2026 NFL Season · One team a week · No repeats · Lose or tie, you're out

Survivor Roadmap

Mid-season plan for weeks 3–8: exact-optimized picks from live DraftKings spreads blended with ESPN FPI, assuming the pool ends within a few weeks. Teams already burned: Los Angeles Chargers, Philadelphia Eagles. 9 entries still alive.

This week — Week 3 · Sun Sep 27

San Francisco 49ers over Arizona Cardinals

Home, −8.5. Backup: Kansas City Chiefs (78%).

79%

win prob

The path from here · Weeks 3–8

SF San Francisco 49ers

vs ARI · home · Sun Sep 27 · spread 8.5

Backups: KC 78% free swap · BUF 72%

79%

public10%

alive79.1%

MIN Minnesota Vikings

vs MIA · home · Sun Oct 4 · spread 10.5

Backups: BAL 78% free swap · SEA 71%

79%

public31%

alive62.4%

NE New England Patriots

vs LV · home · Sun Oct 11 · spread 8.5

Backups: DET 71% free swap · CIN 69% free swap

72%

public23%

alive44.9%

LAR Los Angeles Rams

vs ARI · home · Sun Oct 18 · spread 13.5

Backups: NE 74% · SF 73%

83%

public48%

alive37.2%

HOU Houston Texans

vs NYG · home · Sun Oct 25 · spread 5.5

Backups: LAR 70% · SF 68% free swap

70%

public23%

alive26.0%

DAL Dallas Cowboys

vs ARI · home · Sun Nov 1 · spread 10.5

Backups: NE 71% · CIN 70%

77%

public41%

alive20.1%

Strategy vs the pool

Strategy	Survive wks 3–8	8-person pool	12-person pool
Optimal path (this plan)	20.5%	17.1%1.5× fair	12.0%1.6× fair
Ride the favorites	20.5%	17.1%1.5× fair	11.9%1.5× fair
Contrarian	13.4%	15.1%1.4× fair	11.1%1.4× fair

Pool cells show your expected share of the pot (fair share = 1/pool size). 50,000 simulated runs against opponents seeded with the pool's real already-burned teams, picking like the survivor public — clustering on the biggest available favorite.

Read before you pick

Pool state: 9 entries alive. Opponents' most-burned teams: SF by 5 of 8, LAC by 3 of 8, PHI by 3 of 8 — teams they can no longer use but you can.
Farm the bottom. This window picks against: Arizona Cardinals ×3, Miami Dolphins, Las Vegas Raiders.
Re-run before every deadline. Lines move on injuries and QB news. Log the pick in picks.json, run python3 update.py, and the remaining weeks re-optimize.
Survival compounds: the optimal path clears all 6 remaining weeks 20% of the time — the pool will almost certainly end sooner. It's won by outlasting, not by running the table.