College football has seen its fair share of underdog stories over the years. But too many of them occurred in September or in bowl games that counted for little more than pride. At long last, we’re on the brink of a new era. On Sunday, a genuine playoff bracket will be revealed, the first of its kind in FBS history (sorry, a four-team invitational doesn’t count).

With it comes the introduction of one of the most compelling aspects of sports: the chance for meaningful upsets. And, per our past research charting commonalities from upsets across sports, the most likely team to pull off one of those upsets (should they make the field) is Alabama.

Upsets are our niche, dating all the way back to 2006, when we launched the Giant Killers model that projects NCAA Tournament upsets (you know it as Bracket Breakers now). Over the years, we have expanded our reach to identify worthy underdogs in events ranging from the World Cup to the Olympics to the NFL. But all of those competitions had historical data we could analyze in our search for trends.

It will take time to build a robust model unique to the brand new College Football Playoff. (How much of an edge does a first-round home game provide? How much does an underdog’s conference matter in its chances of pulling off an upset?) In the meantime, we can still apply what we’ve learned about upsets in other sports, starting with three key premises that have proven true in just about any sport we have studied.

1. Some underdogs are underrated and under-seeded

Find teams whose underlying statistical strengths outweigh their records, and you’ll pinpoint squads that are likely to overperform in the future. In this case, that leads you to that plucky squad known as … the Alabama Crimson Tide. It also highlights how the current format won’t allow some of the most dangerous teams into the field. More on that later.

2. The best underdogs play high-risk, high-reward styles

Inviting greater variance into the upset equation boosts the chances an underdog can clip a superior opponent. For longshots, inconsistency is a good thing. To examine this, we have looked at the weekly shifts in every FBS team’s basic power rating this season, after adjusting for the strength of their opponents. Our Variability Index measures which programs have the widest range of game-to-game outcomes. Kansas has the craziest gyrations among Power 4 teams, beating BYU and Colorado before getting wiped out by Baylor in the past three weeks.

3. Luck tends to regress to the mean over time

It pays to target underdogs that have been victims of bad fortune, and attack favorites who have received more than their fair share of good breaks. Teamrankings.com measures this by comparing team records with scoring margins. Their Luck Rankings call UCF the unluckiest team in the country: The Knights went just 4-8 but scored 42 points more than their opponents this season.

With all that in mind, let’s take a look at the teams currently sitting on the CFP bubble. We’ll define that as teams that have a realistic chance of playing a first-round road game. We’re not concerned with which teams are most deserving of a bid; we want to know which have the best chance of pulling off a major upset, whether that’s a 12-seed winning a road game against the 5-seed, or a 9-seed taking out the No. 1 team in the quarterfinals.

We will be able to take a deeper dive once we have matchups. And we can address teams that don’t offer particularly strong or weak upset chances – Miami and Clemson, specifically – should they find their way into the bracket. But for now, here are five potential underdogs that favorites should want to avoid and four they should hope to face.

Good dogs

Ole Miss and Alabama

Hey, don’t blame us for shoehorning a couple of the biggest powerhouses in the country into the role of plucky underdogs. The top conferences have expanded to the point where their highly ranked teams can’t all play one another. And the CFP selection committee still hasn’t made its mission clear: Is it out to reward the teams that accomplished the most, or the teams that would make the strongest contenders moving forward?

It’s nearly inevitable for some of the best Power 4 also-rans to end up underseeded. It was also entirely foreseeable, too. Back in May, Oklahoma AD Joe Castiglione asked, “[What] if a team has had a great season and played the toughest schedule in the season and has marquee wins but ends up with a 9-3 record?” So here we are.

No. 1 Oregon has gone 12-0 while scoring 422 points and allowed 194, for a 228-point differential. Alabama is at plus-219 (426-207) against a significantly tougher schedule. Ole Miss is at plus-283 (450-167) against a comparably difficult schedule.

GO DEEPER

How much does the CFP committee punish losing a conference title game?

Advertisement

The Crimson Tide have been unfortunate, winning six games by more than three touchdowns apiece while losing two by a touchdown or fewer. They rate 105th in the luck rankings.

The Rebels fare considerably worse than that, ranking 119th. They’ve walloped South Carolina and Georgia, but sustained three losses by a total of 13 points. Their pre-Thanksgiving game against Florida, where Ole Miss outgained the Gators 464 yards to 344 and had more red zone chances but lost, 24-17, was a particular horror show of uncharacteristic turnovers and fluky plays.

In fact, our Variability Index says Ole Miss has been one of the most consistent teams in the country, with the smallest spread among their very best and very worst performances. The Rebels’ weekly swings have been almost completely due to the strength of their opponents and (mostly bad) luck.

Whatever system you pick — Massey, ESPN’s SP+, the Simple Rating System, etc. — predictive analytics see both Alabama and Ole Miss as top-10 teams. It looks like Ole Miss is out, but if either of these two get in, all we can say is: Favorites, beware.

SMU

SMU enters Saturday night’s ACC Championship Game against Clemson as a 2.5-point favorite. If the Mustangs win, they could land the No. 3 seed.

If they lose, they can still make the Playoff, but probably as the 11- or 12-seed. It’s through that lens that we’re looking at them as a potential underdog. And in that scenario, their slingshot would be very dangerous. SMU ranks 30th in the nation in our Variability Index, the second-highest among all bubble teams — and not because their results have been bouncing all over the place, but because they’ve been improving by leaps and bounds.

SMU entered the national polls after beating Louisville on Oct. 5, and the Mustangs have continued to rise nearly every week since then, pulverizing ACC opponents by ever-increasing margins. In November, they won all four games by double-digits with an average margin of victory of 22.75 points.

SMU has been outstanding in the trenches, rushing for 177.9 yards while allowing just 95.8 per game. (Clemson, for comparison, is at 190.6 and 150.3.) And fairly quietly, quarterback Kevin Jennings has put together a season where he ranks 10th in the country in passer rating.

Point is, SMU is a top-10 team that’s unpredictable because it’s been getting better. If the Mustangs land in the lower reaches of the CFP, they’ll make one hot dog.

Indiana

By now, you know about Indiana’s astounding offense. The Hoosiers have hung 40 or more points on opponents eight times this season. QB Kurtis Rourke has a passer rating of 181.4, and WR Elijah Sarratt is nicknamed “Waffle House” because he’s always open, and he doesn’t even lead the team in yards per reception. That would be Omar Cooper Jr., who leads the country with 21.1 yards a catch.

But Indiana has also allowed just 14.7 points per game, the seventh-lowest total among FBS teams. The Hoosiers’ scoring margin (plus-344) is so huge that analytics systems see them as a top-10 team despite their middling schedule and a loss in their one true test at Ohio State.

Indiana has also had big swings: half a dozen games where they demolished opponents (beating Nebraska by 49, Michigan State by 37), and a handful of others where they won by merely comfortable margins (beating Northwestern by 17, Maryland by 14). Overall, the Hoosiers rank fifth in the country in our Variability Index, the highest among teams with a chance to make the CFP. So they’re very strong and very variable. Even on the road, that’s a recipe for seriously threatening a higher seed.

South Carolina

The Gamecocks aren’t likely to get in the field, but they represent another interesting case when viewed through the underdog lens.

In contrast to Alabama and Ole Miss, the Gamecocks haven’t been unlucky. If anything, their record is slightly better than their season-long numbers. But like SMU, they have been inconsistent because they have been getting better. South Carolina ranks 37th in the country in our Variability Index, second-highest among bubble teams, and you can see why: After three conference losses, including a blowout by Ole Miss, in a four-game stretch, this team turned around and whipped Texas A&M and Vanderbilt, and then overcame Missouri and Clemson by narrow margins.

All the while, South Carolina’s defense has been raising its game. Kyle Kennard now leads the nation with 11.5 sacks and 16 tackles for losses. Nick Emmanwori and Jalon Kilgore have eight interceptions between them. The Gamecocks have allowed just 15.3 points per game over the second half of this season. They’ve already shown that, at their best, they can play with anybody, and they’re showing at the end of the season their best is getting better. Alas, they probably won’t make the field.

Bad dogs

Boise State or UNLV

First things first: Boise State may not even get an opportunity to be a true underdog. Should the Broncos beat UNLV for the Mountain West title, they are likely to earn a first-round bye as the No. 4 or even No. 3 seed. That’s despite ranking just 21st in ESPN’s SP+ rankings. Then again, despite being the higher seed, they should be a significant underdog in the quarterfinals if the No. 5 seed (potentially Penn State or Notre Dame) advances. And the Broncos don’t have the profile of a team that should pose much of a threat.

As we’ve stressed, variability is a key ingredient in an upset. Low floors don’t matter: They are the difference between losing by three or 30. But high ceilings generate unlikely outcomes. Boise State is the antithesis of that type of team. From week to week, the Broncos’ opponent-adjusted power rating has changed (up or down) by an average of only 0.88 points, the smallest bounce in the entire country. They’re also 14th in the nation in “luck” meaning they have likely overperformed against an underwhelming schedule.

Yes, the Broncos took Oregon to the wire in September. And sure, they have Ashton Jeanty doing jaw-dropping things. But the metrics say Boise State’s highest level simply isn’t good enough, and that’s still better than UNLV’s!

The Rebels are ranked in the 30s by most rating systems (and 42nd by Massey), so it’s not like they’re some sort of sleeping giant. They rank in the top half of the country in terms of good luck, and are only in the middle of the pack in our Variability Index. UNLV simply isn’t a Playoff-worthy team and, if they shock Boise State in the Mountain West title game, the Rebels’ stay will be brief.

Iowa State or Arizona State

Whichever team wins the Big 12 Championship Game should savor the moment because it won’t last long. There’s a reason why the CFP committee has consistently ranked Boise State ahead of whichever team has led the Big 12 most of the season. The conference is really weak.

According to ESPN’s SP+ rankings, BYU is the best of the bunch (20th), but the Cougars won’t play for the conference championship. Iowa State ranks 24th – one spot ahead of 6-6 USC. And Arizona State is 39th!

It’s not just that these teams are mediocre (by playoff standards). They also don’t compensate with strong underdog traits. Both teams have been extremely fortunate: The Sun Devils rank ninth in the country in luck rating, and the Cyclones are 15th. Their level of play is also steady. Arizona State is in the top 40 of most consistent teams in the country, which is nice when you’re trying to beat the likes of Kansas and TCU, but not when you need a ceiling-game to beat Notre Dame. Iowa State is the third-most consistent team in the country.

Neither team did much in its nonconference games, unless you’re impressed by Iowa State’s one-point win over Iowa in September or Arizona State’s seven-point win over 2-10 Mississippi State.

In short, one of these teams will win the Big 12, likely play the No. 5 seed on the road … and lose.

(Illustration by Eamonn Dalton; photo of Kevin Jennings: Sam Hodde/Getty Images; photo of Ashton Jeanty: Brandon Vallance / ISI Photos / Getty Images; photo of Jalen Milroe: Jason Clark / Getty Images)