Pick Sheet — 2026 Week 1

Enter the confidence number next to each game. Highest number = most confident. Strategy: Always-favourite (baseline).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
ConfPickGameLine
16 LAC ARI @ LAC -10.5
15 JAX CLE @ JAX -7.5
14 DET NO @ DET -7.0
13 PHI WAS @ PHI -4.5
12 SEA NE @ SEA -3.5
11 CIN TB @ CIN -3.5
10 BAL BAL @ IND -3.5
9 LV MIA @ LV -3.5
8 PIT ATL @ PIT -3.0
7 TEN NYJ @ TEN -3.0
6 KC DEN @ KC -3.0
5 CHI CHI @ CAR -2.5
4 DAL DAL @ NYG -2.5
3 LA SF @ LA -1.5
2 BUF BUF @ HOU -1.5
1 GB FLIP CANDIDATE GB @ MIN -1.5

Generated 2026-08-12 05:23 UTC. Line shown is the pick's own spread.

Optional: your closest game to a tossup is GB @ MIN (1.5). Taking the dog — MIN — at confidence 1 is the pool-optimal deviation: it costs at most 1.5pp of win probability if your rivals pick independently, and gains +36.5pp if they all play the favourites too and the pool ties 30 ways. Another game has the same spread, so which one you pick here is arbitrary — that is fine, the value is in deviating at all, not in the game you choose.

This is a standings decision, not a football opinion. It is insurance against being right and tying, so it is worth most when you expect everyone else to play the chalk. It is the only deviation with evidence behind it — one game, at the bottom.

Why this sheet is boring on purpose. Across 2,119 games (2018–2025) the always-favourite baseline scored 72.39% of maximum points and went 66.4% straight up. Five alternatives were tested against it and none won: a devigged-moneyline ranking ties it at +0.00pp despite disagreeing about the order in 138 of 141 weeks, any re-weighting of spread magnitude is arithmetically the same sheet, and flipping small home dogs is worse by 0.52pp with a 99% interval excluding zero.

So the edge here is not the picks — it is not deviating from them. In a season-long pool the ordinary way to lose is a handful of gut overrides on close games, which is exactly where the market is hardest to beat and where the old “hybrid” strategy on this page used to override it.

The one exception is the flagged game above, and it is an exception about standings, not about football: winning a pool is not the same as maximising points, so a single deviation at the bottom of the sheet buys insurance against tying a field that all picked the same chalk. One game. Never more.