WWorldSportsXAIPredictions

2026 FIFA World Cup

Model predictions

A Dixon-Coles bivariate Poisson model fit on 49,215 senior internationals, then run through 10,000 Monte Carlo tournament simulations. Compared live against consensus-market implied probabilities.

Dixon-Coles bivariate Poisson + Monte CarloBrier skill: +7.0%Backtested on 256 matches48 teams covered

Top 10 favorites

Champion probabilities

From 10,000 Monte Carlo tournament runs.

Model vs. consensus markets

Top mispricings

Where the Dixon-Coles model disagrees most with publicly traded implied probabilities, ranked by |edge| × √(liquidity).

#TeamDirectionModel P(W)Market P(W)EdgeRecent form
1Argentina↑ UNDER26.1%9.0%+17.1pp14W 3D 3L · GD +1.65/g
2Colombia↑ UNDER11.7%1.6%+10.1pp8W 5D 7L · GD +0.6/g
3France↓ OVER3.1%16.1%-13.0pp13W 3D 4L · GD +1.05/g
4Spain↓ OVER8.4%16.0%-7.6pp14W 6D 0L · GD +1.7/g
5England↓ OVER3.2%11.1%-7.9pp15W 1D 4L · GD +1.75/g
6Ecuador↑ UNDER6.1%0.6%+5.5pp6W 13D 1L · GD +0.45/g
7Portugal↓ OVER2.2%7.4%-5.2pp11W 7D 2L · GD +1.3/g
8Germany↓ OVER0.9%5.3%-4.4pp13W 3D 4L · GD +1.4/g
9Uruguay↑ UNDER3.8%1.0%+2.8pp5W 11D 4L · GD +0/g
10Paraguay↑ UNDER2.3%0.5%+1.8pp8W 6D 6L · GD +0/g
11Brazil↑ UNDER11.4%8.5%+2.9pp9W 6D 5L · GD +0.65/g
12Netherlands↓ OVER1.2%3.3%-2.1pp10W 8D 2L · GD +1.5/g
13Senegal↑ UNDER2.1%0.6%+1.5pp15W 3D 2L · GD +1.75/g
14Morocco↑ UNDER3.9%1.7%+2.2pp16W 4D 0L · GD +1.45/g
15Australia↑ UNDER1.2%0.2%+1.0pp12W 4D 4L · GD +1/g

UNDER = model thinks team is undervalued relative to consensus markets. OVER = model thinks team is overvalued.

Stage-by-stage

Tournament progression probabilities

Probability each team reaches each stage of the bracket — based on simulation outcomes, not seeding alone.

TeamR32R16QFSFFinalChampion
Argentina98.9%86.3%65.7%48.2%35.3%26.1%
Colombia93.9%71.5%52.6%35.6%23.7%11.7%
Brazil99.1%79.0%48.2%33.3%18.3%11.4%
Spain97.4%72.1%49.8%31.3%18.6%8.4%
Ecuador97.2%61.7%37.1%19.8%10.2%6.1%
Japan94.4%56.6%31.3%17.1%7.9%4.5%
Morocco94.0%62.4%29.7%16.7%7.7%3.9%
Uruguay91.6%59.2%35.3%18.2%9.3%3.8%
England95.5%60.1%33.3%14.1%6.4%3.2%
France87.5%58.4%28.8%16.7%8.0%3.1%
Paraguay82.4%47.9%27.2%12.8%6.0%2.3%
Portugal78.6%46.7%28.4%14.4%6.5%2.2%
Senegal84.1%53.8%25.6%13.4%5.9%2.1%
Australia77.4%40.9%21.1%9.5%4.1%1.2%
Netherlands86.5%41.3%19.6%8.3%3.0%1.2%
Algeria79.0%43.9%22.6%8.7%2.8%1.2%
Norway77.3%43.0%17.1%7.9%3.1%1.0%
Canada96.2%48.8%19.8%6.9%2.2%0.9%
Germany93.3%42.2%18.1%7.0%2.4%0.9%
Switzerland95.0%46.3%18.1%6.4%2.1%0.7%
Croatia85.0%38.9%17.7%5.4%1.8%0.7%
Mexico93.5%41.4%15.8%6.9%2.8%0.6%
Iran78.1%37.9%16.5%6.1%2.3%0.5%
Belgium78.5%37.2%15.8%5.8%2.1%0.5%

Showing top 24 by champion probability of 48 total teams.

Underlying ratings

Top 30 teams by World Football ELO

The ELO rating is one input to the Dixon-Coles likelihood — but recent goal difference, opponent strength, and venue all feed in too.

#1Spain
2215.3
#2Argentina
2176.4
#3France
2129.3
#4England
2072.4
#5Brazil
2052.9
#6Colombia
2046.3
#7Portugal
2029.2
#8Netherlands
2012.9
#9Ecuador
2012.6
#10Germany
1983.4
#11Croatia
1981.8
#12Japan
1979.6
#13Morocco
1970.5
#14Uruguay
1967.2
#15Norway
1956.6
#16Mexico
1953.9
#17Switzerland
1947.8
#18Turkey
1947.4
#19Senegal
1924.6
#20Denmark
1921.3
#21Belgium
1914.9
#22Italy
1910.5
#23Paraguay
1906.6
#24Australia
1892.6
#25Canada
1886.5
#26Iran
1877.1
#27Austria
1872.5
#28South Korea
1866.2
#29Nigeria
1865.6
#30Panama
1854.6

Did the model actually work?

Calibration backtest — World Cups 2018 + 2022

Across 128 real WC matches, the dots cluster near the dashed diagonal — meaning when the model said 30%, teams actually won close to 30% of the time. The Brier-skill improvement of +7.0% over a uniform baseline tracks this.

Brier score (lower is better)
ModelWindowBrierLog-loss
DixonColes20180.59470.9991
Uniform(1/3)20180.66671.0986
DixonColes20220.64601.1049
Uniform(1/3)20220.66671.0986
DixonColesALL0.62031.0520
Uniform(1/3)ALL0.66671.0986

The math

Methodology

Plain-English description of what the model is doing under the hood.

1. Likelihood

For a match between home team i and away team j, we model the score (X, Y) as a bivariate Poisson with the Dixon-Coles low-score correction:

λ_h = exp(α_i + β_j + γ)         (home expected goals)
λ_a = exp(α_j + β_i)             (away expected goals)
P(X=x, Y=y) = τ(x,y,λ_h,λ_a,ρ) · Poisson(x;λ_h) · Poisson(y;λ_a)

α is each team's attack strength, β the defense strength, γ the home-field advantage (zeroed out for neutral-venue World Cup matches), and τ the Dixon-Coles τ correction that fixes the underestimation of 0-0 / 1-0 / 0-1 / 1-1 scorelines.

2. Time decay

Each match contributes a weight exp(-ξ · days_since_match) with ξ = 0.0019, which means a game's weight halves roughly every 365 days. A friendly from 2014 still informs the fit — just less than last week's UEFA Nations League match.

3. Fit

We minimize the weighted negative log-likelihood with L-BFGS-B, with the identifiability constraint that the first team's attack rating is fixed at 0 and the attack ratings are recentered to sum to zero after the fit. This gives clean, comparable attack/defense numbers across all 333 rated teams.

4. Tournament simulation

With fitted parameters, we run 10,000 Monte Carlo simulations of the full 48-team bracket: group stage → R32 → R16 → QF → SF → Final. For each match we sample a score from the bivariate Poisson, advance the winner, and aggregate stage-reach probabilities across all simulations.

5. Market comparison

For each team, we pull the implied win probability from public consensus prediction markets and compute edge = p_model − p_market. Mispricings are ranked by |edge| × √(liquidity) so large edges in thin markets don't dominate.

6. Calibration

Backtested on World Cups 2018 + 2022 — 128 matches. The model's Brier score (0.620) is +7.0% better than a uniform (1/3, 1/3, 1/3) baseline (0.667). Log-loss tells the same story.

Source code

The full Python implementation is open source: src/model/dixon_coles.py and src/model/simulator.py on our GitHub. Pull it, refit it on your own data, or argue with our priors.

Important: WorldSportsXAI publishes statistical model outputs and editorial analysis for informational and educational purposes only. Nothing on this page is financial, investment, or betting advice. Predictive models — including ours — are wrong about specific events frequently and on purpose; that's how probabilities work. Please bet responsibly, or not at all.