Kentucky is not an excellent team, and Tennessee was a heavy favorite against them entering last week. Still, in our model’s latest College Football Playoff probabilities, the Vols have climbed in a way normally reserved for teams who knocked an opponent out of contention. What happened?
There are two ways to improve your playoff chances, if you’re a CFP hopeful judging your status by where you stand in our model (if you are this specific entity, thanks for making us such a prominent figure in your assessment of yourself—we cherish the responsibility). One is to win a game there’s doubt you’ll win. The other is to make the projection likelier you’ll win future games. Tennessee, to a small extent, did the former on Saturday. To a much larger extent, they did the latter.
The question with Tennessee has been and will, for at least six more days, be this: How good are they? Yes, they beat Alabama. They beat Alabama by a field goal at home. But this was just one data point. Just one solitary data point. One identical to that Texas A&M boasted last year before Alabama won the SEC and A&M finished the year 8-4. It means something, but it can only mean so much.
How much it means, then, is a major subject of debate, and the amplitude of Tennessee’s performance against Kentucky—a 44-6 beatdown in which Tennessee averaged 9.8 yards every time Hendon Hooker threw the football, nearly a first down per pass attempt—has changed the nature of that debate. No, Kentucky isn’t excellent, but they’re not bad, and neither is LSU, and in addition to taking down Alabama, Tennessee’s now beaten each of those two by four scores or more.
Our model uses Movelor, our own rating system, as its input for single-game simulations. Movelor is a derivative of elo, a rating system which originated in chess but can be adapted to fit just about anything. The premise of elo, and therefore of Movelor, is to dole out large changes in rating after surprising results and small changes after unsurprising results. Movelor incorporates margin of victory, and as such, a 38-point result is surprising. Tennessee jumped this week. Tennessee jumped enough to pass 50% in playoff probability. Tennessee jumped enough to take over the fifth spot nationally in Movelor’s rankings. This now makes Movelor’s top five nearly identical to those of SP+ and FPI, two very different rating systems attempting to do a very similar thing. Ohio State, all three agree, is the best team in the country. Next come Georgia and Alabama, in some order. Then, it’s Michigan, and after Michigan, Tennessee. This is now agreed-upon by every rating system we personally reference. Tennessee is, to answer the question of how good Tennessee is, the fifth-best team in the country. The thing is…this can mean a variety of things.
FPI has Tennessee about a 9.5-point underdog this coming Saturday against Georgia. SP+ has Tennessee around a 9-point underdog. Movelor puts the spread at 13. Movelor, for better or worse, is slower to react than SP+ and FPI, and looks only at margin, not at any underlying factors leading into that final score. SP+ and FPI each care, in some way, that Hendon Hooker averaged 9.8 yards per pass attempt on Saturday. Movelor does not. Movelor only cares that the margin was 38. It also, again, is slower to react. It plays a longer game, not whipsawing as much as its comparables but sacrificing a little less than a point of predictive precision relative to those in exchange for its plodding pace. So, take all of this for what you will. Georgia, we all seem to think (Vegas spreads give Georgia an 8.5-point advantage at the moment), is better than Tennessee. But by how much?
The reason we care so much about how good Tennessee is doesn’t necessarily revolve around their chances of beating Georgia. This is the focus, yes, but because Tennessee is so likely to lose, we also care about the margin. Margin, as we say again and again, matters. Blowout losses are worse than close losses for rating systems, for betting markets, and—most importantly—for the College Football Playoff selection committee. If Tennessee finishes the season 11-1 with a close loss to Georgia, their playoff chance will be among the best in the country, should all go the way we expect it to go. This is why our model has them above 50% playoff-likely despite being such a large underdog in Movelor’s eyes. The possibility of Tennessee merely hanging with Georgia is large enough to make them a playoff contender. They don’t need to win the SEC to make the final four.
But, of course, winning the SEC would help, and if they don’t, they’ll need help. Let’s go through the rest of Saturday’s action, where help came and did not come and may soon come again:
The Other Contenders
Ohio State had trouble with Penn State. Ohio State had a lot of trouble with Penn State. The Nittany Lions held the Buckeyes in check on the ground and made gains through the air, making mistake after mistake but still taking the lead midway through the fourth quarter.
Then, the party started. Ohio State scored 28 points in just more than six minutes of gametime, Penn State scored only ten the rest of the game, and the fellas in red headed back to Columbus still undefeated. Maryland still awaits, as does—much more crucially—Michigan, but Ohio State’s almost guaranteed to reach 10-0, and Michigan, who pummeled Michigan State on the scoreboard before an ugly tunnel incident took the day, has a similar path: Two layups ahead of games against Illinois and Ohio State.
The Big Ten is going to put a team in the playoff. This is almost guaranteed. The only scenario I can find in which it wouldn’t happen is one where Michigan, Ohio State, and Illinois (who handled Nebraska in Lincoln) all somehow finish with two or more losses, something that would require Michigan to lose to Illinois, Ohio State to lose to Maryland, and whichever one doesn’t beat the other to lose to the Big Ten West champion, with Illinois also picking up an extra loss along the way.
The question with the Big Ten seems likely to be whether their 11-1 team is better than Tennessee’s, and therefore whether they can get a second bid. For the moment, we’re far enough away to not ask the question directly, but that’s looking like the question.
Alabama wasn’t in action, but they’re a question mark as well, and Mississippi escaping College Station with a win paves the way for a big two weeks around Tuscaloosa. On Saturday, Nick Saban’s Tide will go to Baton Rouge to face a potent LSU team. The next week, they’ll host Lane Kiffin’s Mississippi coming off a week of rest. Teams are similarly known and unknown to us at this point, but if this makes sense, we have more to learn about the Tide than we do about most others. They have more legitimate challengers left.
The specific question mark involving Alabama is related to the SEC West, but it’s also related to Georgia, as Tennessee’s is. If Alabama does win the SEC West (or if Mississippi shocks and wins out from here through the Egg Bowl), the division will send an 11-1 team to likely face a 12-0 Georgia or Tennessee, setting the stage for a whole lot of possibilities, most dramatically one in which Alabama and Georgia and Tennessee circle each other and we get three one-loss SEC teams. Georgia, for their part, quietly beat up Florida, taking a 28-3 lead into the half before letting the Gators have some fun the rest of the way.
Clemson, like Alabama, spent the weekend resting, and it’s looking like a great week to have done that, because Notre Dame might not be what they were thought to be but the Tigers are hardly favored by a field goal in betting markets as they head to South Bend (the spread in the rating systems we reference ranges from five points in SP+’s eyes to a one-point Irish advantage in Movelor’s). Notre Dame went to Syracuse, looked at a team that just nearly beat Clemson, and ran the ball up the middle until they got bored, then resumed when the game got close again, ultimately handing the Orange a 17-point loss and the ACC yet another black eye. Nobody should believe Notre Dame is a top ten team. But it requires some leaps of logic to call Clemson an overwhelming favorite in Indiana, which is why their playoff probability (27.4%) is so low despite their ACC Championship probability (67.4%) being so high. After Notre Dame, their schedule—provided no Tar Heel hijinks in the other division—goes Louisville (H), Miami (H), South Carolina (H), North Carolina (N). They could lose any of those games, but this is the toughest one left. If they go down, the ACC is likely done.
The Other Blowout
Movelor has been high on Oklahoma State all season, and after the Pokes took down Texas nine days ago, we were feeling good about that. The TCU loss stung, but Oklahoma State still looked believably to be the best team in the Big 12.
Not anymore.
One trip to Manhattan, zero points scored, and 48 points allowed later, the Cowboys’ playoff chance joins that of their hosts in the dustbin, each coming in at one tenth of one percent ahead of tomorrow night’s rankings. The Big 12’s hopes rest entirely on the shoulders of a TCU team ranked 21st, 14th, and 10th by Movelor, FPI, and SP+, respectively. TCU, meanwhile, nearly lost to West Virginia in Morgantown, needing to force a late three-and-out to put the game away, (West Virginia, for those not following the Big 12 closely, is in all likelihood going to miss a bowl.) and still has to go to both Austin and Waco. I don’t know who likes that for the Frogs.
The Other Conference
Out west, it was customary silliness. After Utah showed up without starting quarterback Cam Rising (I haven’t seen what the injury is, but it sounds like a week-to-week deal) and still snuck past Washington State on Thursday, all three playoff possibilities looked…like themselves on Saturday. Oregon scuffled early but pulled away from Cal, scoring 28 unanswered over the middle half of the game. USC put up points at Arizona and let Arizona put up points of its own, trailing to enter the fourth quarter but never seeing itself a live underdog. UCLA made quick work of Stanford.
Are these teams good? Are they better than Utah? The first question, for playoff selection purposes, doesn’t really matter. It’s the second one that counts. Oregon—after visiting Colorado and hosting Washington, one of which is a moderate challenge—has to host the Utes. USC, already with a head-to-head disadvantage, may have to stay ahead of Utah in the standings. UCLA has that head-to-head advantage, but like the other three holds a heavy risk of facing Utah again in the league’s championship game, a game in which many, ourselves and FPI included, think Utah would be favored.
So, although Utah isn’t a playoff contender themselves, they hold the keys in California (and, well, Oregon, but that isn’t alliterative). Oregon, of course, has that head-to-head loss to Georgia by…ahem…46 points (margin matters), but they retain the best playoff shot of the others, because they haven’t lost yet in conference play, they get Utah at home, and…come on, it was Georgia. It’s hard to envision a universe where they get in ahead of 11-1 Tennessee (unless Georgia hilariously beats Tennessee by 47), but the committee does like its power conference championships. They’re on the radar.
The Playoff Picture
Where does this all leave us? I’d describe it as follows:
- The SEC Champion will make the College Football Playoff.
- The Big Ten Champion will make the College Football Playoff.
- If Clemson goes 13-0, Clemson will make the College Football Playoff.
- If TCU goes 13-0, TCU will make the College Football Playoff, but this is much less likely than a 13-0 Clemson.
- The fourth spot (and third, if Clemson goes down) will come down to some combination of a one-loss Michigan/Ohio State loser, a one-loss Tennessee/Georgia loser, a one-loss Tennessee/Georgia winner who lost to Alabama, a one-loss Big 12 champion, and a one-loss Pac-12 champion. If Clemson has one loss, that loss will be too bad a loss (without good-enough wins). There are five lanes behind the three clear ones, then, and they will merge into one or two. We will likely lose at least one or two of them along the way, unrelated to the others. But this is where we’re heading.
Categories
Teams, their playoff probabilities, how we organize them in our minds:
Favorites (>80% Playoff Probability)
- Ohio State (92.5%)
- Georgia (82.9%)
Contenders (>10% Playoff Probability)
- Michigan (60.3%)
- Tennessee (57.2%)
- Alabama (53.0%)
- Clemson (27.4%)
Factors (>1% Playoff Probability)
- Oregon (9.4%)
- TCU (5.9%)
- USC (5.1%)
- UCLA (2.8%)
Cotton Bowl Candidates (<1% Playoff Probability, but in the Movelor Top 25)
- Utah (0.8%)
- Penn State (0.2%)
- Oklahoma State (0.1%)
- Minnesota (0.0%)
- Baylor (0.0%)
- Notre Dame (0.0%)
- Kansas State (0.1%)
- Texas (0.0%)
- Illinois (0.6%)
- Mississippi (0.8%)
- Oklahoma (0.0%)
- Wisconsin (0.0%)
- Arkansas (0.0%)
- LSU (0.1%)
- Wake Forest (0.0%)
So You’re Saying There’s a Chance (North Carolina)
- North Carolina (0.8%)
**
We intend to be back tomorrow morning to talk about the rankings tomorrow night. They’ll mean a lot. But they won’t mean everything.