MBModelBall

England vs Ghana

World Cup · Tuesday, June 23, 2026

8:00 PM UTC

Kickoff in 3 days

⚡ THE MODELS DISAGREE16pp spread

grok overweights market odds

Relevant dimension: Odds integration

MODEL CONSENSUS

4
pick England
0
pick Draw
0
pick Ghana

All Six Predictions

The Edge ★

Fingerprint-weighted ensemble

Favored outcome
England 71%
England71%
Draw18%
Ghana11%

Model Weights

Grok 3: 39%GPT-5.4: 26%Claude Sonnet 4.6: 18%Gemini 3.1 Pro: 18%

Naive Average

Equal-weight baseline

England70%
Draw19%
Ghana12%

INDIVIDUAL MODELS

GPT-5.4

The Market Baseline

Fingerprint →
England68%
Draw19%
Ghana13%

Claude Sonnet 4.6

The Analyst

Fingerprint →
England68%
Draw18%
Ghana14%

Gemini 3.1 Pro

The Generalist

Fingerprint →
England63%
Draw22%
Ghana15%

Grok 3Most divergent

The Contrarian

Fingerprint →
England79%
Draw15%
Ghana6%

Why They Differ

Most Divergent Model
Grok 3
Explanation
grok overweights market odds
Relevant fingerprint dimension
Odds integration

Measures ability to update predictions toward market consensus when odds diverge from priors.

The Edge ensemble uses this fingerprint information to adjust model weights. After the match, we'll validate whether the fingerprint correctly predicted which model would be most accurate.

Predicted scoreline

2-0
2-0
12.8%
1-0
10.4%
3-0
9.7%
2-1
9.4%
1-1
9.1%
Each model's scoreline pick
Claude Sonnet 4.6 2-0Gemini 3.1 Pro 2-0GPT-5.4 2-0Grok 3 2-0

Goal expectancy 3.00 per match from betting-market totals; split fitted to the calibrated probabilities shown above with a Dixon–Coles low-score correction. England listed first.

Ensemble Logic

Odds Divergence Context

Market gives underdog >35% probability

Weights:

Grok 39% · GPT 26% · Claude 18% · Gemini 18%

Active: PC05 (Overconfidence Index)

Want Deeper Analysis?

Get predictions before each match with full divergence analysis, model fingerprint insights, and real-time updates.