Home win
75.29%
Union Saint-Gilloise
BEL1 · Belgian Pro League · Match Analysis
Union Saint-Gilloise vs
Oud-Heverlee Leuven
Model-based match probabilities derived from multi-season xG data, team strength, venue performance, recent form, Elo ratings and league context.
Model output
Match probabilities.
Probabilities are model estimates, not certainties or betting guarantees.
Draw
16.80%
X
Away win
7.91%
Oud-Heverlee Leuven
Score model
Most likely scorelines.
Exact-score probabilities from the Poisson + Dixon-Coles goal model.
Additional scorelines
3:17.65%
4:06.47%
4:14.87%
0:04.60%
2:23.39%
Team context
Strength and recent form.
Venue-adjusted team strength, Elo rating and recent xG performance.
Home
Union Saint-Gilloise
Elo1711.2
Overall attack1.82
Overall defense0.95
Home attack1.94
Home defense0.79
Form signal+1.773
Last 5 xGF / xGA2.87 / 0.96
Last 10 xGF / xGA2.53 / 1.02
Last 10 xGD+1.51
League statusSTABLE · from BEL1 2025/26
Away
Oud-Heverlee Leuven
Elo1463.6
Overall attack1.28
Overall defense1.64
Away attack1.13
Away defense1.93
Form signal-0.331
Last 5 xGF / xGA0.98 / 1.27
Last 10 xGF / xGA1.11 / 1.52
Last 10 xGD-0.41
League statusSTABLE · from BEL1 2025/26
Elo difference
+247.6
Elo is one input among several and does not determine the final 1X2 probability on its own.
Head-to-head
Historical matchup context.
H2H values provide context and are not used as a standalone prediction.
Goal distribution
Total goals profile.
The model's probability distribution for the total number of goals.
04.60%
110.32%
220.99%
322.09%
418.23%
512.04%
66.62%
7+5.10%
Technical detail
How the model reached the estimate.
Detailed diagnostics are kept below the headline probabilities so the public report remains readable.
League baseline and model setup
LeagueBEL1
Loaded teams24
Loaded historical matches1179
Target season2026/27
Previous seasons in history4
Season window2022/23 → 2023/24 → 2024/25 → 2025/26 → 2026/27
League tierTier 1
Historical league matches1179
xG matches333 / 1179
Effective xG matches262.77
Observed home xG1.61
Observed away xG1.32
Model home xG baseline1.61
Model away xG baseline1.32
Total league xG baseline2.93
Home-advantage weight1.000
Model configdata/models/BEL1.json
Team attack half-life240.0 days
Team defense half-life240.0 days
Attack strength weight1.000
Defense strength weight1.000
Venue prior matches0.00
DistributionPoisson + Dixon-Coles
Dixon-Coles rho-0.130 (effective -0.130)
Dixon-Coles decay0.000
Goal overdispersion0.000
Implied goal variance H / A2.55 / 0.75
Relative strengths and form diagnostics
Union Saint-Gilloise
Home attack strength1.141
Home defense weakness0.719
Base lambda home2.07
Attack form xGF level2.752
Defense form xGA level (lower is better)0.979
Opponent-adjusted attack residual+1.245
Opponent-adjusted defense residual (higher is better)+0.448
Opponent-adjustment weight0.000
Model attack signal+1.285
Model defense signal+0.488
Opponent blend0.000
Form matches 5 / 105 / 10
Last 5 xGF / xGA / xGD2.87 / 0.96 / +1.92
Last 10 xGF / xGA / xGD2.53 / 1.02 / +1.51
Oud-Heverlee Leuven
Away attack strength0.910
Away defense weakness1.124
Base lambda away0.86
Attack form xGF level1.025
Defense form xGA level (lower is better)1.355
Opponent-adjusted attack residual-0.472
Opponent-adjusted defense residual (higher is better)+0.083
Opponent-adjustment weight0.000
Model attack signal-0.442
Model defense signal+0.112
Opponent blend0.000
Form matches 5 / 105 / 10
Last 5 xGF / xGA / xGD0.98 / 1.27 / -0.29
Last 10 xGF / xGA / xGD1.11 / 1.52 / -0.41
Historical data coverage
Union Saint-Gilloise
Total matches161
Home / away matches80 / 81
xG matches47 / 161
xG home / away coverage23 / 80 · 24 / 81
Effective total24.59
Effective attack / defense24.59 / 24.59
Effective home matches11.95
Effective home attack / defense11.95 / 11.95
Oud-Heverlee Leuven
Total matches131
Home / away matches66 / 65
xG matches37 / 131
xG home / away coverage19 / 66 · 18 / 65
Effective total18.35
Effective attack / defense18.35 / 18.35
Effective away matches8.71
Effective away attack / defense8.71 / 8.71
Historical data coverage
50.90%
This measures usable historical evidence. It is not the probability that the prediction is correct.
Validation
Historical model performance.
Leakage-safe backtesting compares the fitted model with a league baseline over 1159 historical matches.
Positive improvement means the model performed better than the league baseline.
Total goals log-loss vs baseline-0.05%
1X2 log-loss vs baseline+2.96%
BTTS Brier vs baseline+0.23%
Over 2.5 Brier vs baseline-0.52%
Full backtest metrics
Model
MAE total goals1.3639
Exact score log-loss3.0338
Total goals Brier0.8324
Total goals log-loss1.9152
1X2 Brier0.6289
1X2 log-loss1.0443
BTTS Brier0.2499
Over 1.5 Brier0.1776
Over 2.5 Brier0.2517
Over 3.5 Brier0.2178
League baseline
MAE total goals1.3565
Exact score log-loss3.0747
Total goals Brier0.8318
Total goals log-loss1.9142
1X2 Brier0.6515
1X2 log-loss1.0761
BTTS Brier0.2505
Over 1.5 Brier0.1778
Over 2.5 Brier0.2504
Over 3.5 Brier0.2168
Calibration diagnostics
Home win · ECE 9.51%
0- 10% n=3 avg= 6.2% actual= 0.0% 10- 20% n=15 avg= 15.9% actual= 13.3% 20- 30% n=52 avg= 25.6% actual= 21.2% 30- 40% n=313 avg= 37.6% actual= 26.2% 40- 50% n=701 avg= 42.5% actual= 51.2% 50- 60% n=49 avg= 55.1% actual= 69.4% 60- 70% n=21 avg= 65.4% actual= 76.2% 70- 80% n=5 avg= 72.8% actual=100.0%
Draw · ECE 4.05%
10- 20% n=15 avg= 17.9% actual= 26.7% 20- 30% n=1103 avg= 28.4% actual= 24.6% 30- 40% n=41 avg= 31.0% actual= 39.0%
Away win · ECE 7.46%
0- 10% n=4 avg= 9.3% actual= 0.0% 10- 20% n=49 avg= 16.1% actual= 10.2% 20- 30% n=581 avg= 28.0% actual= 22.0% 30- 40% n=446 avg= 32.2% actual= 41.7% 40- 50% n=46 avg= 44.0% actual= 47.8% 50- 60% n=23 avg= 54.1% actual= 60.9% 60- 70% n=6 avg= 64.3% actual= 50.0% 70- 80% n=4 avg= 74.4% actual= 25.0%
BTTS · ECE 2.29%
30- 40% n=3 avg= 36.8% actual= 33.3% 40- 50% n=37 avg= 47.3% actual= 40.5% 50- 60% n=1025 avg= 55.2% actual= 54.0% 60- 70% n=92 avg= 63.0% actual= 52.2% 70- 80% n=2 avg= 70.8% actual= 0.0%
Over 2.5 · ECE 4.40%
30- 40% n=11 avg= 37.9% actual= 36.4% 40- 50% n=812 avg= 49.0% actual= 54.2% 50- 60% n=269 avg= 53.3% actual= 53.2% 60- 70% n=60 avg= 63.5% actual= 50.0% 70- 80% n=7 avg= 75.0% actual= 71.4%
Walk-forward - latest periods
| Period | N | Model 1X2 LL | Baseline 1X2 LL | Improvement |
|---|---|---|---|---|
| 2025-08 | 37 | 1.0250 | 1.0758 | +4.72% |
| 2025-09 | 27 | 1.1079 | 1.1017 | -0.56% |
| 2025-10 | 25 | 1.0204 | 1.0813 | +5.63% |
| 2025-11 | 31 | 1.0767 | 1.0870 | +0.95% |
| 2025-12 | 32 | 1.0577 | 1.1081 | +4.55% |
| 2026-01 | 20 | 1.0965 | 1.0998 | +0.30% |
| 2026-02 | 32 | 1.0101 | 1.0524 | +4.02% |
| 2026-03 | 28 | 1.0072 | 1.0901 | +7.60% |
| 2026-04 | 15 | 0.7823 | 1.0786 | +27.48% |
| 2026-05 | 15 | 0.9888 | 1.0436 | +5.25% |
| 2026-08 | 33 | 0.9379 | 1.0598 | +11.50% |
| 2026-09 | 30 | 0.8722 | 1.0279 | +15.14% |
Football is data.
NILYTIX turns match data into transparent model probabilities and measurable analytical signals.