FATE settling… sims… locks… phase…

FATE

World Cup fantasy draft order
The Wrong Football live
Private league board: World Cup groups decide the draft.
Deepest Survivor decides the draft.Each group owns four teams. Draft slots sort by deepest round reached, then teams there, wins, and goals. i
projected draft order Deepest Survivor Projected final pick for every owner · deepest survivor → teams there → wins → goals
Deepest Survivor draft order
Projecting boardrunning simulations...
FATE DispatchLoading the first projection...Known results stay visible while the owner order initializes.
Draft racevisual timeline · movement

Draft race — expected final pick over time canonical · depth → teams → wins → goals

Expected final pick over time. 1.01 is the first pick; the bottom four are the bottom-four danger zone.

Detailed movement deltas refreshing. Last good board stays visible.
Skip probability river
simulation warming up…
Thicker band = more title chance Glow = your / selected owner Dim = not in focus Dashed = colorblind owner mark
Inspectorowner, risk, roots
Model i seed — predictions proof
Exact tiebreak receiptfrozen-now stack, only when needed

Frozen-now ledger depth → teams → wins → goals

Exact stack from results already locked, plus the remaining gates that can still move the projected board.

Future splittouch a match; compare both futures

Click a match, or choose one here, to split the future into both advancing outcomes and see owner deltas side by side.

Tiebreak detailsprojected final order · full stack on demand

Deepest Survivor standings

draft order law
Who goes furthest wins the better pickProjected rows are ordered by final simulated draft rank.
1
deepest team
2
teams there
3
total wins
4
total goals
not a rule
kickoff timingSchedule tells us when evidence arrives, not how picks score.

Sorted by projected final pick. The detail toggle shows the full tiebreak stack and Legacy What-If.

pick owner · group draft posture canonical stack Legacy What-If accuracy
pick 1.01–1.12 = canonical Furthest Team projection stack = depth → teams at depth → wins → goals Legacy What-If = old additive p10–mean–p90 tier canyon = expected-rank gap bottom-four danger zone locks after the July 19 final
Champion keyprojection · owner colours

Projected champion

Most-likely winner across — live simulations (the right bank of the river). Tap any team in the river to open its group.

Owners

Calibrationgroup + knockout Brier
Group standings & how FATE works ▾

Live group tables from real results, ordered by the real FIFA 2026 Art.13 tiebreak (head-to-head first). M% = the model's chance each team wins its group; ○% = the live group-winner market (Polymarket). Top two (green) advance; third (amber) chases a best-of-8 third-place spot.

How FATE works — the model, in full

In one sentence: FATE plays the whole World Cup nine thousand times — from each team's strength rating, honoring real results as they lock into history — and reports how often each outcome happens. Probabilities are frequencies, not opinions.

  1. Strength — one scalar per team. The toggle above chooses the input foundation: Team EV only, N. Silver PELE ratings when coverage is live, or Blend = Team EV plus PELE and World Elo as available. If an external rating feed is not ready, the toggle shows pending and the model falls back to Team EV.
  2. Match goals — opponent-aware Poisson, λ = 1.32·exp(±ha + Δs); the home edge ha applies only to a 2026 host (USA/MEX/CAN) at home — a World Cup is otherwise neutral, so we add no phantom edge from fixture order. Live match heat is display-only STC from score and clock; no xG or red-card feed is claimed.
  3. Scoreline — Dixon-Coles τ correction (ρ = −0.11) for realistic draw rates. A played match uses its real score every simulation.
  4. Group tables — FIFA 2026 Art. 13: head-to-head before overall goal difference, then a coin.
  5. Third place — the real best-8-of-12 by (points, GD, GF) fill the 32-team knockout.
  6. Knockout — official bracket gates where known; unresolved ties are simulated through regulation, extra time, and penalties.
  7. Monte-Carlo — n = 9,000 tournaments; P(reach a round) is just how often it happens. River mass halves 48→32→16→8→4→2→1.
  8. Sources — inputs: ESPN final scores plus the active strength foundation above. Checks: Polymarket, Kalshi & Manifold title odds; Polymarket group-winner odds; and Nate Silver's dated PELE champion-% snapshot.
  9. Independence (honest) — checks are correlated estimators of one truth, not independent votes. Agreement does not multiply confidence; the Divergence view is the useful signal. Market odds and PELE champion-% are never double-counted into the projection; PELE ratings / World Elo affect FATE only when the strength toggle says they are driving the simulation.
  10. Determinism — a seeded RNG makes every run replayable; 34/34 conformance checks lock the behavior.

It does NOT model: lineups, injuries, suspensions, rotation, xG history, fatigue, travel, weather, referees, in-game momentum, transfers, or motivation/dead rubbers. Disciplined uncertainty — explicitly not prophecy. Full written breakdown →