Five games in and the Premier League table already has a habit of lying. A year ago Liverpool led the division after five wins from five and Tottenham sat third; both finished nowhere near those positions. Expected goals (xG) data — which strips out finishing quality, goalkeeping heroics and pure luck to measure how well teams have actually performed — offers a more honest early-season verdict.

Arsenal still the benchmark, despite Brighton hammering

Champions Arsenal posted the best xG difference across the opening five rounds, even after conceding what the BBC Sport statistician Chris Collinson describes as their heaviest defeat in three years, at Brighton. That context matters: one bad afternoon has not undone a broadly dominant start. Manchester City sit just behind them in xG difference, and Brentford's unbeaten run appears genuinely earned rather than fortunate.

Manchester United are the most intriguing case. Their start has felt erratic to most observers, yet only three clubs have a higher xG difference. The data suggests they have been, in Collinson's words, «simply fine» more often than the discourse implies.

Brighton's attack is the division's sharpest tool

Going forward, Brighton lead the division by xG generated, ahead of Manchester City, Brentford and Manchester United. What makes the Seagulls' numbers particularly striking is their clinical edge: 15 non-penalty goals from 10.7 xG across five matches is ruthless conversion by any measure.

The only side more clinical in front of goal is Newcastle, who have scored nine goals from just 4.5 xG — essentially twice the expected return. Head coach Matthias Jaissle will be well aware that sustaining that level of overperformance across 38 games is close to impossible, and the underlying numbers confirm Newcastle have a lot of work to do to create chances at a rate that matches their ambitions.

At the other end of the shooting spectrum, Coventry, Sunderland, Nottingham Forest and Tottenham have each scored around three goals fewer than their xG suggested they should have. Whether that reflects a genuine finishing problem or a short-term dip in luck is the question their managers will be trying to answer.

Defensive standings: Arsenal clear, London clubs struggling

Defensively, Arsenal remain in a class of their own. Even factoring in the Brighton result, no side has conceded as few «expected» goals. Forest's Oliver Glasner has built a solid defensive base even if the goals haven't flowed at the other end, and Andi Iraola has done similar work at Liverpool.

The picture in London is less encouraging. Crystal Palace, Fulham and Chelsea have all been too open at the back (note: Chelsea's underlying defensive xG against is not quite as alarming as the 12 goals they have conceded suggests, so some misfortune is in the mix).

Aston Villa, Crystal Palace and Newcastle rank as the three least impressive sides overall by xG difference — Newcastle's position in the top half making that the most surprising finding. Villa fans may take some comfort from last season's precedent: the club made a similarly slow start by these metrics before finishing fourth.

Promoted sides refusing to be passengers

The headline from the promoted clubs is encouraging for the neutrals. Both Hull and Ipswich have made what Collinson calls «genuinely solid starts», suggesting neither side will be easy to push towards the drop zone. Everton and Leeds have also conceded three goals fewer than their defensive xG against implied, hinting at strong goalkeeping or a decent slice of luck — probably some of both.

The sample size caveat applies throughout: five matches is a tiny window, fixture difficulty varies, and the market for who goes down or finishes top four shifts on the thinnest of evidence at this stage. But xG at least gives a clearer baseline than the raw table — and right now it says Arsenal and City are where you'd expect them, while Newcastle and Villa have work to do.

Frequently asked

What does xG mean in football?
xG stands for expected goals. It measures the quality and quantity of chances a team creates or concedes, based on factors like shot location and type, to estimate how many goals a team 'should' score or concede — before finishing ability and luck come into play.
Which Premier League team has the best xG so far this season?
Arsenal have the best overall xG difference (goals created minus goals conceded, on an expected basis) through the opening five matches, with Manchester City close behind. Brighton lead the division purely for attacking xG generated.
Why are Newcastle near the bottom of the xG table if they're winning games?
Newcastle have scored roughly twice as many goals as their xG suggested they should — 9 goals from just 4.5 xG. They've been extremely clinical, but sustaining that level of finishing over a full season is very unlikely, which is why the underlying numbers flag them as a concern despite their results.