ATROPOSPlan the whole season, not 3–4 weeks
The standard advice — "plan a few weeks ahead" — is measurably wrong in big-format contests. Same weekly options, two strategies, 200,000 simulated seasons: greedy best-team-available survived 0.009% of the time; a full-season planned route survived 0.025%. Nearly 3x, before ownership and splitting are even considered.
The wall a 4-week plan can't see
The 2026 World Championship demands ten picks in weeks 12–16. On current lines, week 13 offers exactly two favorites strong enough to trust for its two mandatory picks. Whether those teams are still in your inventory was settled in September and October — a horizon of three or four weeks arrives at the gauntlet with whatever happens to be left.
Planning ≠ predicting
A season plan is not a bet that December lines hold. Measured drift: ~1 point one week out, ~10 at five weeks, ±13–15 at three months. So the plan discounts distant favorites accordingly and gets re-solved every week as lines move. The route is a living object; what it preserves is optionality at the bottlenecks, not a frozen prediction.
And the discount nobody applies
Hoarding for a week the contest never reaches is pure waste: 12.7% of simulated seasons are over before week 15 plays. Correct future value = win probability x drift discount x P(you're alive) x P(the contest is, too). The field skips the last two factors, which is one reason its hoard-then-dump pattern — ~95% of holders firing a standout spot simultaneously — keeps repeating.
FAQ
How far ahead should you plan survivor pool picks?
The whole season, before week 1. Common advice says three to four weeks; in Atropos's simulations of the 2026 World Championship, a greedy strategy taking the best available team each week survived 0.009% of the time, while a globally planned full-season route survived 0.025% — nearly triple — from the identical weekly options. The gap comes entirely from late-season weeks whose supply is determined by choices made in September.
Why isn't planning 3-4 weeks ahead enough in survivor pools?
Because the binding constraint sits farther out than four weeks. In the 2026 World Championship, week 13 demands two picks and offers exactly two trustworthy favorites; whether you can use them was decided by picks made two months earlier. A four-week horizon walks into that wall with the same empty inventory as no plan at all.
Doesn't planning far ahead fail because lines change?
Plans must be re-solved weekly, not abandoned. Measured line drift is ~1 point one week out, ~10 points five weeks out, and ±13–15 points at three months. The right structure is a full-season route re-optimized every week as lines move — planning and updating are not opposites. Distant weeks are planned with shrunk probabilities that respect the drift.
Should you plan around your survivor pool ending early?
In big contests, yes. 53.6% of simulated 2026 World Championship seasons end with the entire field eliminated, and the contest is fully over before week 15 plays in 12.7% of them. Teams saved for weeks that never happen are wasted inventory — future value must be discounted by the chance nobody is left to use it.
How full-season routes and books are built → · See the whole season on the grid