College Football: How a 16-Team FBS Playoff Might Work

Use your ← → (arrow) keys to browse more stories
College Football: How a 16-Team FBS Playoff Might Work

I'm going to leave my anti-bowl rant for later in the week, as the 20th approaches. Not that the bowls shouldn't exist; after all, 6-6 teams need some reason to feel good...

No, that's not it. It's the excitement and the drama of a knockout playoff to bring the best teams in the nation to fight it out, unshackled by conference affiliations and (after the first round kicks off) untainted by the "beauty contest" aspect of the polls.

Leave it all on the field.

 

16-TEAM PLAYOFF BRACKET (seeded)

BCS conference winners seeded 1-6 according to BCS ranking of Week 15 (Dec. 7);
Non-BCS conference winners seeded next according to BCS ranking (if in Top 25);
Remaining teams considered “at-large” and seeded according to BCS ranking.

  1. Oklahoma (BCS 1)
  2. Florida (BCS 2)
  3. USC (BCS 5)
  4. Penn State (BCS 8)
  5. Cincinnati (BCS 12)
  6. Virginia Tech (BCS 19)
  7. Utah (BCS 6)
  8. Boise State (BCS 9)
  9. Texas (BCS 3)
  10. Alabama (BCS 4)
  11. Texas Tech (BCS 7)
  12. Ohio State (BCS 10)
  13. TCU (BCS 11)
  14. Oklahoma State (BCS 13)
  15. Georgia Tech (BCS 14)
  16. Georgia (BCS 15)


Team “bumped” from the playoff due to Virginia Tech's inclusion: BYU (BCS 16).

Buffalo (MAC), East Carolina (C-USA), and Troy (Sun Belt), the three remaining FBS conference champions, are not in the BCS Top 25 and are thus ineligible for the playoff.

Other rankings could conceivably be used, but the AP has already asked out of being used to determine participants in the BCS and is likely to do so with the playoff scenario as well.

Using the final pre-bowl BCS ranking represents a compromise to ensure the BCS conferences retain some measure of influence in the process (though not overwhelmingly so as they do now). The seeding of the six BCS conferences as 1-6 in the bracket is the only other concession.

This sort of playoff may come about as the result of political action by the United States Congress, financial incentives from major media interests (such as ESPN, CBS College Sports, etc.), and public pressure from sports fans disappointed with the purgatory that is the BCS—neither the status quo ante (a dozen "big" traditional bowls, with no attempt to create a playoff), nor the utopian "pure" playoff system free of any BCS involvement (seeding without regard to BCS conference affiliation).

What this represents is politics as the art of compromise, and sport as the art of  competition. There are fairer and finer ways of bringing the FBS college football season to a close, and this is one small contribution to that end.

So, in addition to my previous article (one of my first on Bleacher Report) outlining the possibilities for a pragmatic, workable, eight-team playoff, I shall now present an equally pragmatic, workable example of a 16-team playoff.

FIRST ROUND (Sweet Sixteen): Dec. 26-27

Higher seeds are “home” teams, though there are no "home" games per se; the location of each group is set by the 1-4 seeds (with attention paid to pre-existing conference affiliations with certain bowl sites when relevant).

It is possible for those who wish to consider the possibility of seeds 1-8 hosting first round games to imagine them doing so, with only the Quarterfinal games ("Elite Eight") taking place at the four bowl sites. However, it is very likely that minimizing travel costs between games will be a more important factor.

Therefore the "@" should be taken as a technical designation only (for uniforms, etc.).



Group A (Fiesta Bowl)
16 Georgia @ 1 Oklahoma
9 Texas @ 8 Boise State

Group B (Orange Bowl)
12 Ohio State @ 5 Cincinnati
13 TCU @ 4 Penn State

Group C (Rose Bowl)
14 Oklahoma State @ 3 USC
11 Texas Tech @ 6 Virginia Tech

Group D (Sugar Bowl)
10 Alabama @ 7 Utah
15 Georgia Tech @ 2 Florida

 

QUARTERFINALS (Elite Eight): Jan. 1

Winners of the first round games meet at the same bowl site within the week—hence Dec. 26 through Jan. 1, which is seven days. When Jan. 1 falls on Sunday, the first round games move to Dec. 27-28 and the quarterfinals to Jan. 2. The traditional date of Jan. 1 (or 2) for the four BCS bowl sites is preserved.

Ideally the date would be left to these four bowl games, but it may be difficult to force the non-playoff bowls to move to a different date.

 

SEMIFINALS AND FINAL (Final Four): Jan. 7 and Jan. 12

The winners of the four quarterfinal games then advance to a 5th site for the “Final Four”—for example, the Edward Jones Dome in St. Louis (pictured at the top of the article).

Note that the 15 playoff games take place at five sites: the four BCS bowl sites and a fifth, neutral site that may change from year to year (as the NFL Super Bowl and the NCAA Final Four), or be set in one location (as the NCAA College World Series).

Each of the five sites hosts three games. This represents a bonus for the four BCS bowl sites, as they host one currently (and the title game every fourth year).

The concessions *by* the BCS (the concessions *to* the BCS have already been mentioned) are the addition of the fifth site, presumably to be organized in full or in part by the NCAA (not the BCS conferences), with the winner of the final being officially recognized by the NCAA and presented with an NCAA trophy (this is not the case currently).

 

What follows is a simulation run on “ESPN College Football Playoffs Simulation (powered by AccuScore)”: http://proxy.espn.go.com/ncf/bowls08/bracket

The simulation was run only once, and for those who question the results, note that the percentages given after the score are the probabilities of each team winning based on 10,000 (!) simulations of every possible matchup among the teams allowed in the simulation. Note that only the Top 25 teams in the final pre-bowl BCS rankings are available for the simulation.

Where the probabilities (percentages) of a team's chances of winning do not match the result, an upset has occurred, and I've labeled it as such.

I could run the simulation with these particular 1-16 seeds over and over, and get different results each time. If I were to run it 10,000 times, then the "average" results of the 10,000 times would more closely match the probabilities, and the upsets would disappear into those "average" results. Of course, the games are played only once.

 

FIRST ROUND (Sweet Sixteen):

Group A:
Oklahoma 46, Georgia 29 (79% - 20%)
Boise State 41, Texas 25 (18% - 81%, upset)

Group B:
Ohio State 28, Cincinnati 17 (69% - 31%)
Penn State 27, TCU 20 (59% - 40%)

Group C:
Oklahoma State 35, USC 29 (41% - 58%, upset)
Texas Tech 39, Virginia Tech 27 (67% - 32%)

Group D:
Utah 32, Alabama 19 (21% - 78%, upset)
Florida 50, Georgia Tech 24 (87% - 12%)



QUARTERFINALS (Elite Eight):

Group A: Boise State 53, Oklahoma 31 (16% - 83%, upset)
Group B: Ohio State 34, Penn State 28 (38% - 61%, upset)
Group C: Texas Tech 43, Oklahoma State 39 (52% - 47%)
Group D: Florida 48, Utah 25 (81% - 18%)

 


SEMIFINALS AND FINAL (Final Four):

Semifinal 1: Ohio State 31, Boise State 22 (61% - 38%)
Semifinal 2: Florida 46, Texas Tech 32 (67% - 32%)

Final: Ohio State 42, Florida 27 (24% - 75%, upset)



***IMPORTANT*** Note that the locations of the games are not specified in the simulation and thus do not have an effect. We can assume that, in the simulation, all games take place at a “neutral” site (i.e. in cyberspace, or on the ESPN or AccuScore servers). In my playoff model on the previous page, I assign the games to specific locations, but do not project the results.

So... about this idea for a sixteen-team playoff...

Is it perfect? No. Nothing on this Earth is perfect.

Is it better than the "Broken Championship System?" I'd like to think so. Let me know what you think, both of the prospective matchups created by the seedings and of the playoff model in general. And feel free to try running simulations of this playoff model or another one of your choice.

Load More Stories

Follow B/R on Facebook

College Football

Subscribe Now

We will never share your email address

Thanks for signing up.