Premier League
Retrospective analysis — the model is trained only on matches played before this one.
The model's probabilities for a home win, draw and away win — with reasoning and recommended markets.
Unlock with ProAlready have an account? Log in
Man City
Bournemouth
Expected goals, over/under and both teams to score — priced by the model before kick-off.
Unlock with ProAlready have an account? Log in
The model's top final scores with the best bookmaker odds.
Unlock with ProAlready have an account? Log in
Half-time score: 4–1
| # | Team | K | V | U | T | +/− | P |
|---|---|---|---|---|---|---|---|
| 1 |
Man City
|
8 | 6 | 0 | 2 | +12 | 18 |
| 2 |
Arsenal
|
8 | 5 | 1 | 2 | +6 | 16 |
| 3 |
Man United
|
8 | 5 | 1 | 2 | +4 | 16 |
| 4 |
Crystal Palace
|
8 | 5 | 0 | 3 | +4 | 15 |
| 5 |
Leicester
|
8 | 4 | 3 | 1 | +2 | 15 |
| 6 |
West Ham
|
8 | 4 | 2 | 2 | +6 | 14 |
| 7 |
Tottenham
|
9 | 3 | 5 | 1 | +4 | 14 |
| 8 |
Everton
|
8 | 3 | 4 | 1 | +4 | 13 |
| 9 |
Liverpool
|
9 | 3 | 4 | 2 | -2 | 13 |
| 10 |
Southampton
|
8 | 3 | 3 | 2 | +3 | 12 |
| 11 |
Swansea
|
8 | 2 | 4 | 2 | +0 | 10 |
| 12 |
Watford
|
8 | 2 | 4 | 2 | -1 | 10 |
| 13 |
Norwich
|
8 | 2 | 3 | 3 | -2 | 9 |
| 14 |
Stoke
|
8 | 2 | 3 | 3 | -2 | 9 |
| 15 |
Bournemouth
|
8 | 2 | 2 | 4 | -2 | 8 |
| 16 |
Chelsea
|
8 | 2 | 2 | 4 | -5 | 8 |
| 17 |
West Brom
|
8 | 2 | 2 | 4 | -5 | 8 |
| 18 |
Aston Villa
|
8 | 1 | 1 | 6 | -5 | 4 |
| 19 |
Sunderland
|
8 | 0 | 3 | 5 | -10 | 3 |
| 20 |
Newcastle
|
8 | 0 | 3 | 5 | -11 | 3 |
Calculated from our results data. “xG +/−” is expected goal difference — a positive number with a low league position suggests an undervalued team. Ties on points are sorted by goal difference; some leagues use other rules.
K. De Bruyne
M
A. Smith
M
Green columns are the outcomes that came in.
Man City
Newcastle
Tottenham
Man City
Man City
West Ham
Crystal Palace
Man City
Man City
Watford
Mike Dean · 154 kamper i vår database
🟥 Red cards: 0,22 per match
The analysis is a statistical estimate based on the teams' results and underlying performances (xG) — not a guarantee. See the model's results.