Home win
37.04%
St Johnstone
SCO1 · Scottish Premiership · Match Analysis
St Johnstone vs
Dundee United
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
26.30%
X
Away win
36.66%
Dundee United
Score model
Most likely scorelines.
Exact-score probabilities from the Poisson + Dixon-Coles goal model.
Additional scorelines
2:26.22%
0:06.08%
2:05.70%
0:25.64%
3:14.17%
Team context
Strength and recent form.
Venue-adjusted team strength, Elo rating and recent xG performance.
Home
St Johnstone
Elo1433.2
Overall attack1.33
Overall defense1.65
Home attack1.47
Home defense1.42
Form signal-0.418
Last 5 xGF / xGA1.21 / 1.65
Last 10 xGF / xGA1.26 / 1.65
Last 10 xGD-0.39
League statusSTABLE · from SCO1 2024/25
Away
Dundee United
Elo1487.1
Overall attack1.42
Overall defense1.64
Away attack1.31
Away defense1.72
Form signal+0.555
Last 5 xGF / xGA1.88 / 1.22
Last 10 xGF / xGA1.66 / 1.29
Last 10 xGD+0.37
League statusSTABLE · from SCO1 2025/26
Elo difference
-53.9
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.
06.08%
113.48%
223.59%
322.40%
416.59%
59.83%
64.86%
7+3.17%
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
LeagueSCO1
Loaded teams14
Loaded historical matches960
Target season2026/27
Previous seasons in history4
Season window2022/23 → 2023/24 → 2024/25 → 2025/26 → 2026/27
League tierTier 1
Historical league matches960
xG matches276 / 960
Effective xG matches219.17
Observed home xG1.59
Observed away xG1.29
Model home xG baseline1.61
Model away xG baseline1.27
Total league xG baseline2.88
Home-advantage weight1.125
Model configdata/models/SCO1.json
Team attack half-life240.0 days
Team defense half-life240.0 days
Attack strength weight0.763
Defense strength weight0.969
Venue prior matches2.00
DistributionPoisson + Dixon-Coles
Dixon-Coles rho-0.081 (effective -0.081)
Dixon-Coles decay0.016
Goal overdispersion0.000
Implied goal variance H / A1.49 / 1.48
Relative strengths and form diagnostics
St Johnstone
Home attack strength0.972
Home defense weakness1.042
Base lambda home1.66
Attack form xGF level1.232
Defense form xGA level (lower is better)1.650
Opponent-adjusted attack residual-0.182
Opponent-adjusted defense residual (higher is better)-0.187
Opponent-adjustment weight0.000
Model attack signal-0.208
Model defense signal-0.210
Opponent blend0.000
Form matches 5 / 105 / 8
Last 5 xGF / xGA / xGD1.21 / 1.65 / -0.43
Last 10 xGF / xGA / xGD1.26 / 1.65 / -0.39
Dundee United
Away attack strength1.010
Away defense weakness1.055
Base lambda away1.33
Attack form xGF level1.802
Defense form xGA level (lower is better)1.247
Opponent-adjusted attack residual+0.373
Opponent-adjusted defense residual (higher is better)+0.203
Opponent-adjustment weight0.000
Model attack signal+0.363
Model defense signal+0.193
Opponent blend0.000
Form matches 5 / 105 / 10
Last 5 xGF / xGA / xGD1.88 / 1.22 / +0.65
Last 10 xGF / xGA / xGD1.66 / 1.29 / +0.37
Historical data coverage
St Johnstone
Total matches122
Home / away matches61 / 61
xG matches8 / 122
xG home / away coverage4 / 61 · 4 / 61
Effective total7.06
Effective attack / defense7.06 / 7.06
Effective home matches3.51
Effective home attack / defense3.51 / 3.51
Dundee United
Total matches122
Home / away matches61 / 61
xG matches46 / 122
xG home / away coverage23 / 61 · 23 / 61
Effective total24.03
Effective attack / defense24.03 / 24.03
Effective away matches12.05
Effective away attack / defense12.05 / 12.05
Historical data coverage
29.13%
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 940 historical matches.
Positive improvement means the model performed better than the league baseline.
Total goals log-loss vs baseline+0.24%
1X2 log-loss vs baseline+6.86%
BTTS Brier vs baseline+0.67%
Over 2.5 Brier vs baseline+1.16%
Full backtest metrics
Model
MAE total goals1.2931
Exact score log-loss2.9768
Total goals Brier0.8253
Total goals log-loss1.8746
1X2 Brier0.5894
1X2 log-loss0.9903
BTTS Brier0.2497
Over 1.5 Brier0.1699
Over 2.5 Brier0.2472
Over 3.5 Brier0.2185
League baseline
MAE total goals1.3025
Exact score log-loss3.0841
Total goals Brier0.8267
Total goals log-loss1.8791
1X2 Brier0.6424
1X2 log-loss1.0632
BTTS Brier0.2514
Over 1.5 Brier0.1705
Over 2.5 Brier0.2501
Over 3.5 Brier0.2197
Calibration diagnostics
Home win · ECE 5.29%
0- 10% n=4 avg= 8.5% actual= 25.0% 10- 20% n=31 avg= 15.5% actual= 22.6% 20- 30% n=103 avg= 24.8% actual= 14.6% 30- 40% n=201 avg= 36.0% actual= 36.3% 40- 50% n=369 avg= 44.1% actual= 50.1% 50- 60% n=111 avg= 54.3% actual= 54.1% 60- 70% n=84 avg= 64.4% actual= 75.0% 70- 80% n=24 avg= 74.1% actual= 87.5% 80- 90% n=11 avg= 85.0% actual= 90.9% 90-100% n=2 avg= 92.4% actual=100.0%
Draw · ECE 3.10%
0- 10% n=8 avg= 8.6% actual= 12.5% 10- 20% n=48 avg= 17.0% actual= 8.3% 20- 30% n=880 avg= 26.6% actual= 23.9% 30- 40% n=4 avg= 30.7% actual= 25.0%
Away win · ECE 3.20%
0- 10% n=32 avg= 6.4% actual= 3.1% 10- 20% n=135 avg= 15.3% actual= 16.3% 20- 30% n=330 avg= 26.3% actual= 23.3% 30- 40% n=282 avg= 33.7% actual= 33.0% 40- 50% n=86 avg= 44.9% actual= 54.7% 50- 60% n=53 avg= 53.6% actual= 64.2% 60- 70% n=14 avg= 65.4% actual= 64.3% 70- 80% n=8 avg= 74.0% actual= 50.0%
BTTS · ECE 2.32%
30- 40% n=3 avg= 37.8% actual= 33.3% 40- 50% n=91 avg= 46.9% actual= 44.0% 50- 60% n=829 avg= 54.1% actual= 52.1% 60- 70% n=17 avg= 61.6% actual= 47.1%
Over 2.5 · ECE 5.06%
30- 40% n=6 avg= 38.5% actual= 50.0% 40- 50% n=481 avg= 48.7% actual= 51.8% 50- 60% n=401 avg= 53.0% actual= 60.1% 60- 70% n=40 avg= 63.5% actual= 60.0% 70- 80% n=11 avg= 73.8% actual= 54.5% 80- 90% n=1 avg= 82.7% actual=100.0%
Walk-forward - latest periods
| Period | N | Model 1X2 LL | Baseline 1X2 LL | Improvement |
|---|---|---|---|---|
| 2025-10 | 23 | 1.0089 | 1.0453 | +3.48% |
| 2025-11 | 22 | 0.9725 | 1.1161 | +12.87% |
| 2025-12 | 34 | 0.9366 | 1.0299 | +9.06% |
| 2026-01 | 25 | 0.8051 | 1.0731 | +24.98% |
| 2026-02 | 29 | 0.9630 | 1.0379 | +7.22% |
| 2026-03 | 16 | 1.1526 | 1.0513 | -9.63% |
| 2026-04 | 18 | 0.9239 | 1.0121 | +8.72% |
| 2026-05 | 24 | 1.0380 | 1.0573 | +1.82% |
| 2026-07 | 1 | 1.6304 | 1.3147 | -24.01% |
| 2026-08 | 18 | 0.9214 | 1.0569 | +12.82% |
| 2026-09 | 23 | 1.0485 | 1.1083 | +5.40% |
| 2026-10 | 6 | 0.8378 | 1.0221 | +18.03% |
Football is data.
NILYTIX turns match data into transparent model probabilities and measurable analytical signals.