Premier League 2026/27 — Projected Final Table
Our AI's projected final Premier League 2026/27 table, simulated from every remaining fixture.
| # | Team | vs last week | avg Pts | avg GD | Title % | Top 4 % | Rel % |
|---|---|---|---|---|---|---|---|
| 1 | Arsenal | ▲ 1 | 80.6 | +42.7 | 51.2 | 97.5 | 0.0 |
| 2 | Manchester City | ▼ 1 | 78.3 | +38.6 | 35.3 | 95.3 | 0.0 |
| 3 | Chelsea | ▲ 2 | 70.9 | +25.0 | 8.0 | 74.7 | 0.0 |
| 4 | Brighton & Hove Albion | ▲ 5 | 68.2 | +26.9 | 4.4 | 63.2 | 0.0 |
| 5 | Liverpool | ▼ 2 | 60.9 | +13.9 | 0.4 | 22.2 | 0.1 |
| 6 | Brentford | ▲ 8 | 60.5 | +13.5 | 0.4 | 20.3 | 0.1 |
| 7 | Everton | ▲ 12 | 56.9 | +6.9 | 0.1 | 8.8 | 0.2 |
| 8 | Leeds United | ▲ 2 | 54.6 | +2.1 | 0.1 | 5.0 | 0.6 |
| 9 | Newcastle United | ▼ 1 | 54.3 | +3.0 | 0.0 | 5.1 | 0.7 |
| 10 | Ipswich Town | ▲ 7 | 54.1 | -0.3 | 0.0 | 4.6 | 0.9 |
| 11 | Manchester United | ▼ 7 | 52.1 | +0.3 | 0.0 | 2.8 | 1.4 |
| 12 | Hull City | ▲ 6 | 45.7 | -12.7 | 0.0 | 0.2 | 8.3 |
| 13 | Tottenham Hotspur | ▼ 7 | 43.9 | -13.7 | 0.0 | 0.1 | 13.3 |
| 14 | Nottingham Forest | ▼ 1 | 42.3 | -12.6 | 0.0 | 0.1 | 18.5 |
| 15 | Crystal Palace | – | 39.9 | -19.5 | 0.0 | 0.0 | 28.9 |
| 16 | Sunderland | – | 38.6 | -20.9 | 0.0 | 0.0 | 36.1 |
| 17 | Bournemouth | ▼ 5 | 38.3 | -19.7 | 0.0 | 0.0 | 37.3 |
| 18 | Fulham | ▼ 7 | 37.8 | -19.1 | 0.0 | 0.0 | 39.5 |
| 19 | Aston Villa | ▼ 12 | 37.3 | -23.9 | 0.0 | 0.0 | 43.8 |
| 20 | Coventry City | – | 32.6 | -30.5 | 0.0 | 0.0 | 70.4 |
Movement shows how each club's projected finish has changed since the previous weekly run.
Last updated: Tue, 1 Sept 2026
How this projection works
Every remaining fixture is simulated using FootyPulse's AI win / draw / away probability and predicted score, calibrated as a bivariate-Poisson so outcome and goal difference stay internally consistent. Completed fixtures contribute their real result.
This projection is a weekly snapshot, re-run automatically every Tuesday morning at 06:00 UK time, once the full gameweek (including Monday-night matches) is finished and results are synced. It doesn't update live during matches.
Each entry in the projection selector is the actual projection published at that moment — saved run by run and never recalculated. The list grows every week through the season.
Note: for 13 of 380 fixtures (3.4%), our model couldn't fully match both the predicted scoreline and the stated win / draw / away probabilities at the same time — we used the closest realistic match instead. This affects a small number of individual match simulations.