← Premier League 2026-27

Prediction Arena

Pick any two sides, set the scene with the storyline sliders, and see how the model reads the match. Then go beat it.

v

First side plays at home — home advantage is built into the league model.

Model verdict

Arsenal 47%Draw 28%Liverpool 26%
Arsenal rating
1673
Liverpool rating
1663
Arsenal expected goals
1.35
Liverpool expected goals
0.92

Most likely scorelines

  • 1–014%
  • 1–113%
  • 0–010%
  • 0–110%
  • 2–09%

Beat the model

Call it before you see the model lean.

Live Form Model

Members

This season's form read against each club's established level — it updates after every final whistle and is honest about when this season disagrees with that baseline.

Checking your membership…

How the model works

Online Elo with an estimated home-advantage term plus a recency-weighted Poisson goal model, updated after every final whistle.

Matches this season
40
Home-field bump
17 Elo
Goals per side
1.42
Live formTuned to this season's form — 40 matches in.

Proven on this season's results: about 5% sharper than a coin-flip on log loss. Predictions are guidance, not betting advice.

Power ranking

Strongest sides by the model's learned rating.

  1. 1Arsenal1673
  2. 2Liverpool1663
  3. 3Manchester City1642
  4. 4Chelsea1576
  5. 5Brighton & Hove Albion1571
  6. 6Newcastle United1569
  7. 7Nottingham Forest1545
  8. 8Bournemouth1537
  9. 9Brentford1528
  10. 10Everton1499