Premier League table 2026/2027
- Leader Manchester City 15 pts
- Most above xPts Manchester City +5.66 against expected points
- Most below xPts Bournemouth −3.93 against expected points
The standings, with the expected figures beside them: what each club's chances were worth, match by match, next to what it actually took.
Showing non-penalty xG. Penalties are removed from xG, xGA and xGD; xPts is unchanged.
| # restore the table order | Team | P sort descending | W sort descending | D sort descending | L sort descending | GF sort descending | GA sort descending | GD sort descending | Pts sort descending | npxG sort descending | npxGA sort descending | npxGD sort descending | xPts sort descending | Pts vs xPts sort descending |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Manchester City MCI | 5 | 5 | 0 | 0 | 13 | 5 | +8 | 15 | 9.34 | 8.00 | +1.34 | 9.34 | +5.66 |
| 2 | Arsenal ARS | 5 | 4 | 0 | 1 | 8 | 4 | +4 | 12 | 7.72 | 3.15 | +4.57 | 10.36 | +1.64 |
| 3 | Brighton & Hove Albion BHA | 5 | 3 | 1 | 1 | 16 | 5 | +11 | 10 | 10.64 | 9.22 | +1.42 | 8.12 | +1.88 |
| 4 | Brentford BRE | 5 | 2 | 3 | 0 | 10 | 4 | +6 | 9 | 9.48 | 6.08 | +3.40 | 8.29 | +0.71 |
| 5 | Leeds United LEE | 5 | 2 | 3 | 0 | 7 | 3 | +4 | 9 | 8.11 | 5.72 | +2.39 | 7.75 | +1.25 |
| 6 | Liverpool LIV | 5 | 2 | 3 | 0 | 7 | 4 | +3 | 9 | 6.91 | 5.13 | +1.78 | 8.43 | +0.57 |
| 7 | Everton EVE | 5 | 2 | 3 | 0 | 6 | 3 | +3 | 9 | 7.39 | 6.86 | +0.53 | 7.24 | +1.76 |
| 8 | Hull City HUL | 5 | 2 | 2 | 1 | 6 | 4 | +2 | 8 | 5.37 | 6.36 | −0.99 | 5.43 | +2.57 |
| 9 | Newcastle United NEW | 5 | 2 | 2 | 1 | 9 | 9 | 0 | 8 | 4.32 | 7.87 | −3.55 | 4.08 | +3.92 |
| 10 | Chelsea CHE | 5 | 2 | 1 | 2 | 10 | 12 | −2 | 7 | 6.84 | 9.64 | −2.80 | 5.82 | +1.18 |
| 11 | Ipswich Town IPS | 5 | 2 | 0 | 3 | 7 | 11 | −4 | 6 | 7.47 | 8.64 | −1.17 | 7.03 | −1.03 |
| 12 | Manchester United MUN | 5 | 1 | 2 | 2 | 8 | 8 | 0 | 5 | 9.36 | 6.99 | +2.37 | 7.99 | −2.99 |
| 13 | Nottingham Forest NFO | 5 | 1 | 2 | 2 | 4 | 5 | −1 | 5 | 5.88 | 5.20 | +0.68 | 8.59 | −3.59 |
| 14 | Sunderland SUN | 5 | 1 | 1 | 3 | 6 | 10 | −4 | 4 | 8.85 | 7.83 | +1.02 | 7.35 | −3.35 |
| 15 | Crystal Palace CRY | 5 | 1 | 1 | 3 | 6 | 11 | −5 | 4 | 6.26 | 8.81 | −2.55 | 4.76 | −0.76 |
| 16 | Aston Villa AVL | 5 | 1 | 1 | 3 | 4 | 9 | −5 | 4 | 4.40 | 8.66 | −4.26 | 4.50 | −0.50 |
| 17 | Bournemouth BOU | 5 | 0 | 3 | 2 | 6 | 8 | −2 | 3 | 7.39 | 6.83 | +0.56 | 6.93 | −3.93 |
| 18 | Coventry City COV | 5 | 1 | 0 | 4 | 1 | 10 | −9 | 3 | 4.88 | 7.39 | −2.51 | 4.33 | −1.33 |
| 19 | Fulham FUL | 5 | 0 | 2 | 3 | 5 | 8 | −3 | 2 | 7.92 | 8.14 | −0.22 | 5.86 | −3.86 |
| 20 | Tottenham Hotspur TOT | 5 | 0 | 2 | 3 | 2 | 8 | −6 | 2 | 4.96 | 6.97 | −2.01 | 5.86 | −3.86 |
- More points than xPts
- Fewer points than xPts
How to read this table
- xG
- Expected goals: the summed chance quality of the shots a club took. A chance scored nine times in ten is worth 0.9.
- xGA
- Expected goals against: the same measure for the shots a club allowed.
- xGD
- xG minus xGA: how far a club's chances outweighed its opponents', or the other way round.
- xPts
- Expected points: the points a club's match-by-match xG would be worth on average. Each match's two xG figures become chances of a win, a draw and a loss, and a win is worth 3 and a draw 1.
- Pts vs xPts
- Actual points minus expected points based on match xG. Positive, drawn in green, is more points than the chances were worth on average; negative, in red, is fewer. It measures the gap and says nothing about its cause.
- npxG
- Non-penalty xG: xG with penalties removed, offered where the competition's figures support it. It changes xG, xGA and xGD; xPts is always worked out from all xG.
Expected points are read off the same score matrix that prices the board: each played match's two xG totals are taken as scoring rates, and the team's share is 3 × P(win) + 1 × P(draw), summed over the season. Understat arrives at the same figure by simulating each shot, which needs per-shot data, held here for %{competitions}. Passes per defensive action and deep completions are not shown because no source here ingests the event data behind them.