Michigan Museday Sanity Checks Postseason Ideas
In 1997, back when 7-point leads were comfortable and safeties were meant for hitting people while corners did the covering* undefeated and no-brainer No. 1 Michigan went to the Rose Bowl. That was pretty cool. We faced Washington State and Ryan Leaf back when he was Ryan Leaf and not Ryan Leaf, Woodson made that interception to stop the comeback keep Michigan in striking distance (wow I forgot that context), and woo forever.
Meanwhile the conferences that weren't the Big Ten and Pac Ten were into their third year of a "bowl coalition" to match up the best two teams possible. Undefeated kick-ball-in-OT-vs-Mizzou Nebraska went to that and beat the tar out of Tennessee. The AP declared the next day's Daily cover something to hang on your wall forever, the coaches gave Osborne his send-off gift, and it didn't matter that there was a co- because starting next year there would be the perfect championship system to determine an unquestionable champion…for 1997.
-----------------------------------
* This wasn't at all true unless you literally had Charles Woodson in your backfield. Dude should win an award for that or something.
-----------------------------------
Getting' Jiggy With It isn't working. This has been the BCS's problem since its inception. In 1998 it was 
the perfect system to pit the last big conference undefeateds against each other, but then it left out an undefeated minor conference team and arbitrarily selected one of several similar 1-loss teams to face unquestioned numero uno Tennessee. Every year there was at least some complaint they patched with an overreaction on the next one. Team A beat Team B beat Team C who got in? Overrate head-to-head and dump half the computers. Too influenced by pollsters? Let's get more computers. Teams running up the score? Dump the margin of victory. AP and Coaches No. 1 USC left out of a three-horse race? Overrate the polls. Undefeated SEC team left out of a three-horse race? Overrate schedule strength. Boise State keeps going undefeated by playing Wyoming 12 times? Autobid the little guys. Two Big Ten teams about to rematch? Oh the pollsters can jig the system. Wait the pollsters are jigging the system? Kick 'em out and get our own pollsters. Two SEC teams about to rematch? Dammit.
This is what a process looks like when it has no forethought. I could say the same about many playoff proposals. Every year there's a perfect system that would be perfect for that year if we had that system. What we should be asking for is a system that would be good enough every year.
Good enough is good enough. Math says if you found the best team in a 120-team league after 12-13 games of unbalanced schedules, you just got lucky. What we're shooting for here is something where only the homeriest homer of Domer will be claiming their team got duked. The last team in should have an ironclad case, were they to emerge victorious, to be the No. 1 overall team, but the first team out should not have a very good case to be given that chance.
Autobids are bad (for this). This includes conference champions, sorry. The championship games help clear things up by giving contenders an extra bellwether. However a two-division format means 8-4 teams can beat 12-0 teams they lost to the first time. The Big Longhorns Conference still technically exists. So does the Big East. Bowl tie-ins for conference champions are great and should stay but nothing should be automatic about a playoff.
The right process is some kind of playoff. I'd be fine if it just went back to bowl games and polls to determine the National Champion, but the game has gone national and there's money to be made.
The question is how many teams should be in it. The current system has two teams. The Plus-One proposal discussed by the conferences last year is basically four with some measure of flexibility. Brian wants six, which is the fewest that will accommodate undefeated mid-majors most years. Hinton proposed 10, which reasonably fits most of the good 2-loss teams.
What I'd like to do here is UFR the BCS years past and see which of these playoff systems, the BCS, a Plus-One, Brian's, or Doc Saturday's, would have been best.
1998: Slightly similar to this year, with one undisputed team on top, then lots of 1-loss teams to pick from. Four-teamer is #1 Tennessee (12-0), #2 FSU, #3 Kansas St, and #4 Ohio State. Six teamer includes #5 UCLA and #10 Tulane. Ten teams nets #6 Texas A&M with 2 losses, #7 Arizona with 1 loss, #8 Florida with 2 losses, and #9 Wisconsin with 1 loss (the Big Ten Champ). Ideally: Brian.
1999: The first obvious matchup of two undefeated BCS teams, #1 Florida State, and #2 Virginia Tech. Clear #3 Nebraska stands apart from a ton of 2-loss teams like Tennessee, Bama, Michigan, Wisconsin and MSU. 1-loss KSU is in there too. 10 teams works if you take Marshall over 3-loss Florida or Penn State. Ideally: BCS
2000: #1 Oklahoma, #2 Florida State, who lost to #3 Miami, who lost to #4 Washington. #5 V-Tech, and #6 Oregon State also had 1 loss each. After that is a lot of 2-loss BCS teams. The BCS system generated all sorts of controversy for teams 2-4 being mostly indiscernible, and lo more overreactive rules were written into the BCS codec. Ideally: Brian.
2001: Another year where 1 is clear but the rest ain't. #1 Miami, then #2 Nebraska, #3 Colorado with 2 losses but who just beat Nebraska, #4 Oregon with Joey Harrington. Getting to six includes 2-loss SEC teams #5 Florida and #6 Tennessee. You're leaving out 1-loss Illinois and 2-loss Texas here but 2-loss Tennessee was a shoe-in for the national championship game until falling in the SEC championship. An expanded field of 10 also draws in Stanford and Maryland. Ideally: Plus-One.
2002: #1 Miami, #2 Ohio State, HUGE GAP, #3 Georgia, #4 USC, #5 Iowa, #6 Washington State. This is the year you want to just skip to an N.C. game because the top two are undefeated and everyone else has 1 or 2 losses against easier schedules. A 10-team playoff includes Oklahoma, Kansas State, Notre Dame, and either Texas or Michigan. Could you really build a strong argument that the 2002 team is a national title contender? Ideally: BCS
2003: A top tier of three 1-loss teams: #1 Oklahoma, #2 LSU, #3 USC, then and easy cutoff between #4 Michigan and #5 Ohio State, #7 Florida State. Again you're picking between 2-loss teams for the 6th spot. Here I drew in FSU over Texas for winning their conference (not an auto-bid but it can count). Whichever team that is would have to play in Ann Arbor under the Brian plan to avoid having a repeat of M-OSU in the same place a week after The Game. The next four teams would include Texas, Tennessee, Miami (YTM), and either K-State or 1-loss Miami (NTM). This is a great case for a 4-team 
playoff, a decent case study for a 6-teamer, and shows how a 10-teamer is getting down to 1-loss MAC teams. Ideally: Brian.
2004: Again a clear top tier: #1 USC, #2 Oklahoma, #3 Auburn. A fourth is #5 Cal or #4 Texas, a sixth undefeated #6 Utah. Undefeated #9 Boise State is out there too. Expanding to 10 includes 2-loss Georgia, Virginia Tech, and 1-loss (not Big East yet) Louisville. Ideally: Brian.
2005: #1 USC, #2 Texas, BIG GAP, #3 Penn State, #4 Ohio State, #5 Oregon, #6 Notre Dame or maybe #11 WVU? Like '02 this is a "just play the NC" year. Twice in four years is enough to write a fix into the system for this sort of thing (more on this below). A 10-teamer includes Georgia, Miami (YTM), Auburn, and either VT, WVa., or LSU, or ??? – there are fully 10 two-loss BCS teams. Ideally: BCS.
2006: A one and many situation again. #1 Ohio State, then pick one from #3 Michigan (no need for shenanigans), #2 Florida, #5 USC, #4 LSU, #8 Boise State. I slotted in undefeated Boise over 1-loss Louisville and Wisconsin, and also moved USC over LSU for winning their conference. Going to 10 includes them plus probably Auburn and Oklahoma; after that is Brady Quinn's 10-2 Notre Dame who don't belong near an NC game except in ND fans' minds. Ideally: Brian.
2007: Sixer would have #1 Ohio State, #2 LSU, #3 Virginia Tech, #4 Oklahoma, #5 Georgia, and #10 Hawai'i. This might as well be 2011 with another pretty sure-fire #1 and some confusion after that. This would be a hard call between a BCS game (LSU's a strong #2 while the other 1-loss team is #8 Kansas) and a 6-teamer. Going to 10 includes Mizzou, USC, Kansas, and West Virginia, who are indiscernible from Georgia and VT but cuts off before 10-2 Arizona State. Ideally: Doc Sat.
2008: This was the season that wasn't played. Henri the Otter of Ennui wins. Okay fine this is a mess of seven 1-loss teams at the top and two undefeated mid-majors, one of which played Michigan and respectable MWC schedule. Sixer ends up with #1 Oklahoma, #2 Florida, #3 Texas, #4 Bama, #6 Utah, and #5 USC. Sorry #9 Boise State. After that there's 1-loss Texas Tech and Penn State and 2-loss Ohio State. If you're okay with leaving out Boise for USC it's Ideally: Brian.
2009: It's not 2004 despite three undefeated BCS teams since the Big East was by now a mid-major. #1 Alabama and #2 Texas in easy, and #3 Cincy and #4 TCU after. Going to six includes #5 Florida and #6 Boise State. Only Florida among the six has a loss. The next four are Oregon, Ohio State, Georgia Tech and Iowa, all with 2 losses so 10 teams would only muddle things that are fine, but this year would work well as BCS, Plus-One, or the Cook Six Plan. Ideally: Brian.
2010: Top two are easy #1 Auburn and #2 Oregon. Top six hauls in #3 TCU, #4 Stanford, #5 Wisconsin, and #6 Ohio State. Again this is tailor-made for six teams (three undefeated, three with one loss). It's tempting to go with the NC format, TCU be screwed, but six is just fine. The 2-loss Sooners and Razorbacks, and 1-loss MSU and Boise would draw into a 10-team field. Ideally: Brian.
2011: Two is a rematch of #1 LSU and #2 Alabama. Four is #3 Oklahoma State, plus either #4 Stanford or #5 Oregon who beat them. 
And #7 Boise State, now with BCS scheduled teams and TCU. I'm giving Boise the entry in a six-team system over Arkansas so we don't have half the field from one conference. Ten teams would be a bitch (Hinton includes Clemson in there—the BCS standings would have four SEC teams in a 10-team field). Ideally: Brian.
So you're saying the boss's system is better?
Yeah, I…wait I have a bolded subconscious alter-ego too now?
No I'm Ace's bolded alter-ego, filling in.
Where's mine?
The coaches like me better. Boom BCS'ed!
Really?
He got bored right around the time you started going over every year since 1998.
:( So final score is Brian 9, BCS 3, and 1 each for a Plus One and Doc Saturday's 10-team bowlstravaganza. So six is the best solution, but far from a perfect solution. This makes sense when you look at an average season. For this I can even give you a
Ch..
..art of how many of each type of contender we've had in 14 Final BCS Standings:
| Team type | Avg. per season |
|---|---|
| Undefeated BCS Teams | 1.4 |
| One-loss BCS Teams | 3.4 |
| Undefeated Mid-Majors | 0.8 |
| Two-loss BCS Teams | 5.8 |
This is a loose argument for a six-team playoff. There's a reasonable chance of having four or five undefeated or 1-loss BCS teams, plus one perfect mid-major, every year. Those mid-majors aren't going away with TCU and Boise joining one of the recently pilfered BCS leagues; you can see Marshall and Tulane popped up before they did. However any given year should expect plenty of 2-loss BCS teams, more than you want to pick from to expand to a field of 10. Six draws an imaginary circle around the top three rows and suggests most years you can get between 5 and 6 comfortably competitive playoff contenders.
But then you still have 5/14 years when that's not ideal in just this little sample. Is that acceptable?
No it isn't. Even if you figure the perfect Plus-One year and the perfect Doctor Saturday year wouldn't bother too many people if we rammed them into a six-team field, what's unacceptable are those three BCS wins. It's better than the BCS's 3/14 but hell some years you just wanna see Ohio State versus Miami (YTM), or Florida State versus Vick, or the Pete Carroll's Hollywood All Stars versus the Vince Young Show. So:
Let's have that!
Let's propose the six-team playoff system I'll call Brian-Plus:
- Six-team field chosen by a select committee/cabal like in basketball
- #3 and #4 hosting #6 and #5 respectively in home field quarterfinals the week after the conference championships (mid-majors who get in will almost certainly fall in that that 5- or 6-seed range to preclude too much blue turf in Round 1)
- Semifinals in Sugar and Orange Bowls on Jan. 1.
- Final a week later in the Rose Bowl.
- All other bowls left alone; bowls can schedule Round 1 losers. Rose Bowl can have its regular game a week earlier with the parade.
…but that seeding committee can also choose to declare a clear national championship game. So basically when they meet they decide a.) Is it two or six this year, and then b.) If it's six who gets in and how are they seeded? On years when there's a clear two-team BCS game we revert to something like the current system, with bowl tie-ins for the regularly scheduled bowl games.
I would also suggest removing one game from the regular season schedule (if only this would solve the FCS problem) so that the conference championships are played over Thanksgiving and Round One of the playoffs be a week after. Maybe that's pushing it.
December 13th, 2011 at 8:21 AM ^
I don't really see the downside in making the "clear BCS" years a "Brian" year also. IOW, if there are two clearly dominant teams, then they will get first-round byes, and then they will stomp the lowly winners of those first-round games, and then they will beat each other bloody in the Rose Bowl in a glorious gridiron battle in for world domination.
Your (Sethopogon) way would be murder on logistics, would it not?
December 13th, 2011 at 8:37 AM ^
Not really. We generally know pretty far out in advance if it's going to come down to just a couple of teams unless X or Y happens. The logistical nightmare is scheduling two home games a week after the championship games. That decision needs to be made very quickly so travel plans can be made and the host city readied. Imagine Michigan coming off a win in Indianapolis and immediately having to prepare for Boise State the next week. That would be their 14th game.
The rest of the games are the bowls.
The best thing would be to have games scheduled far out in advance. This is a big part of the reticence from NCAA to go to a playoff: are Boise fans really going to fill three games with 1 week's notice? Not even M fans will. They give everybody a month to get to the bowl games right now.
This is why it is preferable, if possible, to have a 2-team playoff instead of using the six. It's by far the easiest to plan. You're losing a total of two games: those two home games in the first round. You still have 5 BCS bowls the first two weeks of January, and those don't change their schedules.
December 13th, 2011 at 10:24 AM ^
I like your idea of 2 or 6, but realistically for this to work, you can only have the one neutral site game. The two other bowls aren't going to be able to sell tickets if they don't know whether they're selling national semi-finals or regular old conselation games. Huge difference.
Also, I agree that college football would need to cut back to 11 regular season games... which is why playoffs won't happen. People like to scoff at other concerns, such as finals, student travel, too many games, etc. But in the end, those things do matter. And one of the big ones, to me, is that college kids shouldn't be playing a brutal physical game 16 times in a season. Even ignoring long term quality of life issues, how can you send a future first round pick out there 16 times to risk his professional future? One more shot for a Willis McGahee type injury.
Since the NCAA will require 11 game schedules, there will be a huge resistance to a playoff of more than 4 teams. That extra game is huge for program funding, both from gate receipts and TV contracts.
Edit: As someone said below, the whole 2 v. 6 decision is a bit of a utopian wish than even a reasonable reality. TV contracts would prevent "cancelling" 4 huge money generators.
December 13th, 2011 at 11:15 AM ^
I don't know where you're getting "4 huge money generators" when the difference between a 2-team playof and and a 6-team is two games. I think you're misunderstanding the timeline and the setup.
Sunday after Thanksgiving: Format for this year (2 or 6) is determined, as are participants.
December 4ish: If 6-team format, two games: #6 at #4, and #5 at #3. If 2-team format, no games.
December 5-ish: Bowl games are determined. If 6-teamer, winners move on to a 4-team playoff that uses two BCS bowls then a Plus-One the week after. Losers may be selected by bowl games.
- Six Team System: Two BCS bowls will host #1 v. lowest seed remaining, and #2 v. other remaining team. Other two BCS bowls will host their normal bowl games. These games will all be played on Jan 1.
- Two Team System: The #1 and #2 teams will be scheduled to play each other one week after the Jan. 1 games. Excepting those two teams, the BCS bowls take their usual teams and tie-ins. This is the system we have now.
Note that no extra games have been scheduled. Either way there are 4 games on Jan 1, and one game Jan 8-ish. The difference is in whether the Championship Game on Jan 8 will be filled with the top two teams selected after Thanksgiving (2-team format), or whether they will get the winners of two of the previous week's champions.
Also note that there are only three games that are scheduled with less than 4 weeks notice, and that these are two home games for the #3 and #4 teams in the country against #6 and #5, and the National Championship Game. Imagine if Michigan had defeated MSU, then went to Indianapolis and beat Wisconsin. We would be #3. Would you spend $80 to go to the Big House last Saturday to see us play Oregon? That's rhetorical. Now imagine Michigan wins that game and we are placed in the Orange Bowl against Alabama on January 1. Do you go to Miami for that? Now say we beat Bama and 1 week later Michigan will be in Pasadena for all the marbles. Fuck you say. But it's the national championship so you gotta. For the NC game, plenty of people who missed Miami will go. Or will they?
Maybe this needs two weeks, not one.
December 13th, 2011 at 12:09 PM ^
Think of NCAA tournament vs. NIT. Once you establish a playoff scenario, every other game becomes that much less important. The value of the the 4 playoff games you lose in skipping right to the finals cannot be made up by reverting to a bowl system in years where a clear #1 and #2 are established.
Even granting your point of only losing 2 total games, these are still extremely valuable. Everyone talks about how much money a playoff would generate. If that system was put in place, you'd really think the powers that be would allow the games to not happen? If a Big Ten championship game was worth $15M to $20M per year in television alone? How much would two national quarterfinals be worth? At a bare minimum, lets say $50M. The NCAA is just going to allow that money to slide because there's a clear 1 and 2?
On your other point, I think getting 30,000 people to travel twice spending roughly $400 for a flight, $400 for two nights hotel, and $200 per ticket, plus all other expenditures, is a bit aggressive. Even getting 15,000 to do it twice and 30,000 others to go to one or the other would be tough. And that's just a money issue. Others might simply not have the time. I'd love to go to the Sugar Bowl this year, but with travel around the holidays and work issues, I can't do it.
And yes, you'd need at least two weeks to make it reasonable if the semis are at neutral sites.
Edit: Again, I really like your plan. I just think realities interfere. In a perfect world, the only thing I would change is make the semi finals home games 1 or 2 weeks after the quarterfinals. This screws the bowls, but I don't really see why we should care.
December 14th, 2011 at 10:22 AM ^
I don't think there's a significant difference between a regular BCS game and a semi-final. People will only go if they care about the teams. You'll sell more tickets to a BCS bowl than a Semi-Final, I think. That's because few fans will go to both the bowl game and the National Championship, but a lot will go to either/or. The same people who would go to the bowls would go to the Semi-Final. Then there's a new market reached of people who say "holy shit my team is in the National Championship Game" and they go. You're right -- they probably do need two weeks lead time.
On TV, the semi-final will get more viewers, but again the BCS bowls already get fairly close to what you'd expect from any given regular season's top game, so the uptick will be marginal. However I think a BCS National Championship game makes up for that. There's more interest when there's a clear #1 and #2. That's what helped make Football Armageddon such a big deal--nobody else at that level was still undfeated. The three BCS era games that actually did get two obvious teams were some of the best watched: FSU-VT, OSU-Miami, and USC-Texas are three of the top four in ratings. The first was last year's between Oregon and Auburn, which is another matchup of unbeatens that I put down as a "Brian" format only because TCU wasn't given a shot.
December 13th, 2011 at 8:24 AM ^
But considering how much planning, advertising and money are set in place for these games well in advance, it's also not realistic.
<br>
<br>It sounds like a plus one may actually happen. I foresee it lasting for a while if it does, with many.a tweak over its first dozen years. I hope the first round games are a week after championship weekend at the higher ranked's home field.
December 13th, 2011 at 8:31 AM ^
I am all for a 6 or 8 team playoff, but I think the ability to "declare a national championship game" is a step backwards-
1. I think it is unworkable- how do you market a playoff and sell tickets when the games might not exist at the end of the year based on some committee vote?
2. Many will disagree (in the heat of the moment) on a "clear" #1 vs #2- it will lead to cries of favoritism and <insert conference here> bias etc. Even if 70% agree on a clear #1 and #2 (a pretty strong majority), why disenfranchise a third of the fans?
3. If you support jumping to the NC game with a "clear" #1 and #2, why not skip the whole thing in a year like this where there is a "clear" #1?
4. your earlier point was absolutely correct that a 12 game unequal schedule is not enough to identify the best team, or even the 2 best, so why even include that option.
5. half the fun of March Madness is seeing the underdogs make a run at the title- why remove that
6. you did not comment on the impact of conference championships. Many of those 1 loss teams in 1998-2004 would have an additional loss now because of their conference championship. As this evolves, I could see more of an argument for a good 2 loss team being made. (certainly more this year than in past years)
December 13th, 2011 at 8:42 AM ^
I don't think you could ever reasonably get away with just having a NC game when most of the time it's a 6 teamer. That'd be like basketball saying this year they are only doing 32 because it's just slim pickings after that.
I'd also like to see a bye week after the conference championships for the 3/6 and 4/5 games - 13 games is already quite a bit and I want the teams to be healthy and ready to go instead of just having played a conference championship game and possibly injured and/or unprepared. This would also give us college football on Dec 10th which would be a nice filler between the regular season and crappy bowl games.
Oh, and the games would have to rotate amongst the 4 BCS sites - we can't just exclude one of them because we think it's stupid. Fiesta, Sugar, Orange and Rose all have to be included.
December 13th, 2011 at 9:40 AM ^
The BCS option makes the regular season a little bit more important. Imagine we just got to the end of 2005 and then:
A.) Penn State and Ohio State host Boise State and Oregon, respectively, the week following the season (both win). This sets up undefeated USC vs. 2-loss Ohio State in the Rose Bowl because the Buckeyes already played Texas, and undefeated Texas plays 1-loss Penn State in the Fiesta. At this point Ohio State gets some major NCAA love -- as in the officiating of the game is more mind-blowingly, permament defacingly biased against the Trojans than those two schools' respective NCAA investigations -- and just barely wins. Now you have 2-loss Ohio State and undefeated Texas playing again a week later, and 1-loss USC left out in the cold despite actually having had a better overall season than OSU.
or
B.) Round 1 is canceled. Penn State (Big Ten Champ) plays Oregon (At Large #2) in the Rose Bowl, Florida State (ACC Champ) and West Virginia (Big East Champ) match in the Orange. The Sugar has Georgia (SEC Champ) facing Notre Dame (at large). And the Fiesta gets Ohio State (at large #1) and Virginia Tech (at large #3). Texas plays USC in the National Championship game a week later.
December 13th, 2011 at 2:19 PM ^
Can it be the Fiesta Bowl played at the fiesta bowl (asu)not the online university stadium that is in Scottsdale not the capitol.
The Orange Bowl in orange bowl....note, Miami (YTM) may have knocked it down...not sure.
Rose Bowl is still the Rose bowl, thank God they got brains.
Sugar Bowl ditto.
I agree that either a) it is 11 game regular season schedule; b) the you all had it but another bye week; or c) the same # of games they play now, but schedule it so all 120 teams are connected to each other somehow. (s-o-s, opp. s-o-s). Maybe it would be playing too much of a what if game, but looks to me like one would be able to look at each team (who beat who) and say they are better.
Side note want to be a 2nd generation U of M grad student graduation.... stupid out-o-state resident rules with divorced parents.
December 13th, 2011 at 9:07 AM ^
...either by messing with the matchups or other side effects of the systems (can you imagine the Rose Bowl being a consolation prize for a team that loses?)
I think it would be better, and I don't know why a system like this isn't a part of the conversation, if we went back to the old bowl system and old bowl tie ins, but kept the BCS ranking system. Then, after bowl matchups are picked, the BCS ranking could be used to rank the bowl games according to their highest ranked participant. The top 4 bowls would serve the 1st round of an 8 team playoff. On the pros list you've got an 8 team playoff, complete preservation of the old bowl system and its significance, and only 2 more games than the current system. That's a great list of pros. On the cons side, that system would let some Cinderella teams in and could leave out a team ranked as high as 5. No big deal in my view.
December 13th, 2011 at 9:10 AM ^
But it was hard for Woodson's Rose Bowl Interception to stop the comeback when WSU was winning 7-0 at the time. It may have made the "comeback" possible. Maybe confusing it with Ohio State?
(Only being pedantic because messing up the '97 season is sacrilege)
Edit- (around the 6:05 mark)
(As opposed to the 1:00 mark)
December 13th, 2011 at 9:30 AM ^
At 8:10, picking up Lloyd?
December 13th, 2011 at 9:33 AM ^
I was about to post the same observation. I'm pretty sure that it is.
December 13th, 2011 at 9:54 AM ^
Good times.
December 13th, 2011 at 9:32 AM ^
I really like the article, up to the bit about declaring an NC game. Why? Here's why.
The BCS, as it is now, is often accused of being esoteric and too political. It uses hidden proceses (the computer algorithms) and voters of dubious intelligence (Herbie, i'm looking at you) to magically determine who plays the NC game, leaving out obviuosly qualified teams (USC, 2003). Why would we want to give a selection committee power to change the rules year-to-year? Could you imagine BSU, sitting at #3, complaining because the selection committee picked the two one-loss teams that were far and away better than all the rest to play for the NC while leaving them out? Perhaps, in that situation, we'd have them going with the playoff, but maybe not! and there's no way to check the authority of the committee. I say give the top two teams a bye (Brian's proposal), and play by the same rules every year.
December 13th, 2011 at 9:43 AM ^
Although I see what you're doing here, I have to agree with several posters above: you can't switch the format arbitrarily based on the year. This would somewhat invalidate the system. My point: one year, the national champ plays only the consensus #2 team and is declared the champion (because there's "slim pickins" after #2); the next year, a very good team (possibly unbeaten) given a #3 seed must win three playoff games to win the championship. The difference between winning one game and three against excellent opponents (especially after a long season) is enormous, and arbitrarily setting that difference based upon the season's results will be unacceptable.
The teams must buy into a set, agreed upon format before any games are played.
December 13th, 2011 at 9:52 AM ^
I like the idea of an actual playoff for the National Championship and I don't really like the idea of some committee being able to arbitrarily declare that there is a pretty clear #1 and #2 and we don't need the rest of the mess. To me, that would put too much power in the hands of that committee and thus would open up to more shenanigans than we already have. It seems like there are already too many backroom deals that allow the selection of a "champion".
If it were up to me, the playoff would only include Conference Champions or Co-Champions. I am of the opinion that if you can't win your conference, then you have no business playing for a national championship. I know that's not fair to teams that play in tougher conferences or tougher divisions within their conference...but life isn't fair. But given the current state of the conferences, I recognize that's not entirely fair either as there are clear disparities amongst the conferences. So I don't know.
But I think the main goal should be to eliminate the polls completely except for seeding the champions.
December 13th, 2011 at 10:20 AM ^
I think this looks like a reasonable approach, except that with the money to be made from these games, I doubt that the committee will ever not play the extra games. I had something similar in mind to meet Brian's goal (which I think is a good one) of a "Restricted Field", while making sure that truly reasonable teams aren't left out.
What if it were similar to "making the cut" in a golf tournament? I'm going to reverse the logic to be exclusive, rather than inclusive, so I'll call it an inverted PGA cut method (with a nod to the inverted veer). You have an 8-team format (home field for 1st round, no byes), but this to make the cut, you also have to have no more than 1 more loss than the leader. So if there's an undefeated team, no 2 loss teams make the tournament. If you end up with 7 teams, #1 gets a bye. If you end up with 6, it's Brian's format. 5 teams, 1-3 get a bye. If you have 1 undefeated team and everyone else has 2+ losses, maybe you go back to 8, but in this situation, I'd personally consider that #1 team the best in football regardless of tournament outcome.
December 13th, 2011 at 11:14 AM ^
In golf every competitor is literally playing the same field so making cuts based on the leader makes sense. In football there is so much variability in schedules just going on losses makes no sense.
Putting on my Nostradamus hat, if this system came to pass I envision non-conference schedules chalk full of FCS teams to avoid losses and thus having no interesting games to watch in September.
December 13th, 2011 at 11:52 AM ^
I would argue that:
- Plenty of teams schedule the lollipop guild for their non-conference schedule right now,
- The number of losses is an additional exclusion criteria in my case, so you'd still have to make the top 8 even if you go undefeated. This makes scheduling a risk-reward endeavor much as it is today.
But, I could see a possible improvement being that we cound number of FBS wins rather than total losses. You can include conf. championship games in that total or not depending on how much you want to reward conference champions. If there's a 13 win team out there, who's played all FBS schools, no team with fewer than 11 wins is invited to the party. (note, my proposal would also exclude 11 win teams that the committee would not have included in their top 8) I know there's still discrepancy with strength of schedule, but that plays in to whether you make the 8 to begin with.
December 13th, 2011 at 10:22 AM ^
It's not that hard.
Quarter finals: 8 teams in with top 4 seeds getting the home game (SEC teams will cry if they have to play in cold weather).
Semi finals: final 4 with top 2 getting home game
After the to week playoff, fill in all the bowls to preserve the bowl tradition.
December 13th, 2011 at 10:31 AM ^
I'm pretty sure in '99 'Bama was #3 going into the '00 Orange Bowl. In other words, are you using post-bowl records to determine your results? If so, they are inaccurate as to determine which system would've been best post-regular season.
I am for the following:
- 16, 8-team conferences in D-1A, determined by geography and total athletic-dept. budget and resources
- The members of these conferences must compete in all sports. Any sports programs in D-1A in certain sports but not others cannot be in one of these conferences.
- Every year each conference will randomly draw 4 other conferences with which to determine its non-conference match-ups. For example, let's say the B1G draws the ACC, WAC, Sun Belt, and Big 12. The non-conf. match-ups will be determined a la the B1G-ACC hoops challenge; i.e., last season's 1 vs. 1, 2 vs. 2, etc.
- In conference, each team will play every other team in its conference, for a 7-game conference schedule.
- Only computer rankings will be used to determine playoff standings. Computer methodology will be reviewed by committee of ex-players and coaches.
- Playoff field is 16-teams, remaining bowl-eligible teams may compete in bowl games b/w 12/20 and 1/1. Playoff games are held every Saturday for four weeks beginning the 2nd full-week of January.
- Playoff field is determined by taking teams with at most 2 losses.
- If the field has more than 16 teams, the lowest ranked 2-loss teams are dropped.
- If the field has less than 16 teams, the highest ranked 3-loss teams are added.
- The field is seeded 1-16 by rankings.
- The first round will be hosted by higher-seeded team.
- The quarter-finals, semi-finals, and finals will be at a neutral site, rotating b/w the Rose, Sugar, Fiesta, Orange, Cotton, Cap-One, and Outback Bowls (with the Outback site changing to Lucas Oil Stadium so the northern teams don't always have a road disadvantage)
December 13th, 2011 at 10:58 AM ^
Step one. Disband all conferences
Step two. Completely blow up all new conferences
Step three. End all non-conference rivalry games
I think you can get just about 100% agreement that no playoff system can be good enough to to be worth those changes.
December 13th, 2011 at 11:30 AM ^
He has a point though. The biggest problem with the current system is the disparity amongst the conferences. That is the knock every year against Boise State. They go undefeated or nearly undefeated and they have one "real" opponent on their schedule. So in order for a playoff to work, you'd have to completely blow up the current conference arrangements but that's not going to happen...
December 13th, 2011 at 10:42 AM ^
Thanks for going back and looking at each of the years. I was thinking about doing this, but I don't remember most of the media attention around a lot of these games. My version would have been based solely off of final record, which doesn't take into account teams that lose very close games or beat top teams. I do take issue with a couple of points in your method, though.
First, I don't like that you give no meaning to conference championships, or at least winning your conference, especially outright. I would say that the #7 team in the country, who won their conference, is more deserving than the #6 team in the country, who finished 2nd, or even 3rd, in their conference. The whole issue this year is that Bama had their chance at LSU and lost. OSU did not have a chance to play LSU, played in a tougher conference, and had the same record as LSU, although their loss was admittedly worse. Forcing conference champions, even if you limited it to conference champions in the top 10 or so, would provide a better field than just picking the top 6.
Second, I dislike that instead of determining the best number of teams to accomodate and assigning points to those sizes, you just seemed to eyeball how the systems fit and made it work. For example, in 2004, it seems that you chose Brian's just by eyeballing it, but you say there is a clear top tier of 3 teams. It seems to me that 4 teams would be idea to solve this, rather than 6. If you go to the second tier to include 1-loss BCS teams or undefeated mid-majors, then I think you have to go to at least 8. In this case, what I would have done, is assigned 0 points for a 2-team setup, 1 for a 4-team, 0 for a 6-team, 1 for an 8-team, and 0 for a 10-team. Then, just add up the points at the end to see which ones are ideal. Once you figure out how many teams are desirable, select the system based around which system with 6 or 8 teams (assuming those two are the ones that come out on top) would have fit most of those years. So, if 6 teams is ideal, then check if top 6 is ideal or if it's some combination of top teams and conference champions.
Another thing to take into consideration, which isn't really a point of issue, is that, in 2007, LSU leapfrogged a bunch of teams in the final week. LSU was ranked 7 the week before the championship games. Missouri and WVU lost. #3 OSU moved back to #1.There were still Georgia 10-2, 11-1 Kansas (Only loss to Mizzou), and 10-2 VaTech ahead of them. Had the voters not had to choose to raise LSU to #2 to fit the system, LSU likely remains at #5 or so. You said that Doc Sat (10-team) would be ideal, so it doesn't really matter, but I thought it was worth pointing out. There might have even been some controversies like this from the late 90s or early 00s that I don't remember.
December 13th, 2011 at 11:02 AM ^
Thanks to Seth for doing his analysis and thanks to you for saving me from writing a long reply.
December 13th, 2011 at 11:22 AM ^
First, I don't like that you give no meaning to conference championships, or at least winning your conference, especially outright.
This doesn't say that. What I said is no autobids. The idea of letting a cabal decide this stuff is to let them decide what a conference championship means in context, rather than hamstring them with autobids. They should absolutely consider a team that won their conference. For example in 2001 the fact that Nebraska didn't win its conference could mean they and (2 loss) Colorado end up the #3 and #4 teams, and Pac Ten Champion (1 loss) Oregon would be the team that gets a bye.
Second, I dislike that instead of determining the best number of teams to accomodate and assigning points to those sizes, you just seemed to eyeball how the systems fit and made it work.
I left out a decision chart that was a little more scientific than this. What I did was list the last in and last out for each year under a 2-team, 4-team, 6-team and 10-team scenarios, and then go down the line to see where the easiest cutoff decision could be made. When I'm home tonight I'll try your suggestion.
December 13th, 2011 at 11:50 AM ^
Thanks. I think it would add some clarity if you posted your chart too. You know what they say about pictures.
And another option you might consider for a post-season proposal is to have guidelines for a variable number of teams. Instead of just allowing them to choose 2 or 6 teams, let them calculate a cutoff and have predetermined rules choose the format.. Something like:
If 6 teams are picked, Brian's format.
If 8 are picked, bowl games +1.
If 2 are picked, straight to championship.
If 3 teams are picked, add the highest ranked team and a 4 team playoff.
That would probably be better than just forcing them to pick 2 or 6. Just another thought...
December 13th, 2011 at 11:00 AM ^
With the limited sample available the conference championships serve as an extension of the championship tournament.
I'd argue that 8-4 teams are often better than 12-0 teams. The whole premise of the Big-East and 'Longhorn conference' argument is that they play weak schedules - which also applies to teams that play soft non-conference slates.
The problem is that voters tend to focus only on losses, when just the opposite should occur. It would be better if losses were ignored entirely.
The real problem is the subjective voting that encourages teams to avoid tough matchups. The only way to get rid of this problem is to remove the voters from the equation and limit the pool to conference champions.
If the 8-4 team can beat the 12-0 team, then go on to beat 2 other elite-level teams, they'll have proven deserving.
December 13th, 2011 at 11:49 AM ^
I don't think an 8-4 team should be in after beating a 12-0 team. What you're doing with that is saying Game 13 alone is more important than Games 1, 4, 6 and 8.
Play this scenario out with me. Say Ohio State's 12 best players are found to have been at the head of a murder, cocaine and kidnap-for-prostitution ring for the last four years, and that the head coach, AD, and president spent months covering it up and lying about it in formal documents. Under apparent NCAA regulation levels of Ohio State that would earn those 12 players a 3-game suspension starting after this year's bowl game.
Without their best players, the Buckeyes lose three nonconference games, then get those players back and go 7-1 up to The Game. Meanwhile Michigan knocks off Bama in Dallas, and blows through our schedule. We meet them in Columbus and beat them soundly, but Ohio State has still won their division with only two Big Ten losses, since they beat (also 2 conference losses) Penn State and Wisconsin. Then Ohio State wins in Indianapolis on a last-second freak play that the video replay guy totally blows.
Ohio State is 9-4 and Big Ten Champ. Michigan is 12-1 and everybody has them ranked higher than the Buckeyes because that game in Columbus and the rest of the schedule should mean more.
An autobid would mean Ohio State has to be in.
I don't have a problem with subjective voting on its own -- my problem is that the people who get to vote are either media with significant biases and a lack of knowledge, and coaches who don't bother to look at teams they don't play, and who have even stronger biases.
I think the guys who decide the seeding for the basketball tournament do a fantastic job. This is because they are dedicated to being fair and impartial experts. Autobids are an overreaction to things like this year when LSU got in. The way to do it I think is to remain as flexible as possible so that the consensus best result is never precluded by an arbitrary rule.
December 13th, 2011 at 12:55 PM ^
If the winner of your conference championship is in the playoffs, the loser of that game shouldn't be. In your example, however, Ohio State presumably wouldn't make the playoffs, so Michigan could still make it in to represent the Big Ten. If its a 2006 situation, on the other hand, and we have a conference championship game, whoever wins that game should be in and the other team out. Do we really want the possibility of Michigan/Ohio State playing three times in one season?
December 13th, 2011 at 1:10 PM ^
Game 13 is really Game 6 or 7, at best. You must acknowledge that half the schedule (give or take, but usually more) is not significant in differentiating Top 25 teams from one another, let alone parsing the Top 5. A Houston/Boise type team plays zero, one, or two 'significant' games a year. A team like Michigan or Alabama may play five or six...or two. So, given the pervasive voting criteria (comparing losses), the idea that "everybody has them ranked higher" is not convincing.
This is the biggest issue in college - getting more significant data. Any system that encourages cupcake non-conference matchups reduces significant data. As a result we have subjective, inconsistent, arbitrary and ultimately incorrect subjective rankings.
Scheduling precludes much inter-conference interaction. In contrast, on a relative scale of what we know about college football, the conference's best is relatively clear. How a conference champion is decided won't be perfect either, but that should be a 'local' issue. Be it a do-or-die conference title, or a subjective vote of power, each conference should bring forth it's best.
I want to protect the value of the regular season as much as possible, but my biggest worry is letting teams like Houston through and encouraging the Kansas State and Wisconsin approach to non-conference scheduling. Because of limited sample size, there is no alternative (in college football) to win-or-go-home scenarios. The further you extend those, the bigger your 'tournament' is, the more the champion is determined "on the field."
Your OSU scenario is legitimate, but so is a scenario where LSU losses a key player or four to injury, gets beaten by Oregon, West Virginia, Florida, and Auburn early but then goes on to beat Alabama, Arkansas, etc. The LSU wins the SEC over Georgia and vanquishes the top contenders in the post-season matchups. This kind of scenario plays out in basketball and few complain. Duke is the best team, but they lost to whoever in the tournament, so they don't win the title. Everyone is OK with this, even Duke, because the system is clear and the stakes are known in advance - unlike what we have today.
It's possible that a workable system will still require some subjective voting, but I believe this should be limited. Your suggestion about a Basketball tournament-style selection committee is one I agree with. That gets rid of 90% of the trouble with the ranking system. However, the fact remains that you have a limited number of games to decide a champion. Making only conference champions eligible is a good starting point. We don't need auto-bids necessarily, but we do need to exclude teams that don't win their conference, even if the popular choice for strongest team is excluded. You win the do-or-die games when you have to - otherwise, as Brian, and other playoff advocates argue, why bother playing at all?
I'd like to see the emphasis for college football return to what Bo, Lloyd, and Woody would have advocated - focus on winning your conference. The non-conference should be viewed like the bowls are today --> entertaining exhibitions for fans. They should be viewed by coaches as preparation for the 'regular' season of conference games. Then, the post-season should be used to let the elite conference champions determine a 'national' champion.
December 13th, 2011 at 1:32 PM ^
What do you do about a situation (kind of like this year in the SEC) where the best teams are in the same division? And say Georgia had 4 losses instead of 2, had previously played LSU at Georgia and been destroyed, but then somehow managed to squeak out a win in the conference championship game? Do you leave 1-loss LSU (and Bama) at home and let Georgia in? Do you leave everyone in the SEC at home? If so, you've undermined the system. People will not view it as legitimate.
The other issue I have with your analysis is that losses should matter notwithstanding the excuse for them (even if you have injuries or suspensions, even if your field goal kicker has an off day, etc.). Now maybe we should discount early season, non-conference games generally. Just view them as exhibitions, win or lose (although that seems somewhat counter to your insufficient data point). But once you've decided what games should count, if a really good team happens to stumble and bumble its way through a portion of those games, the stumbling should matter when it comes time to determine a national champion, even if everyone recognizes they are a very good team.
December 13th, 2011 at 2:58 PM ^
There is no avoiding the do-or-die scenarios to decide things "on the field" - not with only 13 or so games. Problematic scenarios will occur in all proposals. There is no solution for A beat B, B beat C, and C beat A. That is the inherent of nature of having small sample sizes attempt to solve difficult problems.
I'm arguing that there must be clarity above all else. The rules of the game need to be known in advance. Even if they are stupid and arbitrary rules (like the Big10's no-repeat Rose Bowl invite thing), they let you know what the stakes are in advance.
Of course, there are better scenarios for the SEC than bringing together the best of the geographic divisions it created for scheduling purposes. As someone suggested earlier, the SEC could simply schedule a post-season game to decide whatever issues are unresolved for maximum entertainment/satisfaction/ratings/whatever. This year, that's probably an Alabama-LSU matchup, but even then you still have the same problem you raise above - Bama already lost to LSU at home.
If the SEC decides a rematch is acceptable for it's conference championship game, then it (and everyone else) would live with the consequences. It would send forth Georgia to the national playoff and if Georgia won there, the playoff they would be 'national champion' just as Butler or George Mason or any other 'undeserving' school could have been national champions in basketball.
December 13th, 2011 at 4:17 PM ^
and I agree that ideally you want the path to a National Championship to be as clear as possible from Day 1.
But you are arguing for a Playoff of ONLY conference champions because that makes the rules clear and eliminates subjectivity. But which conference champions? You could say all of them, but that means a 3-loss Big East and ACC and Mid-Amercian Champion gets in but a 1-loss Alabama team stays home. There's no way the SEC, e.g., is going to agree to that kind of inequity of opportunity at a National Championship, and rightfully so. It also means you've got to move to an 11 team playoff system (and yet the team everyone thinks is the 2nd best in the country isn't in it).
If all conference champs don't get in, you have to have a system for excluding some over others. Now you are back to the uncertainty you despise. And now you have the possibility of the Georgia example above. Say Georgia as a 4-loss conference champion doesn't make the cut, and Alabama and LSU with only one loss aren't eligible because they didn't win the conference. So now the SEC, who has arguably the two best teams in the country, has NO representation whatsoever in your playoff system. That possibility also isn't going to fly.
I suppose you could say just BCS conference champions, and the two highest ranked champions of the non-BCS conferences for an 8 team playoff. That's at least feasible. But I still don't like that, for example, a 3-loss Big East or ACC team automatically gets in...that is hard to justify. As much as I don't like untested assumptions about conference superiority, I also don't like just totally assuming away conference inferiority. The idea that, for example, 3-loss Clemson could beat a previously unbeaten LSU in the National Championship game and would be crowned National Champion isn't a system I care for very much.
December 13th, 2011 at 5:07 PM ^
Impossible to remove subjectivity entirely. You still have to pick the top 6 (or whatever) conference champions. It's true that there would be some uncertainty for fringe champions (like Clemson or TCU or WVU this season). I'd use a selection committee to pick from those, which would reduce the liklihood of teams that played easy schedules getting in and the prevelance of the loss-comparison approach. Like any other proposal, its imperfect but an improvement.
The difference between the MGo system and mine is that it effectively extends the size of the playoff to include conference championship games (and reduces the size of the pool by excluding teams like Boise, Alabama, and Stanford that lost to their conference champ.) The subjectivity is there, but it's about choosing the 6th best conference champ compared to the 7th or 8th instead of trying to make determinations about conference strength on almost no evidence.
This is viewed by Brian and others as watering-down the pool of team quality, but I think that's based on very flawed assumptions about who 'had the better season'.
What's really significant in the differences in our proposals isn't the quality of the playoff matchups, but ehe ancillary effects. With the focus on conference games and demphasis of the consequences of non-conference losses, college football becomes a more entertaining sport. Non-conference games would still matter a great deal, but your national title hopes would not be eradicated with a 'quality loss'. Tougher non-conference scheduling is rewarded. The significance of the polls becomes no more than in basketball. Competitive matchups increase. Clarity comforts sports fans around the country as a conference championship becomes a pre-requisite to a national one.
If I was King of the World, I'd extend this even further. I'd mandate that to be eligible for the playoff you had to be part of an 9 to 11 team conference that plays round-robin schedules. [Or just have an 16-22 team mega-conference that's really just an agreement between two conferences to play.] You'd also have to be willing to host or travel to face another conference champion the first week of December.
December 13th, 2011 at 5:31 PM ^
Say Georgia manages to beat LSU in the conference championship game but still isn't good enough to be considered one of the top 6 conference champions. Now the SEC, with arguably the #1 and #2 teams in the country, has NO TEAMS in the playoffs. This system basically tells conferences they stand to be punished in the National Championship picture if they have a conference championship game.
I'd be happy to have a system that only allows one team per conference and that favors conference champions if they are viable. But if the conference champion isn't viable but another member of the conference is, the conference shouldn't be excluded from the playoffs altogether. You could even limit this exclusively to conference championship game situations. Allow the loser of the game to potentially make the playoff, but only if the winner doesn't make it.
December 13th, 2011 at 7:07 PM ^
First off, in your scenario Georgia is 11-2 and SEC champ, coming off 11 straight wins after an 0-2 start against 2 tough teams. They'd be a lock to be in the top 6. But just for the sake of argument, let's say they're 9-4 instead (losses to GaTech and New Mexico State), still beat LSU for the conference title, but aren't considered a top 6 team (unlikely, but possible). I still don't see that as a huge problem. Certainly not relative to what we have now, which is a handful of ultimately meaningless games and a national title game that solves nothing and will leave only one team's fanbase remotely satisfied.
In that scenario, the SEC decided to bring a weak team forward as it's championship representative. That's on them. That's the downside of having a rigid structure to championship games. The upside is drama/ratings/money that come with staging a winner-take-all championship game.
If a conference wants to call the highest ranked team at the end of the season their champion, fine. If they want to have the winner of the championship game, fine. If they want to define a system tie-breakers, that's fine too. Whatever the conference thinks is best.
They system doesn't punish conferences in any way. It encourages them to figure out a system wherein they bring forward their best. This shift of responsibility is better than asking a national poll of uninformed pundits and coaches to figure out who is the best and how runners-up in one conference compare to champions in another based on scant evidence.
You're right that the SEC loses in the above scenario. And if they went with polls to decide their representative that would defeat the purpose of many of the aims I identified above. But, again, its a better system than what we have and what is being proposed. The finality and clarity of it remain intact.
December 13th, 2011 at 7:09 PM ^
There should probably be an OUT for situations like this. Perhaps a rule where if a team is ranked in the top 10 and the conference champ is 10 or more spots behind them...
That'd make the conference championship irrelevant (as the SEC's was) but so be it.
I favor letting the conferences figure this stuff out for themselves, not national voters.
December 13th, 2011 at 1:42 PM ^
You're right, absolutely right, about the "useless" games. That's an issue for another day unfortunately. One quick fix would be for them to treat all FCS games as ties, or just not count them at all. They're exhibitions anyway. Sadly there's not much differentiation that can be made other than schedule strength to catch the true difference between playing Houston/TCU/Boise and, say, Toledo.
With the LSU scenario however, no, what happens when you're injured should count. One of the oldest coaching platitudes out there is that the guy who's number 2 or number 4 on the depth chart has to be ready to play if No. 1 goes down. Injuries are a part of the game, and as in basketball the depth behind your starters is a significant factor.
This is the important part: I don't want the six teams most likely to win. I don't want them seeded by Vegas odds. The playoff should reward what happened on the field, not who would win a rematch. If Michigan would have beaten MSU 9 times out of 10 this year and just happened upon a garbage tornado, well, kudos to MSU for having the kind of team that can win in a garbage tornado.* That hypothetical LSU team lost early because they didn't have depth (the real LSU, thanks to oversigning, would actually benefit greatly from depth playing a bigger role), and that should be counted against them.
I agree wholeheartedly with your last paragraph, but I think that's even less realistic than what I proposed. The conferences are too big and have divisions and championship games now. The game should be regional, but those days are over.
------------------------------
* That was only part of it. My end take on that game this year is that Michigan is basically an inside option team and Michigan State is built--with Worthy and those blitzing linebackers--to best stop an inside team. MSU cheated against the inside game and in those conditions and with Denard's condition we couldn't make them pay for it. Nebraska's an outside team and that hurt them. Wisconsin's an off-tackle team and that again goes away from MSU's strengths. Had they faced Ohio State later rather than with all 5 suspensions in play that was a perfect game for MSU to lose.
December 13th, 2011 at 2:37 PM ^
As it is the preliminary and primary means to access inter-conference superiority, I don't think non-conference scheduling is a seperate issue at all. Teams schedule cupcakes not only for financial reasons, but because not losing is valued above all else. Fans also become less tolerant of losses because they understand the end-game and importance of rankings and profile. Any playoff proposal should address how non-conference scheduling is influenced. [I don't think the FCS distinction solves the cupcake issue, it just increases demand for the EMU's of the world. It's a fine band-aid though.]
Health, like randomness, it's not a concept most people want to value or acknowldege, but is probably a determining factor every year. Yes, injury is part of the game and depth should be rewarded, and so on, but properly accounting for it is yet another subjective element and criteria for which there is no consensus. Michigan was not deep this year, but they finished 10-2 largely because they managed to avoid any significant injuries. We can pretend that we only care about what happens on the field, but when off-field events so heavily influence the outcomes of our 'game of inches' that seems as ludicrous as it is necessary.
"I don't want the six teams most likely to win...reward what happened on the field, not who would win a rematch." I completely agree and yet we reach different conclusions about how the objective should be achieved. To allow a non-champion into a playoff, or worse yet two teams from the same conference, encourages a second judgment to be made that will either be redundant or contradictory. There is no time for that. Anyone can dream up (or call to mind) examples where it's 'better' to have rematches or non-champions but it's based on extremely weak and subjective evaluations of superiority, plus the exercise can be performed just as easily to make the opposite argument.
The more finality to game outcomes the better. I'm fine with the proposed 6-team playoff as a means to settle what the regular season (usually) can not. That is - which the best team in the nation? But if your 8 to 10 game conference slate can not settle who the better team is within a group of 12, how can 1 or 2 additional games? That kind of system is just asking for contradictory results and an unsatisfying conclusion.
It's not about being 'regional' per se - the new Big East conference is anything but - it's about finding a conclusive answer within a subset of teams with a large (-enough) sample of relevant data before attempting to answer a larger, more difficult question.
It's the lack of clarity that people have a problem with, particularly in sports, where a great appeal comes from the clarity and finality. Winner/loser, champion, forever. If you give teams that don't win a title a second chance it muddles things severely. It creates scenarios like this year where a) a conference championship game with the #1 team in the country is meaningless to the national title game and b) the outcome of the biggest game of the year between #1 and #2 is ultimately meaningless. And this isn't a rare event either. Had the structure of football been slightly different in 2006, the Michigan Ohio State game would have been meaningless as well (a point I actually argued at the time only to be proven wrong thanks to CBS, Danielson, etc.). If coaches ever acknowledge this publically you face the professional sports problem of resting players for more important games later to come...this would, I think we all agree, be tragic.
December 13th, 2011 at 2:47 PM ^
Well now we're rewarding things that are not about who the best teams are, or the quality of their seasons.
It is quite possible that the best two teams in the country play in the same conference. I hate rematches too, and we can make sure the system selects against those just like we can weight a #7 conference champion over a #6 team they beat.
But ultimately sometimes we're going to get that rematch.
There is a way to make a 4-team playoff with just conference champions, and I outlined it when I did a Museday or Dear Diary awhile back proposing conference realignment options. If you get down to four conferences that matter, those four conference champions can battle it out. We don't have that system.
December 13th, 2011 at 3:12 PM ^
I'm not going to restate all my points above but the 'best teams' are the ones that win the games that matter most.
You can either define those 'games that matter most' and the process for being invited to them in advance OR you can let those games be decided based on a undefined, subjective, and constantly changing criteria system that leaves everyone feeling frustrated and confused. I thought the LSU-Alabama game mattered, but it turns out it didn't. Neither did Georgia-Boise or Georgia-LSU.
I favor minimizing that subjective element that leads to so much confusion and frustration.
One end-of-the-year ranking to select participants, amongst conference champions, to compete in direct do-or-die tournament is a significant improvement.
It doesn't have to be 4 teams, it can be 5,6,7 or 8.
Even then, there is nothing that would preclude Oregon from rematching with LSU. (Or in an alternative universe Georgia-Boise), but in my world no one would have a problem with such a rematch because it's clear in advance that one game is more important than the other, it's not an issue where voters decided to declare a previous matchup irrelevant. Instead, they played their way into that rematch.
December 13th, 2011 at 11:07 AM ^
Semifinals in Sugar and Orange Bowls on Jan. 1.
Final a week later in the Rose Bowl.
AGH! SETH! You are smarter than this! The 2007 ACCCG picture is infamous, but it's not proof, as people have said, that Virginia Tech fans don't travel. They sold out their Orange Bowl allotment that year. It's proof that nobody will travel twice.
I know, I know - corporate interests will gobble up the tickets and make sure they don't go unsold. I can't imagine a world in which we should be encouraging this.
I like that that you suggest that first-round losers would go to bowls. It is what I have been proposing for years, so of course I like it. But what say you to the notion that the decision-makers are never satisfied with keeping any tournament the same size in perpetuity? Six seems like an idealistic number, but what about when the powers that be decide that six is a great moneymaker but more is even better! (And how would you ever convince them to not pressure the selection committee to never drop it back to two?) I mean, a 68-team NCAA tournament with the danger that we almost went to 96 should tell you all you need to know about the foolishness that is expecting that ideal sizes will ever stay that way.
Also, yes, it's definitely pushing it to tell the ADs they can't have that extra home game. 12-game schedule is here to stay. Forever. We're never going back to 11, there's too much money in it.
December 13th, 2011 at 11:29 AM ^
Because I can't realpolitik this. If six is ideal then we should start with four so they can expand into six? No. Make it six and stay at six.
Remember if NCAA accepts this system it means they read a blog and thought "hey this person from the internet has a great idea," and actually implemented it. That is such a massive giant stretch of the imagination which would entail a complete reversal in the type of thinking that goes on at that level right now -- if we're living in this fantasy land where Delany et al. could even consider the possibility that the Interwebs might produce a better idea than is in their own heads, why can't we pretend those same people would be sane enough to keep a system that is fair and works once they have it?
December 13th, 2011 at 12:53 PM ^
I can't argue with that logic.
December 13th, 2011 at 11:24 AM ^
That would mean potentially playing as many as 16 games for teams that play in conference championship games and goes too far in undermining the traditional bowls...just don't see a realistic appetite for it amongst the powers that be for that disruptive of a change.
A +1 type format is realistic and looks like it might actually be on the way, and can be done well with some tweaks. Those should be to try and preserve the traditional bowls matchups where possible, provide some tiebreaker type rules that will prevent some of the politicing that goes into determining who gets in, promote conference diversity and give some incentive to winning your conference.
My proposals would be:
- Generally use the top teams in the BCS rankings to determine the Playoff Teams.
- However, NO losers of conference championship games IF the conference winner is a Playoff Team.
This is a fair way to eliminate arguably unworthy teams who controlled their own destiny late in the season, and to prevent repeat games. E.g., what's the point of making the winner of the conference championship game potentially have to play that team again to win the NC a game or two later? (Seriously, could you imagine if in 2006 UM and Ohio had a rematch in a B1G conference championship game--and Michigan won--and then they had to play AGAIN in a 4 team playoff system?) A potential replay of a conference championship game is undesirable, playoff spots are scarce, conference diversity is good...this is an easy elimination. Note, however, that if LSU had lost to Georgia in the SEC championship game, LSU would still potentially be eligible to be a Playoff Team because Georgia would not be a Playoff Team.
- Other non-conference champions (i.e., those who did not play in a conference championship game) should have to play their way in.
In most years, they'd play the first team out. However, the way it would work this year is that Alabama and Stanford (both of whom ducked their conference championship game) would play each other, and Oregon (who played in and won its conference championship game) as the first team out would get the last spot. Note that adding this extra game would create some logistical challenges but won't generally require teams to play more than 13 games, e.g., that LSU had to play. (I'm also tempted to say that if the first team out is a champion of a conference with a championship game, they get in automatically ahead of a nonconferece champion, rather than make the winner of a conference championship game play another game on top of the conference championship game.)
- Select the host BCS Bowl games that best preserve traditional bowl matchups.
So, e.g., if a B1G and Pac12 representative are both in the Playoffs, then they play each other in the Rose Bowl in the first round (seedings be damned). If only one of the conferences is in the Playoffs, or neither, then the Rose Bowl isn't a Playoff Bowl that year, and so you preserve the possibility of a reasonably traditional Rose Bowl. After you decide the Rose Bowl, you figure out the other bowls based on the conference champions in the Playoffs. So this year, LSU would play in the Sugar Bowl and Okie State would play in the Fiesta Bowl and they would both be Playoff Bowls. Orange would be Clemson vs. West Virginia. Rose Bowl would be Wisconsin vs. an At-Large (probably Stanford if they lost to Alabama in the play-in game).
December 13th, 2011 at 1:10 PM ^
in FCS ever survive? They play an 11 game schedule, followed by a 20-team playoff. Teams in the championship play either 4 or 5 extra games, for a total of 15 or 16 in the season. Of the 126 teams in FCS, only 2 are going to play 15 or 16, and only 2 others are going to go 14 or 15.
Wtih even an 8 team playoff in Div I-A, this means 2 out of 120 schools might play 16 games and 2 other schools might play 15 (12 regular season, 1 conference championship, 3 rounds of playoffs). Teams are already doing 14 if they're in the conference championship and a bowl game.
No need to shorten the regular season for the privilege of 4 teams fighting it out on the field for a non-mythical national championship.
The problem of course is when playoff bloat happens in the future (Just Say No!). Then there are too many teams playing too many extra games...
Comments