WC 2026 · Forecasting Oxford Football Forecasting
🇹🇷 Türkiye UEFA Elo 1,911
38 / 31 / 31 win · draw · win most likely 1–1
🇵🇾 Paraguay CONMEBOL Elo 1,833 · world 23rd
Group
D
Date
Saturday 20 June 2026
Kick-off
04:00 UTC
Venue
San Francisco Bay Area Stadium, Santa Clara

Fig. V7 Ensemble · Group D

Türkiye v Paraguay — scoreline probabilities

0 Türkiye 0–0 Paraguay · 12.37% 12 Türkiye 0–1 Paraguay · 10.49% 10 Türkiye 0–2 Paraguay · 5.93% 5.9 Türkiye 0–3 Paraguay · 2.03% 2.0 Türkiye 0–4 Paraguay · 0.52% 0.5 Türkiye 0–5 Paraguay · 0.11% Türkiye 0–6 Paraguay · 0.02% Türkiye 0–7 Paraguay · 0.00% 31% 1 Türkiye 1–0 Paraguay · 11.99% 12 Türkiye 1–1 Paraguay · 14.44% (most likely) 14 Türkiye 1–2 Paraguay · 6.86% 6.9 Türkiye 1–3 Paraguay · 2.34% 2.3 Türkiye 1–4 Paraguay · 0.60% 0.6 Türkiye 1–5 Paraguay · 0.12% Türkiye 1–6 Paraguay · 0.02% Türkiye 1–7 Paraguay · 0.00% 36% 2 Türkiye 2–0 Paraguay · 7.55% 7.5 Türkiye 2–1 Paraguay · 7.74% 7.7 Türkiye 2–2 Paraguay · 3.96% 4.0 Türkiye 2–3 Paraguay · 1.35% 1.4 Türkiye 2–4 Paraguay · 0.35% Türkiye 2–5 Paraguay · 0.07% Türkiye 2–6 Paraguay · 0.01% Türkiye 2–7 Paraguay · 0.00% 21% 3 Türkiye 3–0 Paraguay · 2.91% 2.9 Türkiye 3–1 Paraguay · 2.98% 3.0 Türkiye 3–2 Paraguay · 1.53% 1.5 Türkiye 3–3 Paraguay · 0.52% 0.5 Türkiye 3–4 Paraguay · 0.13% Türkiye 3–5 Paraguay · 0.03% Türkiye 3–6 Paraguay · 0.01% Türkiye 3–7 Paraguay · 0.00% 8% 4 Türkiye 4–0 Paraguay · 0.84% 0.8 Türkiye 4–1 Paraguay · 0.86% 0.9 Türkiye 4–2 Paraguay · 0.44% Türkiye 4–3 Paraguay · 0.15% Türkiye 4–4 Paraguay · 0.04% Türkiye 4–5 Paraguay · 0.01% Türkiye 4–6 Paraguay · 0.00% Türkiye 4–7 Paraguay · 0.00% 2% 5 Türkiye 5–0 Paraguay · 0.19% Türkiye 5–1 Paraguay · 0.20% Türkiye 5–2 Paraguay · 0.10% Türkiye 5–3 Paraguay · 0.03% Türkiye 5–4 Paraguay · 0.01% Türkiye 5–5 Paraguay · 0.00% Türkiye 5–6 Paraguay · 0.00% Türkiye 5–7 Paraguay · 0.00% 1% 6 Türkiye 6–0 Paraguay · 0.04% Türkiye 6–1 Paraguay · 0.04% Türkiye 6–2 Paraguay · 0.02% Türkiye 6–3 Paraguay · 0.01% Türkiye 6–4 Paraguay · 0.00% Türkiye 6–5 Paraguay · 0.00% Türkiye 6–6 Paraguay · 0.00% Türkiye 6–7 Paraguay · 0.00% 0% 7 Türkiye 7–0 Paraguay · 0.01% Türkiye 7–1 Paraguay · 0.01% Türkiye 7–2 Paraguay · 0.00% Türkiye 7–3 Paraguay · 0.00% Türkiye 7–4 Paraguay · 0.00% Türkiye 7–5 Paraguay · 0.00% Türkiye 7–6 Paraguay · 0.00% Türkiye 7–7 Paraguay · 0.00% 0%

Cells show P(exact scoreline); the right column and bottom row are the marginal totals P(Türkiye scores k) and P(Paraguay scores k). Grid runs 0–7 goals per side; the 8–10-goal tail holds 0.00% of the mass and is omitted from the cells (not from the totals).

The grid makes Türkiye favourites at 37.7%, with a 31.3% draw. The single most-likely scoreline is 1–1 (14.4%), but no exact score clears 14% — the distribution is broad, as it should be.

Source · Oxford Football Forecasting model

Win · draw · loss

🇹🇷 Türkiye 37.7% Draw 31.3% 🇵🇾 Paraguay 31%

Rounded values sum to exactly 100%.

Expected goals (λ)

🇹🇷Türkiye 1.16
🇵🇾Paraguay 1.02

Poisson means feeding the grid; combined expected goals 2.18.

37.2% Over 2.5 goals P(3 or more goals in the match)
62.8% Under 2.5 goals complement of over-2.5
45.0% Both teams to score P(each side scores ≥ 1)
1–1 Most-likely scoreline modal exact score · 14.4%
Levi's Santa Clara, USA
Heat index 28°C apparent temperature (June–July)
Max temperature 27°C June–July daily high
Humidity 63% relative humidity
Altitude 1m above sea level

Source · Open-Meteo & venue records. Travel and time-zone exposure are per-team — see each side's dossier.

1,911 Elo rating 1,833
2.27 Recent NT form 1.67
€413M Squad value €152M
0.254 Squad form (global) 0.296
0.704 Fitness readiness 0.675
−0.41 Decoupling g −0.05

Türkiye carry the Elo edge (78 points). On the decoupling axis, Paraguay is the side whose squad is valued higher relative to its record.

How a single-match forecast is built

The pairing is scored by the ensemble — Dixon-Coles bivariate-Poisson, the Bayesian hierarchical model and the global LightGBM-Poisson, log-pooled — yielding the 11×11 scoreline grid above. Win/draw/loss, expected goals (λ), over-2.5 and both-teams-to-score are all marginals of that one grid, so they are mutually consistent by construction. The strength inputs shown here feed the models; the forecast is their pooled output, not a manual weighting of these rows. The model matches the market out-of-sample (RPS 0.1891 vs 0.1905); it does not significantly beat it at n = 3. The ensemble, in full →