There’s a narrative that will be pushed that Clemson/Georgia is a College Football Playoff elimination game, and it’s wrong. There’s a counternarrative that will be pushed that Clemson/Georgia is a free play exhibition with no downside for either team, and it’s also wrong.
We still don’t have our model finished (Validate! That! Dataset!), but the last two years of following the playoff rankings closely plus a lot of hours spent studying the rankings in the system’s earlier five iterations have taught us a lot, so it’s with that knowledge that we embark on Week One, in which there is, as there is most weeks in college football, a good deal of playoff impact.
The Big Game: Georgia vs. Clemson
No galaxy brains allowed. Georgia/Clemson is huge. Kirby Smart’s had the talent. He’s knocked on the door. He came as close as one can come to winning a national championship. But his teams have been plagued with inconsistency, and it’s been rare that they’ve looked overpowering.
Some of this is the nature of how Georgia plays football under Smart. Some of it’s that they’ve never had all that good of a quarterback under Smart. Some of it is maybe a mild case of programmatic dysfunction. But whatever the reason, Georgia has had the ingredients and never the cake.
Meanwhile, Clemson’s trying something new. The Tigers have played big September games before under Dabo Swinney, but Auburn and Texas A&M were not, at those times, what Georgia is today. Those were potential New Year’s Six teams. This is a potential playoff team. And this isn’t exactly a home-and-home like those were. It’s a neutral-site game, and while it’s on the South Carolina border, it’s hard to imagine Georgia fans being outmanned anywhere save Pickens County. If the SEC is like Europe in the World War era, with powerful enemies all around, Clemson has been comfortably isolated across the Atlantic…until now. Maybe. It’s also possible Georgia stinks, or that Clemson’s good enough to make it look like Georgia stinks.
Clemson’s done a thing over Swinney’s tenure in which they’ve peaked at the right time. The Tigers play their best football in December and January. In September, they’ve been solid, but they haven’t been the fear-inspirers they are come the holidays. The game’s up for grabs. Clemson’s the favorite, but the game’s up for grabs. And that’s bad news for Clemson.
ESPN’s FPI has its oddities, but it’s easy to access and it’s not entirely out of line with the consensus on these two teams, so let’s take a look, not at what it says about the game, but at what it says about the rest of the way for both the winner and the loser.
FPI’s high on Clemson’s chances. It gives the Tigers a 69% chance of leaving Saturday night a winner. It also gives Clemson a 32.7% chance of going undefeated this season, which means the probability of Clemson going 13-0 rises to 47.4% in simulations in which it defeats the Dawgs. That’s a 50/50 shot at sailing through a schedule where Clemson might not play another ranked team until the ACC Championship.
Georgia does not have it so easy. By virtue of playing in the Southeastern Conference, their undefeated probability, per FPI, is only 2.2%, and rises to only 7.1% in the scenarios in which they defeat the Tigers. September isn’t bad on the guys from Athens after this weekend, and their SEC West opponents—Arkansas at home, Auburn on the road—are a merciful pair. Still, they’ve got a gauntlet ahead of them, with the toughest task, by nature, coming on December 4th in Atlanta, should they play well enough to make it.
The biggest thing placing this game in between those two narratives presented above is that the College Football Playoff selection committee often compares teams head-to-head. The data point acquired Saturday night might not end up changing the playoff field. But it could. Because when teams are being compared head-to-head, the best league’s runner up and other leagues’ champions are sure to share some time in the thunderdome.
Should Georgia defeat Clemson, it should be able to afford an SEC Championship loss, and I say that even if Georgia loses a regular season game somewhere else (provided it’s an excusable loss—Florida in Jacksonville if Florida loses twice elsewhere, or maybe that trip to Auburn). A 12-1 or 11-2 Bulldog team that beat Clemson and survived the SEC has a good chance of coming out ahead of a 12-1 Clemson team that walked through the ACC and pounded UNC or Miami in Charlotte at year’s end. This might not matter—there might be room for both, we might not get to this point at all—but it might. If Georgia wins, it has one of the best cards any contender will have in its hand, and better still, Clemson will be in a boat in which they’ll need a lot of help should they stumble somewhere else along the way.
Should Clemson defeat Georgia, Clemson still won’t be guaranteed a spot. They’ll need to handle the regular season schedule, and they may even still need to win the ACC Championship, depending how good or bad their opponent is and depending how scathed or unscathed other Power Five champions are. But again…they’ll be about 50% likely to win out, and with enough margin for error that they’ll be a presumptive playoff team until something changes. Georgia, meanwhile, will be in the same position they’d be in with a win, except one important step down—they could get in at 11-2, by going undefeated in league play, playing well enough in a loss to the SEC West champion, and getting help elsewhere, but that help couldn’t come from Clemson. There are only four playoff spots available. If Georgia loses the SEC, there are only three spots available. If Georgia loses to Clemson and then loses the SEC, that’s down to two. In a target that small, that’s a big reduction.
This is possibly a game where margin doesn’t matter much. It will shape things to an extent, but mostly as a starting point for the playoff rankings, and while the starting point’s important, it’s easily changed. A blowout and a close loss are different in a lot of scenarios, but while our model won’t have this ironed out, my guess is that this won’t really be one of them. In the event that Saturday night is a blowout and the loser ends up within a sniff of the playoff picture, the narrative will likely dictate that they’ve improved. And it’s possible they will have improved. But it’s also possible they won’t have improved.
In short, this is a big game. I’m going back-of-the-envelope here, but I’d guess Clemson’s playoff probability is 55% with a loss (they’d likely need to win out) and 85% with a win (disaster could still strike, especially within the conference and especially if Georgia loses a few more times), while I’d guess Georgia’s is around 10% with a loss (needing to win out or get very lucky) and 40% with a win (needing to win eleven more games somewhere or get very lucky). Hopefully we’ll have a model for you by the beginning of next week so we can check those. Either way, we’ll be able to check them eventually (Remind me to check these I’m feeling ambitious about the back of my envelope).
The Rest of the Schedule
Tonight’s Ohio State/Minnesota matchup is intriguing, partially, to me at least, because there’s a big gap between the betting markets and Bill Connelly’s SP+ projections. My impression is that SP+ struggles on the margins, like many ratings systems do, so it may be underestimating the Buckeyes, but it’s also possible the world is overestimating the significance of last season, in which the Goph’s only beat Illinois, Purdue, and Nebraska while losing to Michigan, Maryland, Iowa, and Wisconsin. P.J. Fleck might not be a great coach. But he might be, and last year was perhaps not all that indicative of a program’s true quality. I believe Connelly has the Gophers 24th in the country. Pull off the upset, they’ll be in the playoff conversation, and deservedly so. Back on the Ohio State side: It’s one of many games of this kind. Not likely do-or-die, but if things break the wrong way, it could turn out to have been do-or-die. You don’t always know at the time. Careful eating that fried twinkie.
Staying in the Big Ten, SP+ is also high on Penn State and Wisconsin, who conveniently open the season against one another at Camp Randall. Wisconsin could be favored in every game they play until the Big Ten Championship, so while the Badgers aren’t new to playoff hopefulness, this year’s among the more encouraging years for that particular faith tradition. The Nittany Lions do have to play in Columbus this year, and they host Auburn and Michigan and visit Iowa City, so while they might make last year look like a fluke, they might do that by going 10-2 even if they can get past Bucky.
Indiana/Iowa will get some love, but with Iowa having to play in Ames, Madison, and Evanston while hosting Penn State and Minnesota, it’s hard to get too high on the Hawkeyes. I’d guess this’ll be a come-to-Earth moment for the Hoosiers, but the vibes are something over there, and maybe those matter. Certainly a big game, but likely not playoff-shaping.
Moving back to the SEC/ACC side of things, I guess Alabama/Miami has a similar bent to Ohio State/Minnesota. No, it probably won’t shake things up. But if it does, it’ll shake them up in an enormous way.
North Carolina and Notre Dame each have big expectations as they open play on the road against respective washed-up powers Virginia Tech and Florida State. Those expectations are probably more legitimate for UNC, but Notre Dame has the better path by way of not presumably having to play Clemson, and by way of getting UNC at home at the end of October. Neither is all that likely of a playoff contender if they win, but if either falls on their face out of the gate, it’ll be one fewer team to worry about.
In Pac-12 action, UCLA tries to build the conference a decent floor, hosting LSU at the Rose Bowl. I’m not sure this is getting the attention it deserves, but maybe Chip Kelly’s performance at UCLA so far is a better indicator of Chip Kelly’s performance at UCLA in the present and future than I’m giving it credit for being. Perhaps I’m also underestimating LSU.
Texas gets a dangerous one to open the Steve Sarkisian era, hosting Louisiana-Lafayette in Austin. The Ragin’ Cajuns memorably tore Iowa State apart to open last year, and while there’s reason to believe there is no way in hell that the committee lets a Group of Five team in the playoff, we have, to be fair, never seen a Group of Five team beat an otherwise undefeated Power Five champion. Will Texas be an otherwise undefeated Power Five champion? It’s highly unlikely. It’s more likely that they open the year 0-2. But given that the only eyebrow-raisers they’d have to pull off would be winning in Fayetteville, beating Oklahoma at the State Fair, and winning in Ames—all of which are eyebrow-raisers but none of which are jaw-droppers—we’ll at least mention it, if for no other reason than to try to give ULL some deserved credit.
Elsewhere in the Group of Five, Boise State visits UCF tonight in a game that would be a lot more meaningful if it were at the end of the year as part of a Group of Five playoff that those leagues completely have the power to create in an effort to boost their best teams’ strengths of schedule but have not yet, to date, tried to create. But hey, why exhaust every option to make the playoff when…I’ve got nothing. Exhaust every option, guys.
***
Alright, that’s conveniently exactly ten games to follow. Manageable. Round number. Good work by the fates. If any other games this weekend affect the playoff we either won’t realize it or it will be such a big effect that it’ll mean we had some very wrong priors, for reasons legitimate or illegitimate. Also, you want evidence a four-team playoff is more exciting? I skimmed the FCS slate and even if one were to make the FCS the focal point of their life, it would be a rather empty weekend in terms of playoff impact. 24 teams is a lot. So might be twelve.
