The Premier League is the deepest league in world football — a top half full of Champions-League-calibre squads and a bottom half that spends more than most title winners elsewhere. That depth makes single results noisy and season-long simulation genuinely useful: our model rates all 20 squads with results-driven Elo ratings and a Poisson goal model (with shots-on-target-based xG), then plays out the remaining fixture list 10,000 times to turn team strength into title, top-four, Europe and relegation probabilities.
| Pos | Team | Pld | Pts | Proj Pts | Title % | Top 4 % | Europe % | Releg. % |
|---|---|---|---|---|---|---|---|---|
| 2.13 | Liverpool FC | 0 | 0 | 77 | 38.6% | 95.0% | 99.1% | 0% |
| 2.17 | Arsenal FC | 0 | 0 | 77 | 36.4% | 95.1% | 99.2% | 0% |
| 2.73 | Manchester City FC | 0 | 0 | 73 | 21.5% | 89.5% | 97.3% | 0% |
| 6.05 | Manchester United FC | 0 | 0 | 61 | 1.5% | 34.9% | 65.0% | 0.1% |
| 6.54 | Everton FC | 0 | 0 | 60 | 1.1% | 28.6% | 58.4% | 0.3% |
| 7.89 | AFC Bournemouth | 0 | 0 | 57 | 0.3% | 15.6% | 40.6% | 0.9% |
| 7.9 | Brentford FC | 0 | 0 | 57 | 0.4% | 15.8% | 40.9% | 0.9% |
| 9.57 | Aston Villa FC | 0 | 0 | 53 | 0.1% | 7.0% | 23.5% | 2.5% |
| 10.57 | Chelsea FC | 0 | 0 | 51 | 0.1% | 4.8% | 16.4% | 4.4% |
| 11.21 | Leeds United FC | 0 | 0 | 50 | <0.1% | 3.2% | 12.5% | 5.9% |
| 11.35 | Newcastle United FC | 0 | 0 | 49 | <0.1% | 2.8% | 11.7% | 6.6% |
| 11.62 | Brighton & Hove Albion FC | 0 | 0 | 49 | <0.1% | 2.5% | 10.7% | 7.1% |
| 11.66 | Fulham FC | 0 | 0 | 49 | 0.1% | 2.4% | 10.3% | 7.5% |
| 12.68 | Nottingham Forest FC | 0 | 0 | 47 | <0.1% | 1.5% | 6.9% | 11.6% |
| 14.2 | Crystal Palace FC | 0 | 0 | 44 | 0% | 0.6% | 3.0% | 21.1% |
| 15.03 | Sunderland AFC | 0 | 0 | 42 | 0% | 0.3% | 1.8% | 29.5% |
| 15.13 | Coventry City FC | 0 | 0 | 42 | 0% | 0.3% | 1.7% | 30.0% |
| 16.48 | Tottenham Hotspur FC | 0 | 0 | 39 | 0% | <0.1% | 0.6% | 46.4% |
| 17.24 | Ipswich Town FC | 0 | 0 | 37 | 0% | 0.1% | 0.3% | 57.9% |
| 17.85 | Hull City AFC | 0 | 0 | 35 | 0% | <0.1% | 0.1% | 67.2% |
The frozen projection from before matchday 1 — kept unchanged all season so you can judge how the model's pre-season read held up.
| Pos | Team | Pld | Pts | Proj Pts | Title % | Top 4 % | Europe % | Releg. % |
|---|---|---|---|---|---|---|---|---|
| 2.11 | Liverpool FC | 0 | 0 | 77 | 39.2% | 95.3% | 99.1% | 0% |
| 2.18 | Arsenal FC | 0 | 0 | 77 | 36.3% | 94.8% | 99.0% | 0% |
| 2.72 | Manchester City FC | 0 | 0 | 73 | 21.1% | 89.7% | 97.4% | 0% |
| 6.03 | Manchester United FC | 0 | 0 | 61 | 1.4% | 35.5% | 64.8% | 0.1% |
| 6.54 | Everton FC | 0 | 0 | 60 | 1.0% | 28.5% | 58.6% | 0.3% |
| 7.89 | Brentford FC | 0 | 0 | 57 | 0.3% | 15.6% | 41.3% | 0.9% |
| 7.89 | AFC Bournemouth | 0 | 0 | 57 | 0.4% | 15.5% | 41.0% | 0.9% |
| 9.57 | Aston Villa FC | 0 | 0 | 53 | 0.1% | 7.4% | 23.6% | 2.7% |
| 10.55 | Chelsea FC | 0 | 0 | 51 | 0.1% | 4.5% | 16.6% | 4.5% |
| 11.21 | Leeds United FC | 0 | 0 | 50 | 0.1% | 3.2% | 13.0% | 6.0% |
| 11.34 | Newcastle United FC | 0 | 0 | 49 | <0.1% | 3.0% | 11.5% | 6.4% |
| 11.64 | Brighton & Hove Albion FC | 0 | 0 | 49 | 0% | 2.3% | 10.3% | 7.4% |
| 11.7 | Fulham FC | 0 | 0 | 49 | <0.1% | 2.3% | 10.2% | 7.4% |
| 12.67 | Nottingham Forest FC | 0 | 0 | 47 | <0.1% | 1.4% | 6.4% | 11.5% |
| 14.2 | Crystal Palace FC | 0 | 0 | 44 | 0% | 0.4% | 2.9% | 21.7% |
| 15.04 | Sunderland AFC | 0 | 0 | 42 | 0% | 0.3% | 1.8% | 29.0% |
| 15.15 | Coventry City FC | 0 | 0 | 42 | 0% | 0.1% | 1.4% | 29.7% |
| 16.47 | Tottenham Hotspur FC | 0 | 0 | 39 | 0% | <0.1% | 0.6% | 46.1% |
| 17.27 | Ipswich Town FC | 0 | 0 | 37 | 0% | <0.1% | 0.3% | 58.2% |
| 17.84 | Hull City AFC | 0 | 0 | 35 | 0% | 0% | 0.1% | 67.1% |
Every club carries an Elo rating that moves only with results — beat a stronger side and it rises, drop points to a weaker one and it falls. A Poisson goal model (the same engine behind our football totals and handicap predictions, with a shots-on-target-based xG proxy) turns the two teams' ratings into a full scoreline distribution for each remaining fixture.
We then play out the rest of the season 10,000 times, drawing a result for every fixture from those distributions and re-building the table each time. The percentages above are simple counts across those simulated seasons: how often a club finished first, in the top four, in a European spot, or in the relegation places.
The model only knows results and shot data — it cannot see transfers, injuries, managerial changes or squad depth until they show up in performances, so early-season projections lean heavily on last season's level. Probabilities are estimates, not promises: a 5% relegation chance still happens one season in twenty. We publish the projections in full; the value bets our model finds on individual matches are part of the paid plans and are never shown here.