The Games and the Committee: Every College Football Playoff Scenario

In 48 hours, some way and somehow, we will have twelve playoff teams. Will the committee survive until then? Will the committee survive the aftermath? Will the ACC?

We’ve got nine conference championships ahead of us. Six or seven of them feature potential playoff teams. By the end of tomorrow night, there’ll most likely be some hard decisions sitting in the committee’s lap. What happens on the field matters most. But, this being college football, the fate of the 17 (19?) playoff contenders ultimately rests in voters’ hands. Back in the day, it was AP voters who decided who won it all. Now, a committee of administrators selects and/or seeds twelve playoff teams.

We’re going to do four things here.

• First, I’m going to show you my seeding probability grids. These are my best guess of every team’s likelihood of landing in each individual slot, built off of our college football model, my knowledge of that model’s shortcomings, and single-game win probabilities from betting markets. (Don’t put too much stock into betting markets for playoff selection itself—the limits are low and the take is huge, so they’re not as efficient as famously efficient spreads and totals.)
• Second, we’re going to walk through the four sections of that grid, outlining what could happen in each area and how I got the numbers I got.
• Third, we’ll share our model’s FPA’s from Tuesday night’s rankings. These are just a measure of how surprising those rankings were, for those who follow our model closely.
• Fourth, I’m going to offer my own opinions on the playoff, from the high-level stuff to the minutiae. We’re leaving this last in an effort to keep opinion and analysis as distinct as we can, and to make it easy to opt out if you don’t really care about my opinion. (I get it. I’ve personally been avoiding sports media for a few weeks now because the playoff discourse gets so annoying.)

Before we do that, though, one quick reminder:

Nobody has anything to whine about.

Notre Dame would be a lock if they’d beaten Texas A&M. Miami would be in the ACC Championship if they’d beaten SMU or Louisville. Alabama would be a lock if they’d beaten Florida State. Texas would be a lock if they’d beaten Florida. James Madison would lead the race for the fifth automatic bid if they’d beaten Louisville. All BYU had to do (has to do, more accurately) is beat Texas Tech once. BYU might even be in if they’d played the Red Raiders close the first time.

The committee might get things wrong. Whether it does or doesn’t is a matter of opinion. But everybody in this thing had the opportunity to keep it out of the committee’s hands. Everybody in this had the opportunity to make the playoff by taking care of business. Notre Dame, Alabama, Miami, and Texas each lost to a team at least ten points worse than them. BYU let Texas Tech score six times before they got on the scoreboard. Nobody’s getting screwed. If they are, they screwed themselves.

Seeding Probabilities

One thing here: Numbers in yellow are the only ones that are precise. Everything else is an imprecise estimate. In reality, there’s a range around each one, but trying to reflect that imprecision only made this all more confusing.

Ok. Here’s the master grid:

Here it is broken down into sections:

As for how we got there…

The Scenarios, Part I: The Race for #2

Whoever wins the Big Ten Championship is going to be the 1-seed. That’s one of two certainties. Betting markets have Ohio State 66% likely to win as of Friday morning (all these probabilities are as of Friday morning). That leaves 34% for Indiana.

Behind them, the 2-seed will come down to the Big Ten loser, the Big 12 champion, the SEC champion, and mayyyybe Oregon. Here’s what can happen:

If Georgia wins (55% likely)

In the scenario where all the favorites win, our model has Indiana 94% likely to take the 2-seed, Georgia 5% likely, and Texas Tech 1% likely. The Texas Tech number is almost definitely wrong. It’s hard to see any scenario where Texas Tech jumps Georgia. Our model thinks it can happen if Texas Tech wins by a lot and Georgia wins close.

Should the Georgia number be higher? I’m inclined to think so. Our model doesn’t have a great gauge of how big the gap currently is between Indiana and Georgia. The committee has shown that it thinks highly of SEC teams’ body of work, though, and not particularly highly of Oregon’s or USC’s. We don’t really know how the committee feels about Indiana and Ohio State. Those two have been so far ahead of everyone else all fall (Texas A&M drew a weirdly easy SEC schedule) that we have no comparison. The Oregon and USC treatment implies the committee might not think as highly of the Big Ten loser as our model expects.

What if Indiana beats Ohio State? In that scenario, our model expects Ohio State to get the 2-seed 99% of the time. Again, I think our model’s overconfident and that Georgia has at least a realistic chance. The debates basically boil down like this:

• Georgia vs. Ohio State: Ohio State still might be the best team in the country, even if they lose what’s effectively just the de facto Heisman Championship. Both teams beat Texas, and the bulk of both games was fairly convincing. Georgia can point to wins over Mississippi and Alabama, though.
• Georgia vs. Indiana: Indiana beat Oregon, who’s ranked higher than anyone Georgia beat. Indiana lost to Ohio State, who’s ranked higher than the team who beat Georgia. Georgia, though, has won national championships recently. Indiana only technically had a football program for most of its history. That shouldn’t matter! But in a committee room that’s going to spend very little time on this part of the rankings, it probably matters.

My gut says Georgia gets the 2-seed. For the probability grid, I split the difference between my gut and our model and called it a 50/50 choice.

One thing I’m more confident about: I don’t think Texas Tech will jump Ohio State or Indiana for the 3-seed, partly because I don’t think the committee would like a bracket that had Ohio State and Indiana potentially matching up in the semifinals. When there’s a plausible way to avoid rematches, I think the committee takes that route. Not a lot of evidence to back that up, though. Just hints from 2020 and 2021, when the committee members were different anyway.

If Alabama wins (45% likely)

Our model doesn’t see Alabama jumping Texas Tech with a win, let alone Indiana or Ohio State. Here, I’m inclined to agree. Maybe Texas Tech has a chance at the 2-seed in this scenario, but I’m skeptical given the committee’s treatment of the Big 12 this year. Again, the committee won’t spend very much time on any debate over #2 vs. #3.

If Indiana, Alabama, and BYU win (3% likely)

Can Oregon get the 2-seed in a maximum chaos scenario? Our model doesn’t think so. Even if Indiana blows out Ohio State, our model expects Ohio State’s best wins (Texas, Michigan) and status as a better team on paper to outweigh Oregon’s best wins (USC, Iowa). We think our model’s right about that, and even if it isn’t, the probability of this combination of results is low enough to not worry about it too much.

The Scenarios, Part II: Byes and Home Games

Ohio State and Indiana are locked into byes, then. Georgia and Texas Tech control their bye fate. I’m comfortable guessing that the order of Oregon, Mississippi, Texas A&M, and Oklahoma won’t change. The questions that leaves:

How High Can Alabama Rise?

Where our model doesn’t have Alabama:

If Alabama and BYU both win, our model calls Oregon the likeliest team to get the 3-seed, leaving the Big Ten with a top-3 sweep. The model gives Oregon a 61% chance and Alabama a 1% chance, with Georgia staying put the next-likeliest outcome behind Oregon. More on Georgia in a moment.

Could Alabama rise to #4? Our model doesn’t really see that either. Across all scenarios, our model only sees a 1% chance Alabama gets a bye. You can win the SEC, but if you lost to Florida State, you lost to Florida State. Or so says our model.

Where our model does have Alabama:

In most simulations, our model has Alabama rising to the 8-seed, ahead of Oklahoma but still behind the 11–1 contingent. There’s a problem with that, of course. Oklahoma beat Alabama head-to-head. One solution for the committee? Go ahead with it and point out that Alabama beat Georgia twice. Another? Keep Alabama at the 9-seed even with a second win over a top-five team.

Personal thoughts:

I think our model’s being too bearish here on Bama. If I had to guess, Alabama jumps Texas A&M and Mississippi with a win. I’m giving that a 95% conditional probability. Besides beating Georgia twice, Alabama has a disappointing résumé, but those two wins over Georgia would be landmarks only Indiana could potentially match. In that 5% where Alabama doesn’t jump Ole Miss and A&M, I don’t really see Alabama jumping Oklahoma, especially if those teams are seeded 8^th and 9^th. If there’s going to be a rematch, I’d imagine the committee would at least want it in Norman this time, under some subliminal philosophy of “fairness.”

In the 95% where Alabama jumps A&M and Ole Miss, do they also jump Oregon? The wins favor Alabama and the losses favor Oregon. I’d call it 50/50 in a vacuum. Which brings us to Georgia.

How Far Can Georgia Fall?

There’s disagreement over whether conference championship losses should matter. More on that at the end. There’s also disagreement over how much they’ve mattered in the past. Last year, SMU and Iowa State each fell past one inactive team with their conference championship loss. Texas and Penn State were only passed by teams who won their conference championships. Back in the four-team days, movement was still limited, but it was a little more common in areas near the playoff field. Conference championships were understood to often be playoff play-ins. That’s no longer the case, but to an undetermined and malleable extent.

Based on this, we told our model back in August to treat conference championships like this:

• In a quarter of simulations, the loser isn’t punished at all. It’s almost like the game never happened.
• In a quarter of simulations, the loss is treated just like it would be if it was played in the regular season.
• In half the simulations, the loss is treated as half of a regular season loss.

We then told our model that these rules apply the same to all conference championships within a certain simulation. Meaning: If Alabama can drop with a loss, so can BYU.

Where does that leave Georgia? In simulations where we do allow our model to drop teams who lose this weekend, Georgia most often ends up as the 7-seed or the 8-seed. There’s a chance of falling all the way into a road game, but it’s unlikely in our model’s eyes, and it’s even more unlikely in my own eyes. Will the committee really push Georgia down past Mississippi, who they beat head-to-head?

In simulations where Georgia can’t drop, they’re sometimes passed by Texas Tech, and occasionally by Oregon (if Indiana beats Ohio State, Oregon’s résumé improves, but I don’t think the committee will notice that).

This is just a guess, but I’d imagine that if Alabama jumps Oregon, Alabama will also jump Georgia. That leaves us with three possibilities if Alabama wins: Stay put behind Oklahoma (unlikely); jump Oklahoma, Texas A&M, and Mississippi; or jump all the way past Georgia.

What about Georgia vs. Oregon, if Georgia loses? My gut says Georgia stays ahead, but I don’t have much basis for that. Our model says that if Georgia can drop, Georgia will drop, and that means falling behind Oregon. I’m calling it 50/50, with Alabama jumping Georgia in the scenarios where Georgia falls past Oregon.

That’s a long way of saying that in the 45% of scenarios where Alabama beats Georgia, I got the following numbers for the order of SEC teams plus Oregon:

• 10.7%: Alabama, Georgia, Oregon, Mississippi, Texas A&M, Oklahoma
• 10.7%: Alabama, Oregon, Georgia, Mississippi, Texas A&M, Oklahoma
• 10.7%: Georgia, Oregon, Alabama, Mississippi, Texas A&M, Oklahoma
• 10.7%: Oregon, Alabama, Georgia, Mississippi, Texas A&M, Oklahoma
• 2.2%: Georgia, Oregon, Mississippi, Texas A&M, Oklahoma, Alabama

That’s before considering Texas Tech vs. BYU.

How High Can BYU Rise?

Does a 12–1 Big 12 champion have a chance at a bye?

It depends who that 12–1 Big 12 champion is.

Our model does see some isolated cases where BYU rises to a 4-seed. In the median scenario, they’re the 8-seed if they win, jumping Oklahoma but not the 11–1 teams. That’s very believable. The bigger question is:

How Far Can Texas Tech Fall?

With the Big 12, the head-to-head question’s a little different than it is for the SEC. If Alabama beats Georgia, Alabama will have swept Georgia. If BYU beats Texas Tech, BYU and Texas Tech will have split their season series.

For our model, this means that in simulations where Texas Tech is allowed to drop, Texas Tech more often than not stays ahead of BYU. Not because our model directly considers head-to-head (Should it? We might find out!), but because Texas Tech usually beat BYU worse than BYU beat Texas Tech, which in a roundabout way leaves Texas Tech with the better overall résumé. Who knows if that’s where the committee will land, but that’s what our model expects. Our model’s median outcome if BYU beats Texas Tech is that Texas Tech is the 7-seed and BYU’s the 8-seed. Or 8 and 9, if Alabama beats Georgia. That would set up a rubber match, at which point I’d personally guess the committee would put the game in Provo because Texas Tech hosted during the regular season. This is all really hazy, but that’s the best our model can offer.

Personally, it seems like there are four possibilities for the committee here:

• Leave Texas Tech where they stand; move BYU up to somewhere in the 1-loss mix.
• Leave Texas Tech where they stand; move BYU ahead of them.
• Move BYU up to somewhere in the 1-loss mix; drop Texas Tech right behind them (I think BYU winning the Big 12 would give them the edge over Texas Tech for the committee).
• Move BYU up to somewhere in the 1-loss mix; drop Texas Tech past all 1-loss teams, and possibly past Oklahoma.

I’ve given those possibilities a 15/15/55/15 split. I’ve also given BYU an equal chance of jumping any of Oregon, Mississippi, Texas A&M, or none of the above. In the none of the above section, I’ve given BYU a tiny chance of staying behind Oklahoma and Alabama. If Alabama does beat Georgia, I’ve given BYU no chance of jumping either Alabama or Georgia. If Texas Tech does drop past all 1-loss teams, I’ve given them a 50/50 shot of dropping past Oklahoma as well.

If Texas Tech wins, could Alabama jump them? This gets back to whether Georgia can drop with a loss. I personally don’t see the committee doing it, but I’ve given that possibility a 25% chance given how down on the Big 12 the committee has been (relative to the ACC, for example). If Alabama and Texas Tech both win, I’ve given Tech a 75% chance of jumping Georgia and staying ahead of Alabama and a 25% chance of staying behind Georgia (provided they wouldn’t have to fall past Oregon), with Alabama jumping Texas Tech if Alabama jumps Georgia. That leaves us with all of this, with a few contradictions ironed out as necessary:

I promise I’m having fun here.

Who Can Get Byes? Who Can Get Home Games?

All of Ohio State, Indiana, Georgia, Texas Tech, Oregon, Mississippi, Alabama, and BYU have believable paths to byes. Those seven plus Mississippi, Texas A&M, and Oklahoma are the ten teams who could host games. Our model doesn’t see a path for Notre Dame to host a playoff game this year, and neither do I.

The Scenarios, Part III: The Bubble

Alright, let’s do this.

Two things to think about, before the scenarios:

• Head-to-head results matter to the committee, but in funky and inconsistent ways. Relevantly this year, head-to-head could matter to the committee if Miami and Notre Dame wind up next to each other. We didn’t build a contingency into our model to account for this. Our model assumes the committee will rank teams based on their whole résumés, with no single game counting more than another (with conference championships the lone exception). We might have to change that next year.
• The committee might want to avoid head-to-head rematches. As things currently stand, both the 7–10 and 8–9 games are slated to be rematches. I don’t think anyone would say this explicitly, but having tried to model these decisions for the last seven years (and having casually observed before that): The committee might move Alabama behind Notre Dame if Alabama loses, solely in order to mix up the matchups. This might have happened in both 2020 and 2021, as we mentioned above. We don’t know for sure.

If Georgia and Texas Tech both win (44% likely)

If Georgia and Texas Tech both win, our model gives the following playoff probabilities:

• Notre Dame: 97%
• Miami: 60%
• Alabama: 24%
• BYU: 19%

Astute observers will notice that Alabama and BYU are both below 25%, despite 25% of our simulations saying, “Don’t drop this team if it loses.” Theoretically, those teams should have at least a 25% playoff chance.

What gives? Well, our model’s still allowing idle teams’ résumé scores to update. And since Notre Dame played Boise State (and Miami played Louisville who played James Madison, not to mention the dozens of other second-level connections between teams’ schedules), Notre Dame and Miami’s résumés could still improve or get worse. More often than not, they improve.

If that doesn’t make sense, just pretend our model says: Notre Dame 100%, Miami 50%, Alabama 25%, BYU 25%. That’s close enough for our purposes.

Zooming out from our model and considering the whims of buffet-fed middle-aged professionals in button-downs…Sure, it’s possible that Alabama could stay ahead of Notre Dame, BYU could drop past Miami, and the committee could say that Miami’s head-to-head win over Notre Dame is now a trump card. This scenario is most relevant if Alabama loses close and BYU loses big. Will the committee do it? It’s so hard to say. Hunter Yurachek’s comments plus all the rankings to date imply that the gap between Notre Dame and Miami is a lot larger than the gap between Notre Dame and Alabama. That should imply that Alabama won’t stay ahead of Notre Dame with a loss, even a close loss, rendering the head-to-head question mostly irrelevant. If Alabama and BYU both lose, Notre Dame should be in, and it’s just a question of whether Alabama stays ahead of Miami. But, it’s easy to see the committee being impressed if Alabama loses close.

For whatever it’s worth, our model assigns the following probabilities if Alabama loses by one score and BYU loses by two or more possessions:

• Notre Dame: 97%
• Miami: 59%
• Alabama: 33%
• BYU: 11%

What we’re really seeing here is our model’s unrealistic but reasonable assumption about the opposite scenario: If BYU loses close and Alabama gets blown out, our model sees a chance that BYU makes the playoff and Alabama doesn’t. That seems prohibitively unlikely, but it’s what our model sees.

The bottom line here, again: If Alabama and BYU both lose, Notre Dame is probably in. Miami’s chance is better than 50/50, but not by a lot.

Personally, I think the situation is unknown enough to put a lot of 50/50 probabilities on it. Let’s assume a…

• 50/50 chance Alabama drops past Notre Dame with a loss.
• 50/50 chance Alabama drops past Miami if they do drop past Notre Dame.
• 50/50 chance BYU drops past Miami with a loss. (100% if Alabama drops past Miami.)
• 50/50 chance Miami jumps Notre Dame if the teams are next to each other.

We’ll carry this into the next three combinations too, but in this scenario, that gives Notre Dame an 87.5% chance of making it, Alabama a 75% chance of making it, and Miami a 37.5% chance of making it.

If Alabama and Texas Tech win (36% likely)

Here’s the nightmare scenario for Notre Dame. If conference championship losers can drop, BYU can fall past Miami here, allowing Miami to be compared one-on-one versus Notre Dame with only one spot available.

Our model doesn’t see any way Miami gets in through this scenario. 0%, our model says. I don’t think our model’s correct about that. But I will say: The committee might still decide that Notre Dame’s overall body of work is better enough than Miami’s to justify inclusion despite the head-to-head loss. The story is this: Notre Dame’s got a better résumé. Right now, Notre Dame’s the better team. Even in August, Notre Dame and Miami were extremely close to equivalent, considering what we know about home-field advantage and the fact Miami was the home team. But Miami beat Notre Dame head-to-head, and if no other teams would be affected by moving Miami ahead of Notre Dame, it seems likely enough to me that the committee would do that.

Using our probabilities from above, this leaves us with a 75% chance of Notre Dame making the playoff and a 25% chance of Miami making the playoff. In half our scenarios, BYU stays ahead of Miami under the “conference championship losses shouldn’t count” idea those with ACC interests championed last year. In the other half, Miami jumps BYU, but in only half of those does Miami jump ND. Is 50/50 the right number here? I don’t know. These are estimates.

If Georgia and BYU win (11% likely)

This creates a three-way race for one spot, with Alabama, Notre Dame, and Miami the contenders in said race. The committee’s options (for the most part—they could theoretically put Texas Tech or BYU between these teams, but that mostly aligns with this first option):

First, they could drop Alabama past Notre Dame but keep Alabama ahead of Miami. This would avoid the head-to-head question and stay fairly in line with what precedent exists: Teams can drop with a conference championship loss, but how far they drop is limited. Our 50/50 assumptions label this 25% likely.

Second, they could drop Alabama past Notre Dame and Miami, leaving us back in the head-to-head debate outlined above. Our 50/50 assumptions label this 25% likely, so 12.5% each for ND and Miami.

Third, they could just keep Alabama ahead of Notre Dame. Our assumptions label this 50% likely.

For what it’s worth, our model has Alabama 10% playoff-likely and Notre Dame 90% playoff-likely in this scenario. These résumés are really close, and our model does expect Boise State to beat UNLV, which should theoretically help Notre Dame.

If Alabama and BYU win (9% likely)

This one’s simple. Notre Dame and Miami are both out and we get a manufactured “Will Texas Tech miss the playoff??” argument that’s very unlikely to sway the committee.

Scenarios, Part IV: Automatic Land

Alternate title:

Is Duke going to make the playoff?

At the very beginning of this post, we mentioned that there might be playoff contenders this weekend. What we meant by that is that it’s theoretically possible that 8–5 Duke is viewed as a worse option, for the committee, than 11–2 UNLV or 10–3 Boise State. Basically, if James Madison loses to Troy we unlock a new level of potential chaos. But we’ll get to that.

If Virginia and Tulane win (29% likely)

It’s Virginia and Tulane.

If Virginia and North Texas win (34% likely)

It’s Virginia and North Texas. Our model does see a 4% chance that JMU makes it in this scenario—that’s in simulations where JMU wins big and UNT wins close—but we wouldn’t buy those odds. Or any odds on JMU jumping the American champion.

If Duke and James Madison win (34% likely)

It’s almost definitely the champion from the American and then James Madison. So long as JMU wins, our model doesn’t see any way Duke jumps the Dukes.

If Duke and Troy win (3% likely)

Yikes!

One spot here definitely goes to the champion of the American. The other spot could go to Duke or to the Mountain West champion (or, theoretically, to Troy or the MAC champion or the Conference USA champion, but our model doesn’t see those teams as anywhere near UNLV or Boise State).

If the committee was ranking a Top 53 instead of a Top 25, our model thinks Duke would be 40^th, UNLV 45^th, and Boise State 53^rd heading into the weekend. In this scenario, Duke’s win over Virginia would be better than even UNLV’s road win at Boise State. Duke would be a power conference champion, which historically has mattered a great deal as a checkbox on a résumé. All of that signals to our model that Duke has a chance if JMU loses, and that UNLV and Boise State do not. Is that correct? There’s only one way to find out.

FPA’s

Moving on to what our model thinks of the rankings:

Heading into Tuesday night, our model expected Miami to jump Alabama but not jump Notre Dame. Instead, as you probably know by now, Alabama jumped Notre Dame and Miami stayed all the way behind BYU. Where does that leave us?

For the uninitiated, Ranking Score goes from 0.0 to 100.0, with 0.0 the worst résumé in the FBS (UMass) and 100.0 the best. FPA stands for Forgiveness/Punishment Adjustment. FPA’s basically how far the rankings stand from our model’s expectations. “New FPA” is this week’s FPA. “Total FPA” is what it sounds like.

To answer basically the only question about this week’s rankings: Yes, the Alabama/Notre Dame thing is weird. Had Alabama entered the week ahead of Notre Dame, it would not have been weird. But Alabama jumping Notre Dame after…that? Weird.

The only explanation that makes sense to me is that some committee members want to reserve the right to put Alabama in if they play Georgia tough tomorrow and BYU beats Texas Tech.

Fair enough, I guess.

The Subjective Part

This is a murderously long blogpost, so if you’re still reading, you probably care about college football. I also care about college football, in case that Big 12–SEC probability grid above didn’t give that away.

The way I see it, there are six areas of debate right now in the CFP discourse. Some of these are always there. Some are unique to this season. We’ll go from the biggest-picture to the smallest.

Area 1: Best vs. Most Accomplished

Should the best teams make the playoff, or the most deserving? The committee says “best,” and has used this as a tiebreaker in the past. In practice, though, it’s more about who’s accomplished, with bonus points for dominance.

I don’t have a problem with accomplishment plus dominance. I don’t really like that the committee says “best,” especially since that isn’t really how they do it, but their effective criteria are fine to me. I think the bigger issue is consistency.

Area 2: Committee vs. Computers

Once upon a time, computers decided who played for the national championship, and everybody threw a fit.

There are problems with using formulas to rank teams. Among those:

• There are a zillion ways to do it. I’m pretty sure even Strength of Record, which ESPN always promotes as something cut and dried, is dependent on the FPI ratings of a team’s opponents. Is that concept the best concept? Is FPI the correct measuring stick?
• A lot of the better ways to do it partially rely on some sort of preseason rating. The college football season’s sample size is too small to nail down how much tougher an ACC team’s schedule is than a team’s from the Sun Belt, or vice versa. If I remember correctly, a desire to not use any preseason priors was part of what made some of the BCS systems so funky. It feels unfair to start teams with some sort of head start, but it’s more accurate. This is inconvenient.
• Any computer ranking has to either include margin of victory (and incentivize blowouts) or not include margin of victory (and ignore dominance). The former approach is disliked. The latter approach opens the door to some frustrating rankings.
• The drama happens in a different way. For as obnoxious as the “What will the committee do?” graphics get during late-game timeouts, a live tracker of each team’s Résumé Score might be even worse.
• There’s no room for nuance. So if you think that injuries or consequential bad calls should matter, or that conference championship results should matter in specific contexts, or that head-to-head results should matter when teams are close to one another in the rankings? No computer system worth anybody’s time is going to give you the rankings you want. I’d personally probably be fine with this, but I saw an interview this week where someone—I think it was Manny Diaz—both asked for computers to take over and said Miami should be ranked ahead of Notre Dame. Sorry, buddy. That’s not what the computers are going to do.

Right now, ESPN’s SOR has BYU four spots ahead of Texas Tech, and it has Texas A&M still third overall, and it has Alabama one spot ahead of Oklahoma and Notre Dame one spot ahead of Miami. If you think people get mad at the committee…wait until you see what the computers do.

Considering all of this, I’m again fine with the current setup in practice. I wish criteria were better defined. I wish there wasn’t so much politicking. I wish the committee were more consistent. But a committee who does a fine job (and I’d argue these committees do a fine job annually) is better than even ESPN’s SOR, which I do think has the right idea in mind.

Area 3: Is Twelve Teams the Right Size?

Every major American sport right now is fighting some version of the same war, a war between its short-term financial interests and its long-term health. For college football, that partly takes the shape of playoff expansion.

The biggest downside of an expanded playoff can be, ironically, that it gets too good. This happened in college basketball, where the NCAA Tournament became so much fun that it started rendering the regular season kind of meaningless. There’s a reason 95% of college basketball coverage in early March revolves around who’s on the bubble rather than who might win the national championship. There’s a reason Steve Sarkisian’s talking points this week implied Texas would rather marginally increase its future playoff probability than ever play a fun nonconference game again.

I don’t think Sark was fully serious, but there’s a kernel of truth in what he said. It’s why Penn State played the nonconference schedule this year that it did. Fans are so playoff-focused that they’d rather watch their team play East Texas A&M than go on the road in nonconference. That’s sad.

On a parallel note, expand the playoff too far and certain games stop mattering very much. Like, for example, tomorrow’s Big Ten Championship. Sure, a trip to the Rose Bowl is fun. But looking at the potential seedings…I’d rather play Mississippi over just about anyone else in the top ten right now, and the likeliest scenario is that the loser of the Big Ten Championship plays Mississippi in the quarterfinals. Why risk injury or show any creative looks if you’re Ohio State or Indiana? You might be better off losing.

Two years ago, Texas made the College Football Playoff because it played Alabama on the road in a nonconference game. Two years ago, every power conference championship was a potential playoff play-in. Yes, that system resulted in a really painful squeeze which left undefeated Florida State out of the final four. But it encouraged big swings and it made conferences matter. College football shouldn’t risk getting too far from those pieces of its identity.

Personally, I like the twelve-team playoff just fine. I’d prefer the four-team system (or in an ideal world, a free-for-all where teams schedule each other until we get a consensus national champion), but I’m fine with twelve teams. I’m scared of going larger, though. I’m really scared of anything beyond 14 teams.

Area 4: Should conference championship losses count? And should teams be forgiven if they lose a big nonconference game?

These are tricky for a lot of the reasons we’ve already said. Twelve games is a small sample. Big nonconference games are important for making college football fun. Conference championships should matter.

I will say: Conference championship games are relatively new. Those aren’t sacred. If a league wants to get rid of its conference championship, by all means, they should go ahead and do that. Conference championships themselves are the things that should matter. Those are the sacred part. Ohio State should care about winning the Big Ten (and I do think it does). SMU should care about winning the ACC (not as sure about that one).

Ultimately, though, conferences don’t want to get rid of their championships. Part of this is financial, but there’s also an element here common to Texas playing Alabama, Michigan, and Ohio State: When you have a chance to lose something, you also have a chance to win something. Last year when SMU was on the brink of missing the playoff, those with ACC interests advocated for a system where Clemson could be rewarded for winning the ACC Championship but SMU couldn’t be punished for losing it. Right now, Sark’s advocating for similar treatment for Texas: All the upside of winning in Tuscaloosa but none of the downside of losing in Columbus. Maybe this is fine with you. It doesn’t sit right with me. Texas absolutely deserves credit for its schedule this year. It really would have been sad for SMU to miss the playoff last year. But there was upside in both those games, and conference championships are part of the schedule. Everyone bought their ticket.

Every game should count.

Area 5: How important should head-to-head results be?

Going off of that!

There was a week earlier this season where all of Virginia, Louisville, Miami, Notre Dame, USC, and Michigan had two losses. Virginia had beaten Louisville, who’d beaten Miami, who’d beaten Notre Dame, who’d beaten USC, who’d beaten Michigan. Should they have been ranked in that exact order? Of course not. Virginia’s ranking shouldn’t set a de facto ceiling for Michigan. Michigan’s shouldn’t set a de facto floor for Louisville.

If two teams’ résumés are really, really close to one another and they played head-to-head, then sure, let head-to-head be a tiebreaker. But every game should count. USC shouldn’t see 90% of its season thrown out the window because it has the same record as Notre Dame and lost to them head-to-head. USC really shouldn’t see 90% of its season thrown out because Virginia beat Louisville.

This gets at a bigger issue with college football media, which is that everyone is tired and the pay isn’t great and there’s a little too much going on to easily wrap your head around it all, so talking heads make full-chested takes without understanding the full picture. Plus, in a media economy that rewards people for being a little dumb, it pays to wail in the streets about Miami and Notre Dame, no matter how silly your argument is when it’s given more than two minutes of examination.

Every game should count.

Area 6: Oklahoma vs. Alabama vs. Notre Dame vs. BYU vs. Miami

As we said up top, whoever doesn’t make the playoff has nothing to complain about. I understand pre-selection politicking. I hate it, but I understand it. I understand disappointment on Sunday. Hopefully it stops by Monday morning. Hopefully everyone handles this like they have some self-respect.

Personally, I think the best way to compare résumés is to look at each team’s twelve (or thirteen) performances lined up from best to worst. Every game should count, right?

Every loss in this group, ranked from most embarrassing to least:

• Alabama lost by two touchdowns at Florida State, a mediocre team.
• Miami lost at home to Louisville, who is unranked.
• Miami lost to SMU, who is unranked.
• Oklahoma lost by three scores to Texas, who’s out of the playoff discussion.
• BYU got blown out by Texas Tech.
• Oklahoma lost at home to Mississippi.
• Alabama lost at home to Oklahoma.
• Notre Dame lost at home to Texas A&M.
• Notre Dame lost to Miami.

Every big win in this group, ranked from most impressive to least:

• Alabama beat Georgia on the road.
• Oklahoma beat Alabama on the road.
• Miami beat Notre Dame.
• Alabama beat Vanderbilt by 16 points.
• Notre Dame beat USC by ten.
• BYU beat Utah.
• BYU beat Arizona on the road.

As far as the bulk of each team’s performance goes…

• Notre Dame has been the most consistently dominant, but their schedule isn’t impressive.
• Miami has been a little less dominant against a little worse of a schedule.
• Alabama has been occasionally dominant, and they’ve played the toughest schedule in the group.
• Oklahoma has been occasionally dominant (in a defensive way), and their schedule’s been almost as good as Alabama’s.
• BYU hasn’t been particularly dominant, but they have one more win than everybody else. Their schedule’s closer to Alabama’s and Oklahoma’s than Notre Dame’s and Miami’s.

What do I make of all of that?

On the extreme ends of the list (the best wins and the losses), BYU has a narrow edge, but if they get blown out again, that edge disappears. Both Oklahoma and Notre Dame’s lists are a wash. Alabama has the worst loss, but also the best win, which combines with a second quality win to make their list roughly a wash as well. Miami is clearly behind everybody else, with only the one good win and with two of the three worst losses of the bunch.

In the middle, I’d take what Alabama and Oklahoma have done over what Notre Dame, Miami, and BYU have done, especially if BYU does lose tomorrow (if they win, they’re in the playoff as an automatic bid anyway). Basically, I’d give Alabama and Oklahoma an 8 out of 10 in the middle, and I’d give Notre Dame and Miami and BYU each a 7 out of 10.

Add the middle to those extreme ends of the list, and Miami gets taken out of the picture while Oklahoma gets bumped safely into the playoff. That leaves Alabama, Notre Dame, and BYU. If Alabama gets blown out, they should be out. If BYU gets blown out again, they should be out. If BYU plays Texas Tech close, they’d probably get my vote. If Alabama doesn’t get blown out, I don’t mind Alabama making it ahead of Notre Dame, though I also don’t mind Notre Dame making it ahead of Alabama. Notre Dame chose to play five ACC teams a year. Notre Dame lost a home game to a team they should have beaten. Alabama lost to Florida State by two touchdowns, a loss almost as bad as what Notre Dame did last year against NIU. Each still has a body of work that’s clearly more impressive than Miami’s.

**