Premier League table: Arsenal and Brighton flattered, Forest the unluckiest

The Premier League table on Saturday 31 October 2026 shows Manchester City and Arsenal level on 21 points, Brighton third on 19, Chelsea fourth on 16 and Liverpool fifth on 15. It also hides some large gaps between what teams have earned and what their chances suggest they should have.

Table figures are from the BBC, updated at 16:57 UK time, so they include the 3pm games but not Tottenham v Crystal Palace or Aston Villa v Fulham, neither of which had finished at the time of writing. Those clubs' places can still move on Saturday night. Expected points (xPTS) are from Understat's match data, summed across the season. Opta's model, quoted by the Guardian, is a different calculation, so the two will not always agree.

Ahead of their numbers

Arsenal have 21 points from eight games and Understat's xPTS puts them at 16.0, a gap of five. They have conceded five goals against an expected 8.1. That is good defending, a good goalkeeper and some luck, and the luck part tends to even out.

Brighton are 19 points on 15.9 xPTS. They have scored 24 league goals from 19.7 xG, and the Guardian, citing Opta before the weekend, has them 6.8 goals above chance quality, the largest gap once penalties are counted. Chelsea were next at 6.7, and the Guardian's piece points to Morgan Rogers (1.87 ahead of his expected goals), Pedro Neto (three league goals) and Jordan Henderson (two from five long-range shots) as unlikely to keep it up. Chelsea have 16 points on 12.5 xPTS, and the Guardian notes that Opta ranks them 12th for expected points.

Newcastle (11 points on 7.4) and Hull (11 on 8.3) are also running warm.

Behind their numbers

Nottingham Forest have six points from nine games and Understat's xPTS gives them 16.6, a gap of 10.6 and the biggest in the division. They have scored six goals from 16.7 xG. The Guardian wrote on Saturday that Forest had gone into the weekend third in Opta's expected table and third from bottom in real life. Against Brentford they had 15 shots to seven on Sky's count and lost 2-0. Liam Delap, a £50m summer signing, has scored once, according to the Guardian, and had three shots with none on target before being substituted. The Guardian says Oliver Glasner has done little wrong beyond chance conversion.

Sunderland have five points on 11.6 xPTS and seven goals from 13.7 xG. Bournemouth have 12 points on 16.6 and have now won six in a row in all competitions. Manchester United are 17th with seven points on 10.8, and ten goals from 14.8 xG. Fulham (8 points on 11.1) also look a little hard done by.

What is fair

Manchester City have 21 points on 19.1 xPTS, so there is no big story. They lost 1-0 to Brighton on Saturday despite 1.92 xG to 1.00 and 21 shots to seven (Sky), and the top of the table is close to deserved. Liverpool (15 on 11.8) and Brentford (14 on 12.6) are a little warm. Coventry (9 on 8.9) and Leeds (11 on 11.8) are about where their numbers say.

What not to conclude

Eight or nine games is a small sample. A single penalty or deflection moves xPTS by a few tenths, and Understat and Opta use different models. Treat the gaps as a prompt to look closer, not a prediction.

The sharpest takeaway is simple: Forest are probably better than the bottom three, and Arsenal's six-point cushion over Liverpool is probably thinner than it looks. The table is the result of what happened. The numbers are the best guess at what would happen next.