Forecasts for the Portuguese football league
Matchday 5 · Updated September 6, 2026
Predicted standings
Predicted standings based on 50,000 simulations. Mean points and probabilities for each outcome.
| # | Team | Pts | Predicted pts | Champion | Relegation |
|---|---|---|---|---|---|
| 1 |  SCP | 13 | 80.2 70–90 | 55% | — |
| 2 |  POR | 15 | 75.9 65–86 | 22%2 | — |
| 3 |  SLB | 10 | 75.5 65–86 | 22%2 | — |
| 4 |  BRA | 7 | 62.4 51–74 | <1% | — |
| 5 |  STC | 11 | 47.5 37–59 | — | <1%1 |
| 6 |  ARO | 9 | 45.3 34–57 | — | 1% |
| 7 |  GIL | 7 | 44.6 33–56 | — | 1%1 |
| 8 |  VSC | 4 | 44.4 33–56 | — | 1% |
| 9 |  FAM | 3 | 42.4 31–54 | — | 3% |
| 10 |  EST | 1 | 38.0 27–50 | — | 12% |
| 11 |  MOR | 4 | 38.0 27–50 | — | 12% |
| 12 |  MAR | 7 | 37.6 26–49 | — | 14%1 |
| 13 |  NAC | 4 | 37.3 26–49 | — | 13% |
| 14 |  ALV | 2 | 37.2 26–49 | — | 14%2 |
| 15 |  ACV | 5 | 36.7 25–49 | — | 17%9 |
| 16 |  EAM | 6 | 36.5 26–48 | — | 16% |
| 17 |  RIO | 3 | 34.5 24–46 | — | 26%5 |
| 18 |  CPA | 1 | 28.6 18–39 | — | 62%1 |
Final points are a distribution, not a number: the bar shows where the 50k simulated seasons fall. These intervals were verified as calibrated over 8 historical seasons — an interval quoted as 90% contained the real outcome 92.7% of the time.
Luck or merit?
Expected points (xPts) computed from each match's expected goals (xG). The difference to actual points measures finishing efficiency and luck.
AROPORMORSCPSTCEAMMARBRAACVSLBRIONACGILALVESTCPAFAMVSCxG data: FotMob
Title race
Championship probability evolution throughout the season.
Fixtures to come
What is left of this matchday, then the next one. Open a fixture for the full preview.
Relegation battle
Relegation probability evolution throughout the season.
Team strengths
Attack and defense parameters estimated by the Bayesian model.
SCPPORSLBBRAFAMVSCSTCGILESTAROALVNACMORMARACVRIOEAMCPAAbout this model
We simulated the rest of the season 50,000 times, varying each match result based on team performance, squad value, and home advantage. The probabilities shown reflect how often each outcome occurs across those simulations.
Model vs Market
We tested the model against Pinnacle's closing line over 504 matches and eight seasons. From matchday 14 it matches the market — the whole deficit is early season.
The 2025-26 season, reviewed
Porto took the title on 88 points while Sporting scored 89 goals and finished second, and Benfica went unbeaten into third. What the xG says about who deserved it — and how our own forecasts held up.
The players, measured honestly
Finishing, attacking contribution, contested possession, goalkeeping — one metric per dimension, each with its own scale and its own uncertainty, and no fake overall score.
Liga 2 in probabilities
Promotion and relegation probabilities for the second tier, from 1,836 matches over six seasons. A deliberately lighter model: goals only, no xG and no squad values.