Defense rankings — 2026 Week 9

Use this for the order, not the number. Ranking defenses this way is about twice as good as ranking them by season-to-date fantasy points (Spearman 0.29 out of sample vs 0.13) — but half of what a defense scores is turnovers and touchdowns, which nothing predicts.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Defense ProjFloor–ceiling Pts allowedfrom tiers SacksTurnovers
1 JAX @ BAL 9.1 2.9 – 16.3 16.1 +2.2 3.0 1.5
2 SEA vs ARI 9.0 2.8 – 16.2 15.8 +2.3 3.0 1.5
3 HOU @ LAC 8.5 2.2 – 15.6 17.4 +1.8 2.9 1.5
4 NYG @ PHI 8.1 1.8 – 15.2 18.1 +1.7 2.8 1.4
5 KC vs NYJ 7.8 1.6 – 15.0 19.1 +1.4 2.7 1.5
6 SF vs LV 7.3 1.0 – 14.4 20.8 +0.9 2.6 1.5
7 NO vs CLE 7.3 1.0 – 14.4 19.8 +1.2 2.6 1.3
8 BUF @ MIN 7.2 1.0 – 14.4 20.5 +1.0 2.6 1.4
9 DEN @ CAR 7.1 0.8 – 14.2 20.9 +0.9 2.6 1.4
10 DET @ MIA 7.1 0.8 – 14.2 21.5 +0.7 2.6 1.5
11 CHI vs TB 7.0 0.7 – 14.2 21.0 +0.9 2.6 1.4
12 LA @ WAS 6.9 0.7 – 14.1 21.6 +0.7 2.6 1.4
13 NE vs GB 6.8 0.5 – 13.9 21.1 +0.8 2.5 1.3
14 PHI vs NYG 6.4 0.1 – 13.6 21.3 +0.8 2.5 1.2
15 DAL @ IND 6.3 0.0 – 13.4 23.5 +0.2 2.4 1.4
16 CIN @ ATL 5.9 -0.4 – 13.1 24.5 -0.0 2.3 1.4
17 CLE @ NO 5.9 -0.4 – 13.1 23.4 +0.2 2.4 1.2
18 GB @ NE 5.7 -0.6 – 12.9 24.1 +0.1 2.3 1.3
19 MIN vs BUF 5.6 -0.7 – 12.8 23.7 +0.2 2.3 1.2
20 CAR vs DEN 5.5 -0.8 – 12.6 24.2 +0.1 2.2 1.2
21 LAC vs HOU 5.4 -0.9 – 12.6 23.9 +0.1 2.2 1.1
22 ARI @ SEA 5.3 -1.0 – 12.4 24.5 -0.0 2.2 1.1
23 BAL vs JAX 4.9 -1.4 – 12.0 25.2 -0.2 2.1 1.1
24 WAS vs LA 4.7 -1.6 – 11.8 26.4 -0.5 2.0 1.1
25 TB @ CHI 4.6 -1.7 – 11.8 27.1 -0.6 2.1 1.2
26 ATL vs CIN 4.2 -2.1 – 11.3 28.3 -0.9 1.9 1.2
27 IND vs DAL 4.1 -2.2 – 11.2 28.4 -0.9 1.9 1.1
28 MIA vs DET 4.0 -2.3 – 11.1 28.5 -0.9 1.9 1.1
29 NYJ @ KC 3.9 -2.4 – 11.0 28.9 -1.0 1.9 1.1
30 LV @ SF 3.2 -3.1 – 10.3 31.1 -1.5 1.7 1.1
Projected points allowed comes from the posted total and the opponent's implied total; the tier column is the expectation across the points-allowed tiers, not the tier a single projection happens to land in. Defensive touchdowns and safeties are the league mean — nothing forecasts them (fitted 2019-2025). Measured walk-forward, one season held out at a time: Spearman 0.295, MAE 4.27 against a 4.41 league-mean baseline, R² 0.079 over 3,136 defense-games. The R² is the honest ceiling: this orders defenses, it does not predict their scores. Study at backtests/kicker_dst/FINDINGS.md.