← all fixtures
France
vs
England
Third-place Play-off · 7/18/2026, 9:00:00 PM
1-1
alt 0-0 · xG 1.04–1.02
4conf
actual 4-6 ✗ missed
Score-probability matrix
0
1
2
3
4
5
6
0
13
12
7
1
13
14
7
2
7
7
3
4
5
6
↓ France goalsEngland goals →
Win / draw / loss
W 35%D 31%L 34%
Market
51% / 24% / 25%

France favoured (35% to win); the model's most likely scoreline is 1-1.

Confidence 4/10. Even the favoured outcome (35%) fails about 65% of the time; a single match is high variance. This one is close; treat the scoreline as indicative only.

How we got to 1-1
  1. 1
    Model fit
    Team strengths fit from 15,899 matches (Dixon-Coles, recency-weighted, low-score correction ρ=-0.044). Neutral venue — no home advantage.
  2. 2
    Team strengths (log scale, vs average)
    France: attack +1.50, defence +1.52
    England: attack +1.43, defence +1.62
    Higher attack = scores more; higher defence = concedes fewer.
  3. 3
    Expected goals
    France: exp(base +0.15 + att +1.50 − opp.def +1.62) = 1.02
    England: exp(base +0.15 + att +1.43 − opp.def +1.52) = 1.06
    “base” is the league scoring level the model fitted.
  4. 4
    Agent adjustments
    λ 1.02/1.06 → 1.04/1.02
    • Tactical home defence ×1.03 — Both sides are in near-identical form (LWWWW) with full availability, so baselines need little adjustment. The one differentiator the model may underweight is defensive solidity: France concede far fewer per game (0.6 vs 1.2), suggesting a tighter structure, so I nudge France's defence up slightly and England's marginally down; attacks stay neutral given equal scoring output and no tactical or stakes edge at a neutral venue.
    • Tactical away defence ×0.98 — Both sides are in near-identical form (LWWWW) with full availability, so baselines need little adjustment. The one differentiator the model may underweight is defensive solidity: France concede far fewer per game (0.6 vs 1.2), suggesting a tighter structure, so I nudge France's defence up slightly and England's marginally down; attacks stay neutral given equal scoring output and no tactical or stakes edge at a neutral venue.
  5. 5
    Scoreline
    A Poisson model on those expected goals (with ρ) gives the score-probability matrix above; its most likely cell is 1-1 at 14%. Win / draw / loss come from summing the matrix.
  6. 6
    Confidence
    4/10 — a sharper peak and a clearer favourite raise it; large agent moves lower it (model agreement 95%).
France vs England — predicted 1-1 · Lambda XI