probability of winning a championship in a BCS matchup as an unfair gift to two lucky teams at the expense of several other worthy teams.

I don't think we can know which are the "best" teams. So I would view an early round playoff loss as a good chance to update my perceptions of team quality, rather than as the imposition of an unfair system on a worthy champion. That's the Bayesian approach. We have our prior perceptions of quality (probability of winning). Then we play the game. And then we update our beliefs about who is and who isn't the better team.

Also, I would argue that the ability to win under the pressure of "win or go home" conditions is a characteristic that we should reward. In my mind, this quality is the essence of what it means to be a championship team.

I agree with you in regards to the differences between basketball and football team dynamics that make them essentially non-comparable. Fortunately we can look to the NFL and 1AA college playoffs for a relevant comparison, which is what I should have done. Basketball's on my brain for some unknown reason now!

by 34hawk on Mar 12, 2007 10:40 AM EDT up reply actions

Short edit to #4

4. Where subjectivity enters the equation and teams are excluded from the opportunity to earn a title, including more teams (if not all) rather than fewer is a more certain way determining that the champion was crowned so over a complete set of potential champions. (This is where I confuse myself because I'm trying to be precise...)

Delete the "Not".

by LD on Mar 9, 2007 11:29 AM EST reply actions

I don't want to interfere with . . .

. . . this great discussion, but, just to provide further grist for the mill, I thought I'd direct your attention to this comment thread over at E.S.P.N., for whatever it might be worth.

by T Kyle King on Mar 11, 2007 6:40 PM EDT reply actions