Home win
43.43%
Werder Bremen
GER1 · Bundesliga · Match Analysis
Werder Bremen vs
SC Paderborn
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
28.99%
X
Away win
27.59%
SC Paderborn
Score model
Most likely scorelines.
Exact-score probabilities from the Poisson + Dixon-Coles goal model.
Additional scorelines
0:16.99%
1:26.95%
2:25.08%
0:24.75%
3:14.41%
Team context
Strength and recent form.
Venue-adjusted team strength, Elo rating and recent xG performance.
Home
Werder Bremen
Elo1478.5
Overall attack1.33
Overall defense1.62
Home attack1.52
Home defense1.39
Form signal-0.383
Last 5 xGF / xGA1.46 / 1.78
Last 10 xGF / xGA1.34 / 1.84
Last 10 xGD-0.50
League statusSTABLE · from GER1 2025/26
Away
SC Paderborn
Elo1484.7
Overall attack0.94
Overall defense1.66
Away attack0.61
Away defense1.50
Form signal-0.721
Last 5 xGF / xGA1.03 / 1.75
Last 10 xGF / xGA1.03 / 1.75
Last 10 xGD-0.72
League statusNEW
Elo difference
-6.1
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.
09.02%
116.53%
226.67%
321.71%
414.03%
57.25%
63.12%
7+1.67%
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
LeagueGER1
Loaded teams25
Loaded historical matches1269
Target season2026/27
Previous seasons in history4
Season window2022/23 → 2023/24 → 2024/25 → 2025/26 → 2026/27
League tierTier 1
Historical league matches1269
xG matches1267 / 1269
Effective xG matches638.73
Observed home xG1.74
Observed away xG1.36
Model home xG baseline1.70
Model away xG baseline1.39
Total league xG baseline3.09
Home-advantage weight0.831
Model configdata/models/GER1.json
Team attack half-life365.0 days
Team defense half-life240.0 days
Attack strength weight1.097
Defense strength weight0.981
Venue prior matches6.50
DistributionPoisson + Dixon-Coles
Dixon-Coles rho-0.120 (effective -0.120)
Dixon-Coles decay0.016
Goal overdispersion0.000
Implied goal variance H / A1.46 / 1.12
Relative strengths and form diagnostics
Werder Bremen
Home attack strength0.898
Home defense weakness1.023
Base lambda home1.51
Attack form xGF level1.421
Defense form xGA level (lower is better)1.804
Opponent-adjusted attack residual-0.054
Opponent-adjusted defense residual (higher is better)-0.184
Opponent-adjustment weight0.000
Model attack signal-0.125
Model defense signal-0.258
Opponent blend0.000
Form matches 5 / 105 / 10
Last 5 xGF / xGA / xGD1.46 / 1.78 / -0.32
Last 10 xGF / xGA / xGD1.34 / 1.84 / -0.50
SC Paderborn
Away attack strength0.850
Away defense weakness0.997
Base lambda away1.19
Attack form xGF level1.031
Defense form xGA level (lower is better)1.752
Opponent-adjusted attack residual-0.559
Opponent-adjusted defense residual (higher is better)-0.250
Opponent-adjustment weight0.000
Model attack signal-0.515
Model defense signal-0.206
Opponent blend0.000
Form matches 5 / 105 / 5
Last 5 xGF / xGA / xGD1.03 / 1.75 / -0.72
Last 10 xGF / xGA / xGD1.03 / 1.75 / -0.72
Historical data coverage
Werder Bremen
Total matches141
Home / away matches70 / 71
xG matches141 / 141
xG home / away coverage70 / 70 · 71 / 71
Effective total27.99
Effective attack / defense42.49 / 27.99
Effective home matches13.63
Effective home attack / defense20.84 / 13.63
SC Paderborn
Total matches5
Home / away matches3 / 2
xG matches5 / 5
xG home / away coverage3 / 3 · 2 / 2
Effective total4.56
Effective attack / defense4.71 / 4.56
Effective away matches1.77
Effective away attack / defense1.85 / 1.77
Historical data coverage
26.34%
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 1249 historical matches.
Positive improvement means the model performed better than the league baseline.
Total goals log-loss vs baseline+1.19%
1X2 log-loss vs baseline+7.92%
BTTS Brier vs baseline+0.78%
Over 2.5 Brier vs baseline+2.40%
Full backtest metrics
Model
MAE total goals1.3963
Exact score log-loss3.1021
Total goals Brier0.8384
Total goals log-loss1.9562
1X2 Brier0.5886
1X2 log-loss0.9889
BTTS Brier0.2385
Over 1.5 Brier0.1358
Over 2.5 Brier0.2321
Over 3.5 Brier0.2377
League baseline
MAE total goals1.4093
Exact score log-loss3.2441
Total goals Brier0.8429
Total goals log-loss1.9798
1X2 Brier0.6498
1X2 log-loss1.0740
BTTS Brier0.2404
Over 1.5 Brier0.1375
Over 2.5 Brier0.2378
Over 3.5 Brier0.2474
Calibration diagnostics
Home win · ECE 3.36%
0- 10% n=25 avg= 7.5% actual= 8.0% 10- 20% n=80 avg= 16.0% actual= 17.5% 20- 30% n=183 avg= 25.9% actual= 21.3% 30- 40% n=279 avg= 35.3% actual= 38.4% 40- 50% n=277 avg= 44.9% actual= 43.0% 50- 60% n=182 avg= 54.3% actual= 53.8% 60- 70% n=125 avg= 64.4% actual= 73.6% 70- 80% n=64 avg= 74.4% actual= 73.4% 80- 90% n=30 avg= 84.4% actual= 93.3% 90-100% n=4 avg= 91.5% actual= 25.0%
Draw · ECE 1.45%
0- 10% n=14 avg= 8.1% actual= 14.3% 10- 20% n=158 avg= 16.1% actual= 17.7% 20- 30% n=1008 avg= 25.9% actual= 26.5% 30- 40% n=69 avg= 31.0% actual= 18.8%
Away win · ECE 2.01%
0- 10% n=74 avg= 6.6% actual= 6.8% 10- 20% n=214 avg= 15.4% actual= 12.1% 20- 30% n=338 avg= 25.0% actual= 26.0% 30- 40% n=281 avg= 34.7% actual= 35.6% 40- 50% n=178 avg= 44.2% actual= 41.6% 50- 60% n=86 avg= 54.4% actual= 57.0% 60- 70% n=47 avg= 64.1% actual= 61.7% 70- 80% n=22 avg= 74.8% actual= 59.1% 80- 90% n=9 avg= 82.6% actual= 88.9%
BTTS · ECE 1.91%
40- 50% n=85 avg= 47.7% actual= 49.4% 50- 60% n=766 avg= 55.8% actual= 56.3% 60- 70% n=378 avg= 63.4% actual= 68.3% 70- 80% n=20 avg= 71.0% actual= 70.0%
Over 2.5 · ECE 4.25%
30- 40% n=21 avg= 37.7% actual= 42.9% 40- 50% n=215 avg= 46.4% actual= 50.7% 50- 60% n=543 avg= 55.2% actual= 60.4% 60- 70% n=343 avg= 64.3% actual= 67.3% 70- 80% n=107 avg= 73.7% actual= 76.6% 80- 90% n=20 avg= 83.4% actual= 80.0%
Walk-forward - latest periods
| Period | N | Model 1X2 LL | Baseline 1X2 LL | Improvement |
|---|---|---|---|---|
| 2025-09 | 27 | 0.8925 | 1.0895 | +18.08% |
| 2025-10 | 28 | 0.8254 | 1.0409 | +20.70% |
| 2025-11 | 35 | 0.9936 | 1.0578 | +6.06% |
| 2025-12 | 27 | 1.0363 | 1.0708 | +3.22% |
| 2026-01 | 42 | 1.0133 | 1.0827 | +6.41% |
| 2026-02 | 35 | 0.9249 | 1.0164 | +9.00% |
| 2026-03 | 31 | 0.8909 | 1.0953 | +18.66% |
| 2026-04 | 36 | 0.9567 | 1.1201 | +14.58% |
| 2026-05 | 27 | 1.1104 | 1.0679 | -3.97% |
| 2026-08 | 9 | 0.8221 | 0.9155 | +10.20% |
| 2026-09 | 27 | 0.9516 | 1.0644 | +10.61% |
| 2026-10 | 9 | 1.1236 | 1.0674 | -5.26% |
Football is data.
NILYTIX turns match data into transparent model probabilities and measurable analytical signals.