Forecast @ 14 July 07:22 ET (Quarter-finals)

12m 56s runtime · $2.21

France leads the semifinal picture at 40.8%, with England and Spain close behind and Argentina fourth. Our international-results ratings and the betting market disagree structurally on France (market much keener) and Spain (market cooler than our ratings), so both readings are blended. Confirmed fitness for Mbappe, Tchouameni, Messi and Rice changes nothing beyond what the ratings already assume.

% win WC

France0.0%
England0.0%
Spain0.0%
Argentina0.0%
Czechia0.0%
Korea Republic0.0%
Mexico0.0%
South Africa0.0%

Where we differ from the market

Our forecast 21.3%Market 21.7%-0.5pp below
Our results-based ratings actually favour Spain more than the market does; with Messi's Argentina and Mbappe's France also in the mix, we blend both views, leaning toward the market's more cautious read on Spain.
Our forecast 40.8%Market 39.1%+1.6pp above
The betting market rates France well above our results-based ratings, seeing Mbappe and Tchouameni fit and rating France's ceiling higher than a pure results model does. We blend both readings, weighted mostly toward the market view, since the gap is structural rather than noise.
Our forecast 16.2%Market 17.3%-1.1pp below
Both our ratings and the market see Argentina as the outsider of the four remaining former champions, with Messi fit to start but not enough to close the gap on the other three.
Our forecast 21.8%Market 21.9%-0.1pp below
Both our ratings and the market read England similarly, so there is little disagreement to resolve here. Declan Rice being fit to start is already priced into the rating and changes nothing new.

Different plausible worlds

Each world is a distinct posterior over team strengths; the forecast is their weight-averaged mixture.

30%Using our results-based ratings
70%Using the betting market's odds

Distribution of winning the World Cup by team

France are the clear favourites at 40.8% to win it all.

40.8% is the weighted average of France’s chance of winning the World Cup across 200 simulated draws.

The betting market rates France well above our results-based ratings, seeing Mbappe and Tchouameni fit and rating France's ceiling higher than a pure results model does. We blend both readings, weighted mostly toward the market view, since the gap is structural rather than noise.

Using the betting market's oddsThe betting market's odds for the four semifinalists with the bookmaker margin stripped out, converted into ratings-equivalent strength shifts.

70%

Using our results-based ratingsThe baseline rating fitted from international results and path difficulty, with no market input, unweighted by fitness news since led-0003 through led-0006 confirm no change to the baseline.

30%

Sources considered by the agent

Internal forecast workings

1Step 1
Read agent-20260713-100102's published mixture (: this scenario 0.30 / this scenario 0.70) and compared its two component worlds against today's baseline sim and today's de-vigged market. this scenario is essentially unchanged (all four teams within ~0.2pp, well under the 0.16pp noise floor), so…
- Prior (2026-07-13): this scenario weight 0.30 (pure fitted strength, no perturbations), this scenario weight 0.70 (four inverted StrengthPerturbations: france +0.2255, spain -0.2194, argentina -0.1095, england +0.0756), published blend France 40.7% / England 21.5% / Spain 20.9% / Argentina 16…
- Continuity read on this scenario: today's baseline (spain 34.9, france 24.1, argentina 23.2, england 17.8) matches the prior this scenario world (34.83/24.33/22.99/17.85) to within 0.2pp for all four teams, below the 0.16pp paired-seed noise floor. Verdict: CARRY, no drift.
- Continuity read on this scenario: today's de-vigged market (france 38.9, spain 22.2, argentina 17.1, england 21.9) has moved substantially from the this scenario world priced into (france 48.62, spain 14.47, argentina 13.69, england 23.22): france -9.7pp, spain +7.7pp, argentina +3.4pp, engl…
- The July 13 published narrative for described England's position using language implying the semifinal was decided ('all other teams zero... eliminated' framing applied loosely to the ranked table) while the France-England SF is in fact unplayed (scheduled July 15 per the run's own journ…
- Noise floor for these comparisons: 0.16pp (paired-seed, 20k sims).
2Step 2
Registered : a two-base this scenario (0.30) / this scenario (0.70) mixture over the four semifinalists. France and Spain gaps are structural (market outside the model's own p10-p90 parameter band); Argentina's gap is borderline (market sits almost exactly at the model's p10); England's gap is…
- registered: this scenario (weight 0.30, no perturbation) + this scenario (weight 0.70, inverted strength deltas france +0.2183, spain -0.2026, argentina -0.1147, england +0.0747).
- Mixture title shares: france 40.5%, spain 21.4%, argentina 16.5%, england 21.6%.
- wq.impact deltas (own-team read, floor 0.16pp): france +15.09pp, spain -12.98pp (self -12.975pp), argentina -6.02pp, england +4.41pp -- all clear the noise floor as raw perturbation effects, but their evidentiary weight differs by structural test.
- Structural test via wq.title_uncertainty p10/p90: france market 0.391 is outside model band [0.186,0.265] (structural); spain market 0.217 is outside model band [0.294,0.413] (structural); argentina market 0.171 sits almost exactly at model p10=0.175 (borderline/weak); england market 0.219 sits insi…
- team_path_tree: France's SF-advance probability is only 43.3% against #1-ranked Spain, the model's own structural reason to doubt a market premium for France beyond the model rating; England's SF-advance vs Argentina is 46.5% in baseline, consistent with a competitive, unresolved match already price…
3Step 3
Repaired 's structural validator defect and re-registered as , reusing the identical two-base worlds and weights (this scenario 0.30 / this scenario 0.70 over france/spain/argentina/england) with unchanged perturbation deltas. Recomputed the missing factor_audit rows, including a fr…
- Reused 's exact worlds/weights: this scenario (0.30, no perturbation) and this scenario (0.70, strength deltas france +0.2183, spain -0.2026, argentina -0.1147, england +0.0747) unchanged.
- Recomputed wq.mixture_spread on these same worlds: france mean 0.408 (p10 0.220, p90 0.528, width 30.7pp, vs_floor 3.10x); spain mean 0.215 (p10 0.128, p90 0.370, width 24.2pp, vs_floor 2.69x); argentina mean 0.161 (p10 0.105, p90 0.243, width 13.8pp, vs_floor 1.30x); england mean 0.216 (p10 0.160…
- Fixed the spain world_direction_contradiction: the this scenario scenario copy for spain now explicitly states the market prices Spain BELOW the model, matching its negative delta sign (-0.2026); all four team copy strings were re-audited for sign consistency (france +, spain -, argentina -, england +…
- Registered a complete factor_audit with checks for bases, previous_continuity, market_gap, ledger_pricing, result_attribution, mixture_spread and a dedicated world_direction_consistency row documenting the repair.
- New artifact reproduces 's title shares exactly: france 40.5%, england 21.6%, spain 21.4%, argentina 16.5%, noise floor 0.452pp -- no re-simulation, no reweighting, no new worlds.