Simulating the 2024 NCAA Football Playoff — Round 1
The bracket is seeded. Who are the objective favorites?
Opinion polls are just that: opinions, subjective value carrying little worth. Each week, this system identifies the Top FBS college football programs in terms of objective performance against strength of schedule (PASS), scoring tempo (sTem), and a true win percentage (tWin%). To learn more, check out the season kickoff, and last year’s primer. As the 2024 season progresses I’ll compare these objective ranks against the biased opinions of the College Football Playoff committee. And Follow me over at BlueSky!
The Playoff is finally here! The college football landscape feels like it really has changed drastically this year. Some lament the changes the transfer portal and overall modernization of the game of college football. Personally, I don’t see the drawbacks — but I’m also not a fan of a classic powerhouse on their way down a ladder. According to the objective, on-field data there’s been more parity in 2024 than in any of the previous years tracked and it’s made following the sport a delight. And here we are today, allowing more teams than ever before a chance at the most clandestine title in all of collegiate sports.
The PASS Top 25
Now that the Army/Navy game is complete, here are the season’s final performance ratings.
A week with just one game is interesting, in that it shows how just one game can create some shifts across the entire landscape. Sure, Notre Dame easily retains the top spot. However, the Irish played both Army and Navy this season and while their performance was very similar in points, the team overall allowed less yards and had more explosive sTem (tempo) against Army. With Army taking the loss, it drags ND’s overall grade down ever so slightly. Other teams see similar slight shifts based on how they performed in the season against those two teams, as well as against any common opponents shared between schedules.
It’s interesting to see what teams were able to maintain a high performance ranking despite taking on more losses. How is 8–4 Louisville ranked so high, when an ACC team like Syracuse has a better record at 9–3 and didn’t make the cut at all? Louisville does have a superior sTem pace of play and a slightly better oSoS (Objective Strength of Schedule), but the Cardinals’ point differential against FBS competition is the biggest reason why:
- Louisville put up 377 points while allowing 279, +98
- Syracuse put up 349 points while allowing 330, +19
There are arguments for and against running up the score in a blowout, and arguments for and against garbage time scoring for a team being blown out. In an effort to objectively balance those factors, PASS leans into scoring on a per play basis, the foundation of the sTem metric. Points per play (PPP) and points allowed per play (PAPP) pulls the curtain back even further:
- Louisville put up 0.55 PPP with 0.34 PAPP, +0.21
- Syracuse scored 0.42 PPP with 0.48 PAPP, -0.06
Syracuse allowed more points per play than they scored, easily the worst ratio of the thirteen 9-wins team this year. And plenty of teams with worse records showed a better edge in their performance against strength of schedule. One could interpret this year’s Orangemen as having really great luck, or having some great coaching and great players who overcame some terrible situations and odds. Thankfully, I’m just gathering the numbers accrued in-game and will leave the evaluating of intangibles up to the louder folks on TV.
The tWin% Top 25
If you prefer wins taking precedence over performance, True Win Percentage ranks all teams by their wins over FBS teams.
If you’ve been following along in previous weeks, there’s not much to see here. Oregon and Notre Dame remain at the top. Army plunges downward to 9th after one of the best service academy games in many years. That brings Navy back into the fold at 24th. Everyone else either stays put, or shifts one spot around those two schools.
The sTem Top 25
The sTem Top 25 grades teams’ offenses and defenses on a per-play basis, to see how efficiently they score and prevent scoring.
Not all of the playoff teams made the sTem Top 25. Clemson is down at #30. That doesn’t necessarily mean they’re outclassed — the team with the higher sTem ranking won 68.17% of total games this year. Anything can happen in games where two teams have marginal gaps in PASS and sTem, as evidenced by the conference championship games where the higher ranked sTem team went 2–7. That’s the biggest upset ratio of the metric to date, even from such a small sample size.
There’s that parity mentioned above. I love that it feels like any of the teams have a shot to win it all, whether or not they were actually one of the top 12 teams.
The Playoff Games — Round 1
We’re finally here — the dawn of the first day of the expanded playoff. Each game is simulated using a head to head matchup of their PASS / sTem data on both sides of the ball. Five hundred simulations of each game reveal an average score, points ceiling & floor, and projected yards, as well as tracking scoring likelihood. Vegas odds for comparison are as of 1:00AM EST on Dec. 20th. I stand excited to present the playoff games in the order that they will air.
Indiana @ Notre Dame
PASS: #6 Indiana 38.369 // #1 Notre Dame 59.832
sTem: #2 Indiana 0.409 // #1 Notre Dame 0.426
Avg. Score/Total: Indiana 29 Notre Dame 30 // Total 59
Ceiling/Floor: Indiana 62 | 0 // Notre Dame 63 | 0
Projected Yds: Indiana 359 // Notre Dame 321
Vegas says: Notre Dame (-7.5)
Simulation odds: Indiana 47.8% // Notre Dame 52.2%
It all kicks off tonight with the highest objectively ranked team in the nation going up against another top performer. Vegas is confident that the Irish loom large as the best team in the state of Indiana. The simulator is more cautious. Both of these teams have very strong defenses and offenses that have created points against almost every defense they’ve faced. sTem’s head to head projection of points actually has Indiana coming out on top with a score of 28–27.
Here’s a look at how each team performed in the simulator and what point totals they were most likely to produce:
A common question is: Why would you only run five-hundred simulations? The answer is the trendline (In this case a polynomial). You can run simulations into the ground, but after five-hundred simulations the trendline doesn’t tend to waver very much unless you’re including some pretty wild variables (The kind that don’t tend to aid a simulation aiming for real world results.) For posterity’s sake, here’s the same graph after two-thousand simulations:
One thing you’ll likely notice: Running gobs of simulations leads to bigger outliers. This x-axis has to move out a bit to account for one result out of the two-thousand where Notre Dame scored 68 points; obviously a result that’s highly unlikely. The trendline may move a hair here and there, but largely remains in place illustrating the same picture. The more simulations you run the more the peaks and valleys of occurrence will flatten towards the trendline, but the line itself doesn’t tend to change in any meaningful way.
So how do you interpret the above? If you focus on the most common simulation outcomes (Say, the top 10%), the Fighting Irish look most likely to score between 28 and 36 points. The Hoosiers are most likely to score between 26 and 35. Do keep in mind the low percentages on the y-axis and the sheer volume of potential outcomes when considering these numbers.
SMU @ Penn State
PASS: #8 SMU 37.371 // #12 Penn State 27.143
sTem: #8 SMU 0.257 // #10 Penn State 0.240
Avg. Score/Total: SMU 28 Penn State 29 // Total 57
Ceiling/Floor: SMU 57 | 0 // Penn State 50 | 2
Projected Yds: SMU 349 // Penn State 413
Vegas says: Penn State (-8.5)
Simulation odds: SMU 49.8% // Penn State 50.2%
This game is as close to a coin flip as you can get without actually being a perfect coin flip. The Nittany Lions won 251 of 500 simulations. Still, there are notes to be made that could swing the odds. Penn State is more consistently grouped in the 20–30 point range, while SMU is more chaotic and evenly spread across low to high, peaking around 16–36.
Clemson @ Texas
PASS: #23 Clemson 17.805 // #7 Texas 38.222
sTem: #30 Clemson 0.159 // #7 Texas 0.271
Avg. Score/Total: Clemson 22 Texas 29 // Total 51
Ceiling/Floor: Clemson 51 | 0 // Texas 56 | 4
Projected Yds: Clemson 323 // Texas 434
Vegas says: Texas (-12)
Simulation odds: Clemson 30.2% // Texas 69.8%
Tigers at Longhorns: Now here’s a game that’s not even close to a coin flip! Vegas seems to be on track with this one, too. The trendlines are further apart — clearly in Texas’ favor. The worst Texas did across all simulations was 4 points. If two safeties are your very bottom floor, you’re probably going to be liking your odds to win. Clemson looks most likely to score 15–22 points while Texas is comfortably ahead in the 25–34 range. Still, there are no sure things in the College Football Playoff, as Texas is just a touch under 70% for grabbing this victory.
Tennessee @ Ohio State
PASS: #25 Tennessee 16.726 // #3 Ohio State 52.177
sTem: #6 Tennessee 0.288 // #3 Ohio State 0.384
Avg. Score/Total: Tennessee 21 Ohio State 25 // Total 46
Ceiling/Floor: Tennessee 52 | 0 // Ohio State 58 | 0
Projected Yds: Tennessee 318 // Ohio State 357
Vegas says: Ohio State (-7.5)
Simulation odds: Tennessee 39% // Ohio State 61%
Finally, we’ve got the Buckeyes and Vols. On one hand, the Vegas spread looks maybe just a touch too big for the average simulation score. On the other, you could note the fact that Tennessee’s got the least-deserving performance against schedule rating of any playoff team. It’s lower than four other SEC schools who would’ve loved to be here in this spot (Alabama, South Carolina, Mississippi, and Texas A&M, in that order). Another fun thing to note is this game’s magnetism for overtime. 14 of 500 simulations went to OT, with Ohio State winning 10 of them.
So there you have it, four games and a whole lot of fun ahead. Remember that higher-ranked PASS teams succeed in just about 72% of games overall, so at least one of these games should product an upset! Good luck to your teams, your picks, and your holiday season kickoff. Hopefully these games will bring you as much cheer as I know they will for me.
For a subscription to full rankings of all 134 teams, a request to hear about your favorite team, or any other related questions, please reach out to me at objectivelydrew@gmail.com. Or drop me a comment right here. If you find this information of value, please consider sharing this or hitting the clap button a bunch. Your viewership is greatly appreciated. Thank you for being here!
