Home win
58.61%
Lille
FRA1 · Ligue 1 · Match Analysis
Lille vs
Le Havre
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
25.35%
X
Away win
16.04%
Le Havre
Score model
Most likely scorelines.
Exact-score probabilities from the Poisson + Dixon-Coles goal model.
Additional scorelines
3:06.86%
3:15.95%
1:24.64%
0:14.48%
2:24.24%
Team context
Strength and recent form.
Venue-adjusted team strength, Elo rating and recent xG performance.
Home
Lille
Elo1587.4
Overall attack1.56
Overall defense1.18
Home attack1.66
Home defense0.94
Form signal-0.121
Last 5 xGF / xGA1.48 / 1.69
Last 10 xGF / xGA1.41 / 1.36
Last 10 xGD+0.05
League statusSTABLE · from FRA1 2025/26
Away
Le Havre
Elo1442.3
Overall attack1.20
Overall defense1.63
Away attack1.13
Away defense1.76
Form signal-0.086
Last 5 xGF / xGA1.56 / 1.63
Last 10 xGF / xGA1.45 / 1.55
Last 10 xGD-0.11
League statusSTABLE · from FRA1 2025/26
Elo difference
+145.1
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.
08.18%
115.46%
225.95%
322.02%
414.82%
57.97%
63.58%
7+2.02%
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
LeagueFRA1
Loaded teams25
Loaded historical matches1343
Target season2026/27
Previous seasons in history4
Season window2022/23 → 2023/24 → 2024/25 → 2025/26 → 2026/27
League tierTier 1
Historical league matches1343
xG matches1342 / 1343
Effective xG matches660.97
Observed home xG1.61
Observed away xG1.31
Model home xG baseline1.61
Model away xG baseline1.31
Total league xG baseline2.92
Home-advantage weight1.000
Model configdata/models/FRA1.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 / A1.82 / 0.87
Relative strengths and form diagnostics
Lille
Home attack strength1.023
Home defense weakness0.796
Base lambda home1.76
Attack form xGF level1.453
Defense form xGA level (lower is better)1.574
Opponent-adjusted attack residual-0.023
Opponent-adjusted defense residual (higher is better)-0.132
Opponent-adjustment weight0.000
Model attack signal-0.007
Model defense signal-0.114
Opponent blend0.000
Form matches 5 / 105 / 10
Last 5 xGF / xGA / xGD1.48 / 1.69 / -0.21
Last 10 xGF / xGA / xGD1.41 / 1.36 / +0.05
Le Havre
Away attack strength0.898
Away defense weakness1.067
Base lambda away0.94
Attack form xGF level1.520
Defense form xGA level (lower is better)1.606
Opponent-adjusted attack residual+0.066
Opponent-adjusted defense residual (higher is better)-0.138
Opponent-adjustment weight0.000
Model attack signal+0.060
Model defense signal-0.146
Opponent blend0.000
Form matches 5 / 105 / 10
Last 5 xGF / xGA / xGD1.56 / 1.63 / -0.08
Last 10 xGF / xGA / xGD1.45 / 1.55 / -0.11
Historical data coverage
Lille
Total matches145
Home / away matches72 / 73
xG matches145 / 145
xG home / away coverage72 / 72 · 73 / 73
Effective total28.21
Effective attack / defense28.21 / 28.21
Effective home matches13.69
Effective home attack / defense13.69 / 13.69
Le Havre
Total matches107
Home / away matches54 / 53
xG matches107 / 107
xG home / away coverage54 / 54 · 53 / 53
Effective total27.48
Effective attack / defense27.48 / 27.48
Effective away matches13.42
Effective away attack / defense13.42 / 13.42
Historical data coverage
83.48%
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 1323 historical matches.
Positive improvement means the model performed better than the league baseline.
Total goals log-loss vs baseline-0.09%
1X2 log-loss vs baseline+6.57%
BTTS Brier vs baseline-0.05%
Over 2.5 Brier vs baseline+0.19%
Full backtest metrics
Model
MAE total goals1.3544
Exact score log-loss2.9759
Total goals Brier0.8304
Total goals log-loss1.9162
1X2 Brier0.6006
1X2 log-loss1.0038
BTTS Brier0.2494
Over 1.5 Brier0.1781
Over 2.5 Brier0.2474
Over 3.5 Brier0.2150
League baseline
MAE total goals1.3377
Exact score log-loss3.0577
Total goals Brier0.8307
Total goals log-loss1.9145
1X2 Brier0.6503
1X2 log-loss1.0744
BTTS Brier0.2493
Over 1.5 Brier0.1774
Over 2.5 Brier0.2479
Over 3.5 Brier0.2147
Calibration diagnostics
Home win · ECE 2.06%
0- 10% n=6 avg= 7.4% actual= 0.0% 10- 20% n=61 avg= 16.3% actual= 9.8% 20- 30% n=225 avg= 25.6% actual= 25.8% 30- 40% n=341 avg= 35.2% actual= 37.5% 40- 50% n=281 avg= 44.7% actual= 46.6% 50- 60% n=219 avg= 54.5% actual= 55.7% 60- 70% n=151 avg= 64.2% actual= 67.5% 70- 80% n=31 avg= 73.5% actual= 77.4% 80- 90% n=8 avg= 83.1% actual= 87.5%
Draw · ECE 2.22%
0- 10% n=1 avg= 9.2% actual= 0.0% 10- 20% n=78 avg= 17.2% actual= 14.1% 20- 30% n=1139 avg= 26.2% actual= 24.4% 30- 40% n=105 avg= 30.9% actual= 24.8%
Away win · ECE 2.18%
0- 10% n=32 avg= 7.7% actual= 3.1% 10- 20% n=262 avg= 15.6% actual= 13.7% 20- 30% n=347 avg= 25.0% actual= 27.4% 30- 40% n=335 avg= 34.9% actual= 37.3% 40- 50% n=215 avg= 44.4% actual= 45.1% 50- 60% n=88 avg= 54.0% actual= 52.3% 60- 70% n=31 avg= 62.8% actual= 67.7% 70- 80% n=11 avg= 73.3% actual= 63.6% 80- 90% n=2 avg= 80.8% actual=100.0%
BTTS · ECE 2.90%
30- 40% n=1 avg= 36.4% actual=100.0% 40- 50% n=116 avg= 47.4% actual= 50.0% 50- 60% n=794 avg= 55.4% actual= 55.0% 60- 70% n=399 avg= 63.2% actual= 55.4% 70- 80% n=13 avg= 70.9% actual= 76.9%
Over 2.5 · ECE 1.79%
20- 30% n=1 avg= 27.5% actual=100.0% 30- 40% n=19 avg= 37.8% actual= 52.6% 40- 50% n=296 avg= 46.6% actual= 47.0% 50- 60% n=640 avg= 54.8% actual= 54.8% 60- 70% n=319 avg= 64.2% actual= 59.9% 70- 80% n=48 avg= 72.9% actual= 62.5%
Walk-forward - latest periods
| Period | N | Model 1X2 LL | Baseline 1X2 LL | Improvement |
|---|---|---|---|---|
| 2025-08 | 27 | 0.8392 | 0.9769 | +14.10% |
| 2025-09 | 27 | 0.9641 | 1.0440 | +7.66% |
| 2025-10 | 36 | 1.0587 | 1.0873 | +2.63% |
| 2025-11 | 36 | 1.0135 | 1.0690 | +5.19% |
| 2025-12 | 18 | 0.8596 | 0.9932 | +13.45% |
| 2026-01 | 31 | 1.0090 | 1.0915 | +7.56% |
| 2026-02 | 36 | 0.9626 | 1.0503 | +8.35% |
| 2026-03 | 31 | 0.9873 | 1.0913 | +9.53% |
| 2026-04 | 35 | 1.0200 | 1.1091 | +8.04% |
| 2026-05 | 29 | 1.1228 | 1.0814 | -3.83% |
| 2026-08 | 18 | 1.0555 | 1.1231 | +6.02% |
| 2026-09 | 27 | 1.0421 | 1.0693 | +2.54% |
Football is data.
NILYTIX turns match data into transparent model probabilities and measurable analytical signals.