Introducing the Performance Gap dashboard: what the table doesn't tell you
Premier League Update
Nine games in, and a clear split is forming between results built to last and results waiting to unwind.
Arsenal are the division’s biggest overperformers, 24 points from an expected 17.03, but it’s defence doing the heavy lifting, just five conceded against an expected 8.59, with only 13 big chances allowed all season, the tightest in the league. Their attack is actually a touch below its chances (14 goals from 15.72 xG), which makes this the more sustainable kind of overperformance, built on control rather than finishing form.
Brighton & Hove Albion are the opposite case. Their 19 points against an expected 14.49 comes from both clinical finishing (24 goals from 17.04 xG) and a defence conceding far fewer than its numbers suggest (seven against an expected 13.59). Two hot streaks running at once is the kind of combination that tends to cool, making them the one I’d watch most closely for a correction.
Manchester City sit closer to their numbers than either, 21 points from an expected 17.44, with goals and chances roughly matching across 35 created and 25 conceded. City look as strong as the table says.
Chelsea’s 16 points lean heavily on finishing, 19 goals from 13.17 xG, while their defence is allowing plenty underneath, 29 big chances conceded, the joint highest in the league. Worth some caution despite a healthy points total.
At the bottom, Nottingham Forest remain the league’s clearest underachievers, six points from an expected 15. Their attack is the issue, six goals from 14.86 xG despite creating 26 big chances, the division’s second most. Defence is solid underneath (12 conceded against 10.36 expected), so this still reads as a finishing problem rather than a structural one.
Sunderland look the most genuinely concerning side in the league. Five points from an expected 12.65, wasteful in attack (seven goals from 13.47 xG) and leaking more than expected at the back too (17 conceded against 14.09 xGA, 27 big chances allowed). Neither end offers much reassurance.
Manchester United sit in similar territory, seven points from an expected 12.91, shaky in attack (10 goals from 14.35 xG) and defence (15 conceded against 12, 29 big chances allowed). Like Sunderland, this looks more structural than unlucky.
Fulham’s story is simpler, nine points from an expected 13.93, almost entirely down to missed chances in front of goal (nine goals from 14.98 xG), with their defence holding up close to expectation.
Aston Villa are worth a mention too, a modest 12 points but riding a defence conceding fewer goals than its chances allowed would suggest, 30 big chances against is the league’s highest, yet only 13 goals have gone in. That gap won’t hold forever.
The pattern repeats: Arsenal and City look built to last, Brighton and Chelsea look due to cool, and while Forest and Fulham’s problems look fixable through finishing alone, Sunderland and United look like sides with issues at both ends.
What the tables do
For every team in the Premier League, Championship, League One, and League Two, I've compared what they have actually achieved, points, goals scored, goals conceded, against what the underlying numbers say they should have achieved. Those numbers come from xG, xGA, and xPTS, measures built from the quality and location of every shot a team has taken or faced, rather than just what went in.
The difference between the two tells the real story. A team outperforming its expected numbers is usually riding good finishing or good luck, and both tend to fade. A team underperforming its expected numbers is usually better than its results suggest, and often gets overlooked by the market as a result.
How to read it
The dashboard has four parts, each answering a different question.
The scatter chart plots every team on two axes at once, attack performance against defence performance. Teams in the bottom left are doing the right things at both ends without the results to show for it yet. Teams in the top right are overperforming across the board and due a correction. It's the quickest way to scan a whole league in one look.
The bar chart lets you flip between points, goals scored, and goals conceded, ranking every team by how far they've strayed from expectation. Green means better than expected, red means worse.
The full breakdown table holds every number behind the charts, actual and expected points, goals, and goals against for each team, sortable by clicking any column.
The big chances table adds something the goals alone can't show, how many clear cut opportunities each team is creating and conceding, regardless of whether they're converting them. Read it alongside the goals vs xG numbers above it. A team creating plenty of big chances but still underperforming its xG is wasting good positions, a different problem entirely to a team that isn't creating anything to begin with.
Why it matters for betting
Bookmakers price games off recent form and public perception, and both lag behind the underlying numbers. A team on a poor run built on strong process is routinely overpriced to keep losing. A team riding a hot streak built on shaky play is routinely overpriced to keep winning. The teams sitting bottom left on that scatter chart, doing everything right without the points to prove it, are consistently where the best value sits before the rest of the market catches up.
I'll update this every week as the season moves on, so it's worth checking back as results come in.
The Betting Desk is run by Neil Potter. Neil built and refined all of the site’s data models, helping it grow into a trusted source of analysis across the Premier League, EFL and major tournaments. His models track key metrics, long term trends and areas where market prices move away from underlying performance.
GambleAware