2015 Death to the BCS Playoffs: Official Bracket

With the exception of the Army-Navy game, the 2015 regular season is in the books. Now, it's for all the marbles.

This weekend, two conferences concluded their regular seasons while the other eight played their conference championship games. The former all had their automatic bids claimed; the latter needed to make it official. Now that they are, it's time to go to the Playoff Laboratory one final time... only this time, I'm doing this for real.

Those ten aforementioned conference champions all received an automatic berth to the 16 team field, while the remaining six spots were awarded at large to deserving teams who couldn't claim their conference crown. So how were those six decided, and how did I work out where everyone should be seeded? That latter question is just as important as the former; higher seeds get home field advantage for the first three rounds.

As for the selection criteria, I am a one-man selection committee who builds a bracket with the help of some metrics. I will use my own Non-Conference Schedule Strength and Playoff Points metrics to get started, and then I follow up with a pair of computer rankings that incorporate margin of victory (because that matters). These rankings are compiled by Jeff Sagarin and by a UCLA faculty member who uses the publicly available formula created by the late David Rothman. I try to combine these factors together to not be biased towards my own metrics while also looking at standings to come up with a playoff field.

Seeding is a little trickier. I'm using the same criteria to assign seedings, but I also want to keep a couple other factors in mind. I want to minimize travel to the best of my ability. Some of that will be unavoidable and TV revenues would make up for travel costs, but at the same time the travel costs would be significantly less than they would for a bowl game. This matters. The other, more important factor is one that I've kept in mind over the years: I want to try to avoid rematches as much as possible. This is vital in the first round, and it goes down in importance as the tournament progresses, but it remains in the back of my head.

So, fourteen weeks of football have all led up to this. Here is the official bracket for the 2015 Death to the BCS Playoffs.

1. Clemson (13-0, ACC Champion)- NCSS: 4, PP1: 68, PP2: 24.08, SAG: 3, ROTH: 3 (LW: 1) 
2. Alabama (12-1, SEC Champion)- NCSS: 3, PP1: 72, PP2: 30.27, SAG: 1, ROTH: 1. (LW: 2) 
3. Michigan State (12-1, Big Ten Champion)- NCSS: 6, PP1: 73, PP2: 30.00, SAG: 9, ROTH: 5. (LW: 3)
4. Oklahoma (11-1, Big 12 Champion)- NCSS: 5, PP1: 73, PP2: 25.64, SAG: 2, ROTH: 2 (LW: 5)
5. Ohio State (11-1, At Large)- NCSS: 6, PP1: 63, PP2: 23.18, SAG: 4, ROTH: 4. (LW: 6)
6. Stanford (11-2, Pac-12 Champion)- NCSS: 5, PP1: 59, PP2: 30.10, SAG: 5, ROTH: 9. (LW: 7)
7. Iowa (12-1, At Large)- NCSS: 5, PP1: 57, PP2: 20.75, SAG: 13, ROTH: 8. (LW: 4) 
8. Notre Dame (10-2, At Large)- NCSS: 26, PP1: 56, PP2: 26.40. SAG: 8, ROTH: 7. (LW: 8) 
9. Florida State (10-2, At Large)- NCSS: 4, PP1: 53, PP2: 20.70, SAG: 6, ROTH: 15 (LW: 9)
10. Oklahoma State (10-2, At Large)- NCSS: 2, PP1: 48, PP2: 19.20, SAG: 16, ROTH: 10 (LW: NR) 
11. TCU (10-2, At Large)- NCSS: 3, PP1: 44, PP2: 17.80, SAG: 7, ROTH: 6 (LW: 11)
12. Houston (12-1, American Athletic Champion)- NCSS: 5, PP1: 50, PP2: 17.45, SAG: 27, ROTH: 13 (LW: 12)
13. Western Kentucky (11-2, Conference USA Champion)- NCSS: 10, PP1: 50, PP2: 15.00. SAG: 40, ROTH: 46. (LW: 13) 
14. Bowling Green (10-3, MAC Champion)- NCSS: 9, PP1: 41, PP2: 15.78. SAG: 41, ROTH: 36. (LW: 14)
15. Arkansas State (9-3, Sun Belt Champion)- NCSS: 6, PP1: 37, PP2: 14.22. SAG: 70, ROTH: 65. (LW: 16)
16. San Diego State (10-3, Mountain West Champion)- NCSS: 6, PP1: 35, PP2: 13.44, SAG: 55, ROTH: 70. (LW: 15)

Out of the playoffs: Florida (10)

To view the official 2015 Death to the BCS Playoffs Bracket, click here.

Photo by Joshua S. Kelly (USA TODAY Sports)

I only bumped one team from the playoffs, as Florida suffered its third loss of the year, and Oklahoma State took their place based on a better resume, while they also have the distinction of a win over TCU, who is in the playoffs. It seemed like the right thing to do. At the same time, it's tough to kick Florida out given their higher Playoff Points totals, but the computers like Oklahoma State more, and I felt like I was snubbing the Cowboys more by leaving them out. Florida also shouldn't have gotten killed by the Seminoles last week. Had they won, we wouldn't be having this argument. For those of you wondering about North Carolina: they were disqualified from consideration for the Death to the BCS Playoffs a couple months ago due to playing two FCS teams this season. They weren't the only team disqualified for such, but they were the only team in the running that this rule impacted. Had they played only one, they probably make it in over Florida State or possibly one of the other at large teams.

Now, let's go over seedings. You have the eternal debate between Clemson and Alabama for the top overall seed, but I stick by what I've been saying for weeks: Clemson is undefeated, Alabama is not. The Tigers get the top overall seed, while the Crimson Tide still get a third home game. Alabama is a unanimous number one according to the computers, and is second in PP1 and top five in PP2.

Photo by Michael Conroy (AP)

I knew who to give second home games to: Michigan State and Oklahoma (see, Sooner fans? I told you all would be well). This was an incredibly tough decision; Oklahoma has the edge in computer rankings, while Michigan State has a higher PP2, a "better" loss (by one to Nebraska, who has a higher PP1, PP2, and Sagarin ranking than Texas, who beat Oklahoma). Both teams are also 2-0 against other playoff foes, so it's close, but I'm sticking with the Spartans for the advantages listed. Ultimately, it's not a huge difference, since both teams get to play at home the first two weeks, but the Spartans will, at least theoretically, get an easier road to the semifinals.

Photo by Jason O. Watson (Getty Images)

This leaves four other teams who will get to host first round games. I had a pretty tough call to get the 5-8 seeds started, especially with rematches in mind as I had my first three teams in that group an easy call to make: Stanford, Ohio State and Iowa. Based purely on numbers, Stanford would be the #5 seed despite two losses; the computers love them and they have among the highest PP2's in the country. However, putting them at the #5 seed would mean Michigan State would theoretically host either Ohio State or Iowa in Round 2, both of whom lost to the Spartans. This, kind of unfairly, pushes the Cardinal down to the #6 seed to avoid either rematch, but it's doable when I look at the numbers. Ohio State is loved by the computers and has the highest PP1 out of that group. They will receive the #5 seed, which drops Iowa down to #7.

Photo by Matt Cashore (USA TODAY Sports)

Deciding on the #8 seed was also incredibly hard, with four or five teams fighting for that last home game. Notre Dame was the #8 seed last week, and they have the best PP1 and PP2 of all remaining teams. The Irish also rank near the top in both computer rankings among all remaining teams with the exception of TCU and Florida State in the Sagarin rankings. The Irish would normally run away with this playoff spot, but giving them this game would send them to Clemson again with a win in their first game. I thought about bumping them for my next highest team in Florida State, but that doesn't solve the rematch problem. That would mean moving either Oklahoma State or TCU into that #8 spot. The computers adore TCU, but I can't in good conscience put them ahead of Oklahoma State because of the head to head result. Oklahoma State got blown out by Oklahoma though, while TCU made a game of it in Norman, losing by only one. With the Cowboys also losing to Baylor at home, I just don't see enough evidence to avoid a rematch. Touchdown Jesus gets a home game, and if that means a Round 2 rematch, so be it.

Photo by Stephen M. Dowell (Orlando Sentinel)

The remaining teams can be split in half to get a clear 9-12 and 13-16, but the seeding question remains. With Florida State being the best remaining candidate (TCU's computer rankings notwithstanding), the Seminoles get the #9 seed. This guarantees Clemson a rematch in Round 2, and while I still don't like the Seminoles' loss to a bad Georgia Tech team, I like Florida State's resume over Oklahoma State's. This puts the Cowboys as the #10 seed, since they have the head to head win over TCU, and TCU's computer rankings give them the edge for the #11 seed over Houston. This may seem like elitism, as the Group of Five conferences have all their auto bids at the end of the bracket, but the numbers back it up.

Photo by Joshua Lindsey (USA TODAY Sports)

That leaves the four remaining conference champions to round out the bracket. I'm sticking with last week's rationale in leaving Western Kentucky ahead of Bowling Green: an extra win, arguably the "better" win between the two teams (over Louisiana Tech), and a higher PP1. While the computers lean Bowling Green, I'm going to stick with the Hilltoppers. Geographically, this minimizes travel for both teams as well, since Bowling Green is better in every metric than the two remaining teams. Meanwhile, Arkansas State finished off its schedule while San Diego State won its conference championship game. Now, the Red Wolves are better in terms of Playoff Points, while the Aztecs hold the edge according to the computers. Travel is negligible here, and there's no rematch in play. For this, I go to a common opponent: Arkansas State beat South Alabama, while San Diego State lost that game. This gives the Red Wolves the 15 seed and a trip to Tuscaloosa, while the Aztecs will face Clemson in Round 1.

Looking at the handing out of berths, the Big Ten and Big 12 each have three teams in, the ACC got two, and Notre Dame made it as an independent. SEC fans are probably screaming at me over only seeing Alabama in, but other than Florida, no other teams were good enough; everyone else finished 9-3 at best. That's not good enough, and blows a gigantic hole in the argument that a 16 team playoff dilutes the regular season. Look at those at large teams again, and tell me none of them deserve to be here. If you have a compelling argument, feel free to leave a comment below.

The bowl games officially kick off on Saturday, December 19th, so that is when the Death to the BCS Playoffs will officially kick off as well. I've scheduled the games to be spaced out pretty much all day (all times listed are in Central time) for 12 hours of football while minimizing jet lag for teams. There will be competing games on during this first week, but I'll follow the same formula in Round 2 and we won't see any schedule conflicts here with multiple playoff games going on at the same time.

This tournament will be somewhat interactive. If you want to predict the results of the 2015 Death to the BCS Playoffs, you can create a free BracketMaker account and follow this Predict It! link. For now though, this is it while teams take a couple weeks to rest up/my simulator gets the 2015 teams entered. To those of you whose teams made the playoffs: Good luck to you!