Writers of Pro Football Prospectus 2008

Most Recent FO Features

FlaccoJoe08.jpg

» Week 3 Quick Reads

We break down the very ugly game Joe Flacco had against Jacksonville, how completely ineffectual the Ravens offense was -- and why it wasn't even the worst game among quarterbacks this week.

06 Nov 2015

Varsity Numbers: Revisiting the Perfect Playoff

by Bill Connelly

Since a) I've been revisiting old Varsity Numbers ideas of late, and b) the initial 2015 Playoff rankings were just released, it seems like a pretty good time to hit on one of my favorite pieces I've written: July 2009's The Perfect Playoff.

One thing is virtually certain: a playoff will happen one day. Maybe not today, maybe not next year, but soon (relatively speaking). What will be the tipping point? Simple sports geology. As Red from Shawshank Redemption would say, "Geology is the study of pressure and time. That's all it takes really, pressure, and time." That, and TV ratings. Another few years of poor ratings for BCS games like Virginia Tech-Cincinnati or Georgia-Hawaii should be all it takes. The playoff ball is in motion -- it's just going to take a while to find its inevitable destination.

So with that in mind, it behooves us, as sports fans, writers, and nerds, to shift the focus from whether there will be a playoff, to what kind of playoff there will be. And only one idea combines the needs and wants of concerned parties and gives everybody a seat at the table. It is the perfect playoff.

(Now, as we knew then and know now, "concerned parties" don't really have a lot of interest in making sure everybody has a seat at the table. We'll just say that's editorial preference on my part.)

Five seasons after I wrote this, we indeed finally got our first playoff, a simple four-team, three-game process. Most of the time, a four-team playoff will work out well -- three teams will be pretty obvious, the fourth will still be pretty worthy, and four teams assures that we don't end up with a three-loss champion at some point and that the regular season still matters significantly, as college football fans prefer it.

But bracket creep is already dominating the conversation. (It doesn't take long, does it?) That Baylor and TCU were left out last year meant people were immediately clamoring for an eight-team draw. We have already seen it come up again this year, even though we don't know if there will even be four title-worthy teams a month from now. It's in so many writers' heads, and they impatiently wait for a chance to uncork "Move to eight already!" for any reason.

Honestly, I like a four-team playoff for the reasons I mentioned above. It's clean, it assures the inclusion of worthy teams, and it makes bowl season even more awesome. College football made a lot of solid decisions in creating this process. (The sport's decision-makers also made some maddening choices when it comes to process and whatnot, but we'll ignore that for now.)

But if we're going to assume that bracket creep is going to happen, and if the sport's power conferences aren't going to go ahead and break apart from everybody else, then said creep would offer an opportunity actually include everybody in FBS.

That's where the Perfect Playoff once again comes into play. Here are the primary steps from my 2009 piece:

1. Tourney Selection Show the Sunday after Thanksgiving. My original idea was that, since this is a 16-team playoff with up to four possible games, we dump conference title games. This obviously doesn't have to be the case. And if we pretend this idea is actually realistic, then money and reality would likely dictate that these games remain.

This, of course, leads to the possibility that the national champion will have played 17 games in a season. Ohio State and Oregon cracked the 15-game barrier last year, but 17 feels like so much more than 15 for some reason. Maybe that's just me. Regardless, this much of a workload would dramatically increase the "pay the players!" argument.

2. The playoffs would start the next weekend, when conference championship games currently take place. Or, obviously, the week after conference championships.

3. The quarterfinals take place the next weekend.

4. On the day after the quarterfinals, bowl pairings are announced. This is one of my favorite parts of this idea: to assure quality bowl matchups, losers in the round of 16 and quarterfinals are dumped into the bowl pool, like Champions League losers are dumped into the Europa League in European soccer. And if you reach the quarterfinals (and are one of the four teams to lose quarterfinal matchups), you automatically get a major bowl bid. So if a Western Kentucky or Appalachian State pulls a first-round upset, the reward is not only a quarterfinal spot, but also a spot in the Fiesta Bowl, Orange Bowl, Chick-Fil-A Bowl, etc.

The downside here is obvious: Think of a team like Florida State last year, which had a ton of soon-to-be draftees. If the Seminoles had to play another game after losing in the quarterfinals, they quite possibly could have mailed it in in a major way. But mail-ins and duds happen regardless, so I guess not much changes there.

5. The semifinals would take place on or around January 1, at a designated [major] bowl. This turned out to be how the current playoff process was arranged.

6. The championship game would take place as it does now -- a week or so after the January 1 bowls (and semifinals) at a designated, rotating host. Ditto.

An updated demonstration

Using 2014's playoff rankings as a guide (and assuming that conference title games were still a thing), here's how a 2014 playoff would have worked out:

Round 1 (December 12-13)

1. Alabama (SEC champion)
2. Oregon (Pac-12 champion)
3. Florida State (ACC champion)
4. Ohio State (Big Ten champion)
5. Baylor (Big 12 champion)
6. TCU (at-large)
7. Mississippi State (at-large)
8. Michigan State (at-large)
9. Ole Miss (at-large)
10. Kansas State (at-large)*
11. Arizona (at-large)
12. Boise State (MWC champion)
13. Marshall (CUSA champion)
14. Memphis (AAC champion)
15. NIU (MAC champion)
16. UL-Lafayette (Sun Belt champion)**

* I flipped KSU and Arizona to avoid a TCU-KSU rematch in the first round.

** Georgia Southern went 8-0 in the Sun Belt, but a ridiculous "you're not postseason-eligible in your first FBS season" rule prevented the Eagles from bowling. I'll assume the same silly rule would prevent them from making the Playoff.

That would leave us with the following first-round matchups:

  • (16) UL-Lafayette at (1) Alabama
  • (9) Ole Miss at (8) Michigan State
  • (12) Boise State at (5) Baylor
  • (13) Marshall at (4) Ohio State
  • (11) Arizona at (6) TCU
  • (14) Memphis at (3) Florida State
  • (10) Kansas State at (7) Mississippi State
  • (15) NIU at (2) Oregon

Based on a combination of ratings and knowledge of how the teams played in the real postseason, you could make a case that we would have seen complete chalk here, with all higher seeds winning. Baylor-Boise may have been interesting, and maybe Ole Miss wouldn't have laid a total egg like it did in the Chick-Fil-A Bowl if there had been higher stakes at play.

But the mid-major field was far weaker last year than it is this year, and the committee's at-large picks would have been pretty weak -- in the final F/+ rankings, Kansas State ranked 26th and Arizona ranked 28th, while teams like Georgia (fourth) and Georgia Tech (eighth) would have been left out. That makes for a pretty top-heavy field (and gives those opposed to a 16-team field all sorts of ammo).

The fun begins the next week, though.

Quarterfinals (December 19-20)

  • (8) Mississippi State at (1) Alabama
  • (5) Baylor at (4) Ohio State
  • (6) TCU at (3) Florida State
  • (7) Michigan State at (2) Oregon

Knowing what we know, the likely winners here would have been Alabama, Ohio State, TCU, and Oregon, though a flawed FSU team would not have gone down without a fight at home, and State-Oregon would have been a battle.

Semifinals and major bowls (December 31-January 1)

The four quarterfinal losers are dumped into the major bowl pool

  • Sugar Bowl/Semifinal: No. 1 Alabama vs. No. 4 Ohio State
  • Rose Bowl/Semifinal: No. 2 Oregon vs. No. 6 TCU
  • Orange Bowl: Florida State vs. Georgia
  • Chick-Fil-A Peach Bowl: Mississippi State vs. Georgia Tech
  • Cotton Bowl: Baylor vs. Michigan State

Oregon-TCU would have been an incredible game, but for simplicity, we'll say the Ducks pull it out. We were hypnotized by TCU's bowl domination of Ole Miss, but Oregon was a better overall team.

Finals

  • No. 2 Oregon vs. No. 4 Ohio State

In terms of bowl structure, there is minimal change here.

An early 2015 projection

Obviously we don't know how things will play out yet, but using the current F/+ rankings and initial playoff rankings as our guide, here's how a 2015 Perfect Playoff might take shape:

  • (16) Appalachian State at (1) Clemson
  • (9) Florida at (8) TCU
  • (12) Memphis at (5) Notre Dame
  • (13) Bowling Green at (4) LSU
  • (11) Iowa at (6) Baylor
  • (14) WKU at (3) Alabama
  • (10) Michigan State at (7) Stanford
  • (15) Boise State at (2) Ohio State

Not going to lie: I still like this idea. You wouldn't see a ton of first-round upsets, but you would only need one or two well-timed David-over-Goliath moments to make the "everybody deserves a seat at the table" sentiment ring true. Seventeen games is obviously a ridiculous amount to play without kids being both compensated and insured, but that's not completely unrealistic.

At least, insuring and compensating players isn't any more unrealistic than the thought of a 16-team playoff that gives a spot to Appalachian State.

Posted by: Bill Connelly on 06 Nov 2015

8 comments, Last at 11 Nov 2015, 12:12am by LionInAZ

Comments

1
by Will Allen :: Fri, 11/06/2015 - 11:51pm

I really think an eight team playoff is optimal, from a t.v. viewership vantage point. 5 automatic bids for conference champs guarantees giant viewership for conference championship games (I think it is inevitable that all 5 have them), and three at large bids doesn't dilute the regular season. Have the quarterfinals at the home fields of the top 4 conference champs, on the 2nd weekend in December (some have said that it isn't feasable for some small college towns to host such a game on relatively short notice, but I don't buy it) and you make winning the conference all that more important, while also occasionally getting to see warm weather teams deal with the elements, which is always fun. The ratings on that weekend would be through the roof. Have the semis and final as they are now.

2
by Alex51 :: Sat, 11/07/2015 - 1:12am

Agreed. Eight teams includes all the teams that really have a good argument to be the best in the nation, using automatic bids for conference champs cuts down on the whining from teams that miss out, and few non-elite teams will make it in. Just right.

3
by White Rose Duelist :: Sat, 11/07/2015 - 10:57am

You should also include any undefeated Group of Five teams, to eliminate the argument that a team can win all its games and still not get to complete for the championship. Though I suppose you could still have a year like 2004, where there are more undefeateds than slots...

4
by Bucs_Rule :: Sat, 11/07/2015 - 3:59pm

If small colleges can't get ready to host the game, then they can just play at the lower seeded home stadium. That should provide plenty of incentive.

I also like 8 more than 16, with 16 teams, games for the top teams won't matter at the end of the year. That will take away some of the allure of college football where a loss is a huge deal.

5
by Muldrake :: Sun, 11/08/2015 - 10:38pm

I can't see enough conference support to really make home field advantage happen. I don't think the B1G would be on board because a game the second weekend in December would be really hard for virtually all the teams in the conference. You have to understand that all these stadiums are nearly a century old and aren't designed to be open in December. They're all winterized so water doesnt freeze in the uninsulated pipes. I believe only Minnesota has a heated field, and in a cold year that could be potentially dangerous to the athletes. The cold could also be dangerous to fans, and definite uncomfortable for many fans.

It may not seem like 2 weeks makes that much of a difference, but in Madison, for example, the average daily high drops 7 degrees down to 33 degrees between Nov. 26th and Dec. 12 (the day the proposed playoff would start this year). The chance of snow being on the ground increases 25% between those 2 dates. Hosting playoffs would involve a lot of investments to the facilities with no promise of ever actually hosting games.

Also, I can't see the SEC ever agreeing to the chance of having to play there either. Or the Big 12 either, really. Can you imagine the Baylor offense in snow?

I understand why everyone would like the games on campuses, but I dont think it's logistically possible. NFL stadiums are designed to play games in December in ways CFB stadiums aren't, and I can't see colleges agreeing to make them that way.

7
by Will Allen :: Mon, 11/09/2015 - 10:00pm

Well, then each conference should be allowed to designate their "home" field for the years one of their champs gets a top 4 seed, and there is no way the B1G shoud ever concede that their champs be consigned to perpetual field disadvantage, as is typically the case with so many bowl games. Good gravy, Nebraska was in the Big 12 until a couple years ago, and Lincoln can have snow and wind in November. I suppose the B1G could designate the domed stadiums in Detroit and Minneapolis, but they really should insist that Chicago, Pittsburgh, Cleveland, ,along with Met Life in Jersey, be allowed to be part of the rotation as well. Hell, Green Bay, too; how awesome would it be to see The Badgers play The Gators at Lambeau in December?

6
by tsmonk :: Mon, 11/09/2015 - 11:59am

Optimal for me would be six and top two teams getting a bye, but I wouldn't bust a gut over 8. This format of 16 looks absurd. For a significant chunk of this formatting you're looking at some really bad football.

8
by LionInAZ :: Wed, 11/11/2015 - 12:12am

The idea of figuring out a college football 'champion' out of 100+ teams that play only 12 games with few head-to-head matchups is, and always has been, a bunch of crock. Especially since the seedings are still driven mainly by poll rankings. There's about as much objectivity as picking winners in a pie baking contest.