Below are the new S&P+ rankings after college football’s Week 12. As you probably would have guessed, there are next to no changes at the top. It’s pretty clear which teams have been 2018’s best, and Week 12 did nothing to change that.

A reminder: S&P+ is intended to be predictive and forward looking.

Good predictive ratings are not résumé ratings, and they don’t give you bonus points for wins and losses. They simply compare expected output to actual output and adjust accordingly. That’s how a given team can win but plummet or lose and move up.

Through 12 weeks, the S&P+ rankings are performing well, hitting 54 percent against the spread and 52 percent on the over/under point totals for the year.

If you’re interested in a decent résumé ranking of sorts, I encourage you to visit this post on strength of schedule. I created a Resume S&P+ ranking and will be updating it on Mondays throughout the rest of the season.

Below, however, are the predictive ratings, the actual S&P+.

(You can find full unit rankings, plus a yearly archive, at Football Outsiders. The offense and defense pages are updated by Monday at the latest.)

This play matters more than that one

Photo by Thearon W. Henderson/Getty Images

Washington State’s 69-28 win over Arizona was one of the strangest games of the week from a statistical perspective. The Cougars destroyed a bad mediocre defense to the tune of 7.7 yards per play, and their Off. S&P+ rating rose from 37.5 adjusted points per game (15th) to 39.1 (14th) as a result.

They also let Arizona’s Khalil Tate find some more traction in the passing game than normal pre-garbage time, and the Cougs’ Def. S&P+ rating actually rose (in a bad way) from 24.3 adjusted points per game (40th) to 25.5 (43rd). So the offense improved by 1.6 points, and the defense regressed by 1.2.

The Cougars also fell from 16th to 20th in overall S&P+. How? Special teams. Specifically, one single kick.

Wazzu’s Special Teams S&P+ rating fell from plus-0.6 points (49th) to minus-0.3 (81st). Wazzu punter Oscar Draguicevich was shaky in his three attempts (34.3 net yards per kick), but Blake Mazza’s missed 27-yard field goal in the third quarter was particularly costly. Field goal efficiency is judged by expected points, and the expected value of a 27-yard field goal is 2.54 points since they’re made 85 percent of the time.

That single miss flipped Mazza’s average from plus-0.1 net points per kick to minus-0.7, and since place-kicking is worth the most value in the Special Teams S&P+ equation (because it’s the most predictive and stable, believe it or not), that meant a net drop in Wazzu’s S&P+ rating despite a resounding win. And since there’s almost no separation between the No. 10 and No. 22 teams (just 2.4 adjusted points per game, top to bottom), that means a fall in the ratings.

Predictive numbers are cranky and anti-social sometimes.

The week’s top movers (good)

Matt Bush-USA TODAY Sports

1. BYU (up 13 spots, from 66th to 53rd)
2. Florida State (up 13 spots, from 87th to 74th)
3. South Carolina (up 11 spots, from 49th to 38th)
4. NC State (up 10 spots, from 45th to 35th)
5. Houston (up nine spots, from 43rd to 34th)
6. FAU (up nine spots, from 61st to 52nd)
7. Mississippi State (up eight spots, from 19th to 11th)
8. Texas (up eight spots, from 41st to 33rd)
9. Ohio (up eight spots, from 57th to 49th)
10. Pitt (up eight spots, from 62nd to 54th)
11. WKU (up eight spots, from 115th to 107th)

Joe Moorhead’s Mississippi State Bulldogs absolutely destroyed Arkansas, 52-6, on Saturday. Is Arkansas very good? Not at all — the Hogs are 2-9 and 87th in S&P+. But nobody else, not even Alabama, did to them what MSU did.

Combined with the fact that the teams in that 10-22 range are so bunched together, this performance allowed the Bulldogs to leap eight spots. There’s always movement in the middle of the rankings, but an eight-spot jump in the top 20 is tough to pull off.

My goodness, UConn

James Guillory-USA TODAY Sports

UConn’s Def. S&P+ rating is currently 52.6 adjusted points per game. It probably goes without saying that that’s last in FBS. But that doesn’t really tell the whole tale: the Husky defense is 7.9 points worse than the second-worst defense.

Actually, you know what? Hold that thought. I’ll have to write about this later in the week.

Top movers (bad)

Jesse Johnson-USA TODAY Sports

1. UAB (down 15 spots, from 31st to 46th)
2. Minnesota (down 14 spots, from 53rd to 67th)
3. Boston College (down 14 spots, from 46th to 60th)
4. Colorado (down 12 spots, from 79th to 91st)
5. Arizona State (down 12 spots, from 52nd to 64th)
6. Utah State (down 12 spots, from 10th to 22nd)
7. Iowa State (down 11 spots, from 34th to 45th)
8. Liberty (down nine spots, from 114th to 123rd)
9. Syracuse (down nine spots, from 39th to 48th)
10. Tennessee (down eight spots, from 70th to 78th)

Minnesota has truly had one of the most bipolar seasons you’ll ever see. Over the course of 12 weeks, P.J. Fleck’s Gophers have now either risen or fallen by at least 10 spots eight times. They’ve risen by 37, 10, and 23 spots at different times, and they’ve fallen by 10, 17, 14, 17, and 14 spots. That’s not supposed to happen, and it doesn’t for almost anyone.

Just in the last five weeks, Minnesota has dropped 14 spots for getting destroyed by Nebraska, risen eight spots for beating Indiana, dropped 17 spots for getting blown out by Illinois, risen 23 spots for blowing out Purdue, and now dropped 14 spots for a disappointing performance against Northwestern. One would assume they’d finish their season with a loss at Wisconsin, but “assume” and “Minnesota” don’t really work in the same sentence this year.

FBS conferences, ranked by average S&P+ rating:

1. SEC (plus-10.9 adjusted points per game, up 0.9 points)
2. Big Ten (plus-5.4, down 0.2)
3. Big 12 (plus-5.1, down 0.3)
4. Pac-12 (plus-3.9, down 0.3)
5. ACC (plus-3.6, down 0.2)
6. AAC (minus-0.5, down 0.1)
7. Mountain West (minus-2.2, down 0.1)
8. Sun Belt (minus-4.3, up 0.1)
9. Conference USA (minus-5.9, same)
10. MAC (minus-6.3, up 0.4)

A unique week of movement here. The top and bottom conferences each improved by decent amounts, and almost everyone else regressed slightly.

---

Another reminder: I have made a few philosophical changes in this year’s S&P+ rankings.

When I get the chance (so, maybe in the offseason), I will update previous years of S&P+ rankings to reflect these formula changes, too.

1. I changed the garbage time definition. S&P+ stops counting the major stats once the game has entered garbage time. Previously, that was when a game ceased to be within 27 points in the first quarter, 24 in the second, 21 in the third, and 16 in the fourth. Now I have expanded it: garbage time adjustments don’t begin until a game is outside of 43 points in the first quarter, 37 in the second, 27 in the third, and 21 in the fourth. That change came because of a piece I wrote about game states at Football Study Hall.
2. Preseason projections will remain in the formulas all season. Fans hate this — it’s the biggest complaint I’ve heard regarding ESPN’s FPI formulas. Instinctively, I hate it, too. But here’s the thing: it makes projections more accurate. Our sample size for determining quality in a given season is tiny, and incorporating projection factors found in the preseason rankings decreases the overall error in projections. So I’m doing it.
3. To counteract this conservative change, I’m also making S&P+ more reactive to results, especially early in the season. If I’m admitting that S&P+ needs previous-year performances to make it better, I’m also going to admit that S&P+ doesn’t know everything it needs to early in a season, and it’s going to react a bit more to actual results.

Basically, I’ve added a step to the the rankings process: after the rankings are determined, I go back and project previous games based on those ratings, and I adjust the ratings based on how much the ratings fit (or don’t fit) those results.

The adjustment isn’t enormous, and it diminishes dramatically as the season unfolds.

Testing this process for past seasons improved performance against the spread a little and, more importantly, decreased absolute error (the difference between projections and reality) quite a bit. I wouldn’t have made the move if it didn’t appear to improve performance.