After more than a century of college football crowning champions in various byzantine ways, the sport entered the modern era of sports by starting a four-team playoff in 2014. The somewhat strange results of the selection committee's first foray into picking four teams led me to believe that their methodology was frighteningly simple. Simple enough even to be replicated by an amateur modeler such as myself. So starting the next year, I did just that. The results have not been perfect, but they've been about close to perfect as I could hope for. Especially given that all my model accounts for is 1) number of losses, 2) rudimentary strength of schedule, and 3) a small adjustment that implicitly discounts 2-loss teams if the season has had minimal chaos. Unfortunately (or fortunately, depending on your desires for a larger playoff), this is the tenth and final year of the four-team playoff era, which means that my model is less than four months from reaching its expiration date. So let's enjoy one last quarter-trip around the sun together before I uninstall R from my computer for good (that's a joke I am never uninstalling it).
Very little here should surprise you. The clear top four teams all have the best odds to make the Playoff. The next dozen or so teams are the next dozen or so teams in the AP poll, and they all have varying levels of outside chances to pull off a special season. This is "boring" in a way as it's often fun when the computers and the poll voters disagree. But it's exciting in that this is the deepest set of good to great teams I have seen in some time, and everyone else seems to agree with this. Should be fun!
Conference Odds
This graphic is slimmed down a bit, as even more conferences have elected to simply pit their two best teams against each other. Salute to the ACC, CUSA, and Mountain West for making the right decision on this front. Overall, nothing sticks out to me too much other than a) SMU being the favorite in the American, and b) the Sun Belt East being an absolute grab bag of decent but not exceptional teams.
Weekly Previews
Here's the best part of this, where I tell you what to get excited about in each week of the season.
Week 1 - 0.910 playoff teams lost (ranked 12th of 13)
1. LSU (67% win probability) vs Florida State - 0.038 playoff teams lost
2. Clemson (86%) at Duke - 0.016
3. Florida at Utah (67%) - 0.008
A real wet fart of an opening weekend. To make matters slightly worst, none of these top three games are on Saturday. But the silver lining of this is that you can focus your Saturday Night on South Alabama-Tulane instead of having to watch Alabama pants someone in the Georgia Dome.
Week 2 - 0.159 playoff teams lost (ranked 10th)
1. Texas at Alabama (73%) - 0.090
2. Notre Dame (74%) at NC State - 0.015
3. Oregon (60%) at Texas Tech - 0.012
There's not necessarily a ton of Playoff implications here, but there is a ton of fun, weird matchups. Baylor-Utah! Wazzu-Sconnie! Oklahoma-SMU? Ole Miss at Tulane??! Kansas State-Troy!!!! Iga Świątek winning her second straight US Open? Coco Gauff winning her first? Let's get weird with it.
Week 3 - 0.819 playoff teams lost (ranked 13th)
1. LSU (78%) at Mississippi State - 0.014
2. South Carolina at Georgia (93%) - 0.012
3. Penn State (78%) at Illinois - 0.010
Look, I picked a good week to go to (tailgate at) an actual game. Washington goes to Michigan State, but it's on Peacock. Oklahoma goes to Tulsa? Whatever, man.
Week 4 - 0.278 playoff teams lost (ranked 4th)
1. Ohio State (75%) at Notre Dame - 0.095
2. Florida State at Clemson (68%) - 0.053
3. Ole Miss at Alabama (86%) - 0.027
Now here is the good stuff. Plus there's plenty of weird, fun depth, largely featuring the one-year-only iteration of the Big 12 (Oklahoma going to Nippert, UCF going to Manhattan, Texas making one last trip to Waco).
Week 5 - 0.158 playoff teams lost (ranked 11th)
1. LSU (65%) at Ole Miss - 0.025
2. Georgia (88%) at Auburn - 0.020
3. Alabama (88%) at Mississippi State - 0.020
Yep, the four "worst" weekends all happen in September. Oh well whatever it's football I'll take it. Clemson going to Syracuse, ND at Duke, South Alabama-JMU, and Utah making one last trip to Oregon State gives this one enough wacky depth to make it fun.
Week 6 - 0.214 playoff teams lost (ranked 6th)
1. Texas (62%) vs Oklahoma - 0.053
2. Alabama (78%) at Texas A&M - 0.041
3. Michigan (82%) at Minnesota - 0.022
Haven't mentioned Michigan yet because they do not play anybody until travelling to Minneapolis in week 6. Meanwhile the Tide will be on their third straight opponent with a decent pulse. Something to keep track of in the jockeying for the top four spots.
Week 7 - 0.173 playoff teams lost (ranked 9th)
1. USC at Notre Dame (52%) - 0.059
2. Oregon (51%) at Washington - 0.020
3. Arkansas at Alabama (91%) - 0.015
A weird week highlighted by a couple of coin flips, plus a few other coin flips lower down the ranks (KSU-TTU, UNC-Miami, OSU-UCLA) that could prove to be pivotal.
Week 8 - 0.268 playoff teams lost (ranked 5th)
1. Penn State at Ohio State (81%) - 0.079
2. Tennessee at Alabama (79%) - 0.050
3. Utah at USC (66%) - 0.038
A really fun week with some of the biggest games to date. Yeah, Bama should get revenge on the Vols, but also this game comes after four straight weeks of SEC West seasoning, which could take its toll on the Tide (one can only hope)? Also Texas goes to Houston in perhaps the funniest game of the weird Big 12...too bad Houston might be pretty listless by this point.
Week 9 - 0.185 playoff teams lost (ranked 8th)
1. Ohio State (84%) at Wisconsin - 0.047
2. Clemson (78%) at NC State - 0.026
3. Oregon at Utah (56%) - 0.025
Weird that there's a random Big Ten-Big 12 matchup in the middle of the season. Should be fun, regardless.
Week 10 - 0.303 playoff teams lost (ranked 2nd)
1. LSU at Alabama (72%) - 0.091
2. Notre Dame at Clemson (61%) - 0.074
3. Washington at USC (71%) - 0.029
A top-heavy week, but a) the top is very, very good, and b) there's some other fun stuff going, like the last edition of Bedlam.
Week 11 - 0.286 playoff teams lost (ranked 3rd)
1. Michigan (59%) at Penn State - 0.076
2. USC (55%) at Oregon - 0.044
3. Ole Miss at Georgia (86%) - 0.027
Two absolutely huge Big Ten games lead the way in a strong weekend. What a conference.
Week 12 - 0.209 playoff teams lost (ranked 7th)
1. Georgia (72%) at Tennessee - 0.063
2. North Carolina at Clemson (80%) - 0.030
3. Michigan (86%) at Maryland - 0.017
I don't like the UGA-Vols game taking place this late, but it does save this weekend from being a completely random grab bag of games. Washington at Oregon State and Texas at Iowa State helps ensure that some of the weirdest corners of college football will be relevant for a week.
Week 13 - 0.340 playoff teams lost (ranked 1st)
1. Ohio State (59%) at Michigan - 0.166
2. Clemson (77%) at South Carolina - 0.026
3. TCU at Oklahoma (67%) - 0.022
Dropping the sarcasm for a minute, it's too bad that Rivalry Week might be a bit of a dud this year. Yeah the big one makes it the #1 week in my rankings, but a lot of other matchups are less important at the moment, which is too bad. Given that we may not see The Civil War, The Apple Cup, and others for a while, I hope all the teams involved have great seasons so that the stakes of this weekend are at an all-time high. It would make both a fitting send-off to conferences as we know them and a strong argument for the primacy of regional rivalries to the allure of college athletics.
No comments:
Post a Comment