World Cup 2026 predictions: 100,000 simulations ✨
A two-layer model: a results layer (Elo, re-rated match by match through March 2026, so current form dominates) blended with a player layer — the market value of each nation's actual 2026 squad, which reprices on the players' form and injuries. We calibrate win/draw curves on the data itself, then simulate the entire official 2026 bracket 100,000 times. The output feeds the same AI engine that allocates your run.
🎲 World Cup 2026 probabilities, simulated
Each team's chance to reach the round of 32, the quarterfinals, the semifinals, the final, and to win the 2026 World Cup — measured as its share of 100,000 full-tournament simulations.
| Team | Squad value | Rating | Round of 32 | Quarterfinals | Semifinals | Final | 🏆 Champion |
|---|---|---|---|---|---|---|---|
 FranceI | $1.76bn | 1776 | 97.0% | 56.0% | 40.5% | 26.7% | 17.3% |
 SpainH | $1.46bn | 1762 | 98.1% | 50.0% | 35.9% | 23.0% | 14.3% |
 PortugalK | $1.18bn | 1738 | 96.1% | 51.1% | 33.1% | 20.5% | 11.8% |
 ArgentinaJ | $946m | 1717 | 95.7% | 43.3% | 27.1% | 16.0% | 8.7% |
 BrazilC | $1.05bn | 1703 | 95.6% | 41.9% | 25.5% | 14.1% | 7.5% |
 EnglandL | $1.51bn | 1698 | 94.7% | 42.1% | 25.1% | 14.0% | 7.4% |
 GermanyE | $1.15bn | 1668 | 93.5% | 34.6% | 20.2% | 10.3% | 4.9% |
 NetherlandsF | $967m | 1666 | 88.4% | 34.8% | 18.3% | 9.3% | 4.4% |
 BelgiumG | $627m | 1632 | 96.2% | 36.8% | 16.4% | 7.7% | 3.4% |
 ColombiaK | $353m | 1614 | 86.7% | 25.0% | 12.7% | 5.6% | 2.3% |
 CroatiaL | $446m | 1614 | 87.8% | 25.5% | 12.5% | 5.5% | 2.2% |
 MoroccoC | $564m | 1602 | 87.0% | 25.2% | 11.3% | 4.8% | 1.8% |
Model probabilities, not market prices — pick a team to see what the live market pays for the same run.
👑 Historical group favorites
Teams with a dominant lifetime record against their own 2026 group — the natural “Through the group stage” candidates.
Brazil3–0 vs Haiti · 2–1 vs Morocco · 8–0 vs Scotland
Build their Road to Glory →Spain3–0 vs Saudi Arabia · 5–0 vs Uruguay
Build their Road to Glory →
Egypt3–1 vs Belgium · 0–0 vs Iran · 2–0 vs New Zealand
Build their Road to Glory →
Mexico0–1 vs Czechia · 8–3 vs South Korea · 2–1 vs South Africa
Build their Road to Glory →England6–3 vs Croatia · 0–0 vs Ghana · 1–0 vs Panama
Build their Road to Glory →Netherlands2–0 vs Japan · 11–8 vs Sweden · 1–0 vs Tunisia
Build their Road to Glory →
USA2–1 vs Australia · 5–2 vs Paraguay · 2–2 vs Turkiye
Build their Road to Glory →
Turkiye2–0 vs Australia · 0–0 vs Paraguay · 2–2 vs USA
Build their Road to Glory →🔥 Rivalry edges in the group stage
Fixtures where one side owns the head-to-head record (3+ meetings, clear lead).
🆕 0 first-ever meetings
These nations have never played each other — no history, pure tournament drama.
📋 Played fixtures — what the model said
Kicked-off group games with the simulation's pre-match numbers, kept for the record.
⚙️ Model card — how every number on this page is computed
Ratings (Elo): every match in the 48 nations' complete head-to-head record (7,503 matches, through March 2026) replayed in date order. Elo re-rates after every match, so each result's influence is overwritten by what came after: today's ratings are driven by recent form, with 660 matches from 2020–2026 doing most of the work. K-factor 35 competitive / 20 friendly, margin-of-victory multiplier, +60 Elo home advantage at non-neutral venues.
Player layer (squad values): the market value of each nation's actual 2026 squad (Transfermarkt, as of June 2026, shown in US dollars at £1 = $1.336) — player-derived by construction, since every player's value is the transfer market's continuously repriced estimate of that player, and the roster is the real 26-man squad. Blended 50/50 with the results layer (z-scored Elo + z-scored log squad value). This is why squad-rich, results-mixed teams like England rate above their recent form.
Squad-turnover check: results ratings rebuilt from 2015-onward matches only (entirely current-generation squads) rank-correlate ρ = 0.95 with the full model, with the same top six in the same order — older data adds statistical convergence, not bias toward past generations. On top of that, live in-match form and injuries are priced by the market the picks route into.
Calibration: draw probability by Elo gap and goal-difference-per-gap are fitted on 6,003 matches from the same pass (after a 1,500-match burn-in) — measured from the data, not assumed.
Simulation: 100,000 runs of the official FIFA bracket (matches 73–104): all 72 group fixtures with host home advantage, standings with points→GD→GF tiebreakers, the eight best third-placed teams allocated to their constrained Round-of-32 slots, then the full knockout tree.
Rivalry signals (Edge Index): full head-to-head record, competitive-vs-friendly split, goal differential, most-recent meeting, and last-5-meetings form — across all 1,128 pairings.
Edge score: win-share differential weighted by sample size — a 5–2 record over 9 meetings outranks 2–0 over 2. Full weight requires 6+ meetings; fixtures with fewer than 3 meetings are never assigned an edge. Strength tiers: ≥0.22 (●), ≥0.32 (●●), ≥0.45 (●●●).
Group favorites: a nation's aggregate record against its own 2026 group opponents, requiring 4+ total meetings and a +2 win margin, ranked by margin over sample size.
From signal to stake: picking a team routes into the Road to Glory builder, where the AI allocation engine spreads your stake across tournament stages using live Polymarket prices, money-on-it, and return upside. History selects the candidate; the live market prices the run.
Guides & docs
Sources: historical results from Michill World Cup 2026 H2H Open Data (CC BY 4.0, 7,503 matches); squad market values from Transfermarkt (June 2026 squads, shown in US dollars); bracket structure from the official FIFA 2026 match schedule; live prices from Polymarket's order book. Not affiliated with FIFA, Transfermarkt, or Polymarket. Model output is context, not a promise of outcomes; predictions carry risk. ↗