Live · updated 2026-09-14

Who Wins
The World Cup?

A stacked ensemble model rates every team, scores every match, then plays out the rest of the tournament 10,000 times a day — tested first on the 2018 & 2022 World Cups before being trusted with 2026.

ESP
Current favorite
Spain 61.4%
↓ SCROLL
104
Matches tracked live
0
Next matches predicted
10,000
Tournament simulations a day
2018 & 2022
Past World Cups tested
Live Tracker

Who Takes The Trophy?

Every remaining team's chance of winning it all, refreshed daily. Right now these numbers come from the global market; our own simulator's numbers join this board once the knockout bracket is locked in -- see the note below.

01
ESP
Spain
61.4%
02
ARG
Argentina
21.1%
03
ENG
England
14.0%
04
COL
Colombia
2.8%
05
SUI
Switzerland
0.8%
Find Your Team

Check Your Country

Every one of the 48 teams: still alive or eliminated, and how their odds have moved.

Algeria
Eliminated
Last tracked
0.1%
Day By Day

The Race, Charted

One point per team, per day. Watch the favorites rise and fall as real results land -- and see who the data backed once the champion is crowned.

Up Next

Our AI vs The World

For every upcoming match: our prediction, right next to what the rest of the world expects.

Just Finished

Latest Results

Every final score below feeds straight into tomorrow's ratings.

Spain1–0Argentina
France4–6England
England1–2Argentina
France0–2Spain
Argentina3–1Switzerland
Norway1–2England
Spain2–1Belgium
France2–0Morocco
United States1–4Belgium
Argentina3–2Egypt
Switzerland0–0Colombia
Mexico2–3England
Portugal0–1Spain
Brazil1–2Norway
Colombia1–0Ghana
Canada0–3Morocco
Spain1–0Argentina
France4–6England
England1–2Argentina
France0–2Spain
Argentina3–1Switzerland
Norway1–2England
Spain2–1Belgium
France2–0Morocco
United States1–4Belgium
Argentina3–2Egypt
Switzerland0–0Colombia
Mexico2–3England
Portugal0–1Spain
Brazil1–2Norway
Colombia1–0Ghana
Canada0–3Morocco
Verified Track Record

Don't Trust Us. Check.

Every prediction is committed to a public, SHA-timestamped git history BEFORE kickoff, hash-chained so it can't be quietly edited, and graded against the market once the match ends. The receipts are one click away.

The signature callJul 14 · we said Spain, the market said France
FRA
France
model 30% · market 40%
0–2
Final score
Spain
model 40% · market 30%
ESP
Our model said
Spain 40% — graded ✓ correct
The market said
France 40% — graded ✗ wrong
Proof it wasn't backdated
Public on GitHub since Jul 11, 07:54 UTC — commit 89a1898
13
Predictions in the ledger
85%
Our model's accuracy
77%
The market's accuracy
2 of 3
Disagreements we won
Every WC26 prediction · hash-chained · SHA-timestamped pre-kickoff13/13 with public git provenance
KickoffMatchOur callMarket callRecorded at
Jul 14France v Spain0–2Spain 40% ✓France 40% ✗89a1898
Jul 11Norway v England1–2England 51% ✓England 51% ✓8d40c02
Jul 10Spain v Belgium2–1Spain 60% ✓Spain 59% ✓5f487f0
Jul 9France v Morocco2–0France 50% ✓France 60% ✓028bc2e
Jul 7Switzerland v Colombia0–0Colombia 43% ✗Draw 63% ✓028bc2e
Jul 7Argentina v Egypt3–2Argentina 73% ✓Argentina 69% ✓3c70218
Jul 7United States v Belgium1–4Belgium 60% ✓United States 37% ✗3c70218
Jul 6Portugal v Spain0–1Spain 50% ✓Spain 50% ✓dd49a99
Jul 6Mexico v England2–3England 42% ✓England 39% ✓34a19b8
Jul 5Brazil v Norway1–2Brazil 55% ✗Brazil 53% ✗34a19b8
Jul 4Paraguay v France0–1France 70% ✓France 81% ✓34a19b8
Jul 4Canada v Morocco0–3Morocco 51% ✓Morocco 51% ✓34a19b8
Jul 4Colombia v Ghana1–0Colombia 79% ✓Colombia 83% ✓34a19b8
01

Every entry's provenance.github_url shows the prediction inside dashboard/data/upcoming_matches.json at a commit GitHub timestamped BEFORE kickoff -- open it, compare the probabilities.

02

The commit history of this ledger file and data/predictions/predictions_log.csv is an append-only daily trail written by the public GitHub Actions runs (Actions tab) -- rewriting it would break every fork, clone, and the Actions log.

03

Recompute the hash chain locally: python scripts/build_proof_ledger.py --verify (each entry_hash = sha256(prev_hash + canonical entry JSON); genesis = sha256('wc26-proof-ledger-v1')).

Ledger wc26-proof-ledger-v1· regenerated and committed by the daily pipeline · every entry's hash embeds the previous entry's hash, so a quiet edit anywhere breaks every hash after it · raw ledger · its commit history
Backtested

Tested On Real History

Before trusting it with 2026, we replayed the 2018 and 2022 World Cups from the group stage onward. The eventual champion's predicted chances climbed round after round in both -- and beat a simple world-ranking guess on average.

2018Champion: France
After Groups
6%
After R16
13%
After QF
31%
After SF
62%
Our AI World-ranking guess
2022Champion: Argentina
After Groups
22%
After R16
21%
After QF
50%
After SF
65%
Our AI World-ranking guess
Under the Hood

How It Works

01

Rate Every Team

Every team carries a strength score that rises and falls with each real result. Recent form counts for more than ancient history.

02

Predict Every Match

Three different predictors — a machine-learning model, a ratings model, and a world-ranking rule — each give a view, blended into one probability per match.

03

Play It Out 10,000 Times

The remaining tournament is simulated ten thousand times a day. Count how often each nation lifts the trophy — that's their chance of winning it all.

Behind The Scenes

What Powers This Site

Every tool used to build what you're looking at -- from the models to the pixels.

Data & Machine Learning
Pythonpandasscikit-learnXGBoostElo ratingsMonte Carlo simulation
Live Data Feeds
The Odds API — live scores & market oddsOpen historical datasets — internationals since 1872, FIFA rankings
This Website
Next.jsTypeScriptTailwind CSSFramer MotionRecharts
Engineering Practice
Git & GitHubpytestDecision log (DECISIONS.md)Backtesting before going live
MLOps Infrastructure
AirflowDockerMLflowSupabaseFastAPI + OpenTelemetry — live serving APIEvidently — daily drift reportsKubernetes — local kind cluster