Two very different ways to stand out

Manchester City finished the 2025/26 Premier League season with the highest expected goals total, but Arsenal produced the strongest balance between chances created and chances conceded. That distinction is the most interesting finding in SmartMatchData's full-season xG figures. City led the attacking measure with 70.89 xG. Arsenal's 65.09 xG was lower, yet its defence allowed just 28.26 xGA, giving Arsenal a league-leading xG difference of +36.83.

These numbers come from 380 completed league matches involving all 20 clubs, with 38 matches per team. Every match in this dataset has both home and away xG values marked as full quality. The analysis describes shot-quality estimates, not an alternative league table or a claim about who should have won the title.

Manchester City led the xG attack rankings

City's 70.89 xG was the highest team total in the league. It scored 77 actual goals, 6.11 more than its xG figure. Chelsea ranked second for xG with 66.66 but scored 58 goals, a gap of 8.66 below its chance-quality total. Arsenal followed with 65.09 xG and 71 goals. Manchester United recorded 64.44 xG and 69 goals, while Bournemouth rounded out the top five with 62.17 xG and 58 goals.

Chelsea's gap is a reminder that chance creation and finishing are separate parts of the story. The xG total says the team generated opportunities that a statistical model valued at roughly 66.66 expected goals across the season. It does not establish why 58 goals were scored. Finishing execution, goalkeeping, shot context and ordinary variation can all contribute, and this dataset alone cannot isolate those causes.

Arsenal's defence made the difference

Arsenal conceded chances worth only 28.26 xGA across 38 league games, the lowest total in the dataset. Its 27 actual goals conceded were close to that estimate. Manchester City had the second-lowest xGA at 43.95 and conceded 35 goals. Liverpool followed on 47.70 xGA, then Manchester United on 48.07 and Brighton & Hove Albion on 49.30.

The distance between Arsenal and City in xGA was 15.69 expected goals over the season. That gap was larger than the 5.80 xG advantage City held over Arsenal in attack. In other words, City's edge in chance creation did not outweigh Arsenal's advantage in limiting the quality of opponents' opportunities.

Expected goals difference tells the fuller story

xG difference is simply a team's xG minus its xGA. Arsenal led at +36.83, ahead of Manchester City (+26.94), Manchester United (+16.37), Chelsea (+14.31) and Liverpool (+12.43). It is a useful way to compare the balance of attacking and defensive chance quality over a common number of matches.

The metric is not a points projection, a prediction of the next result or proof that a team deserved a particular finishing position. It also does not capture every defensive action, tactical decision or game-state effect. Still, when every team has played 38 matches and the xG coverage is complete, the comparison is more informative than contrasting totals from unequal samples.

What the league-wide numbers show

Across the 380 matches, home teams scored 580 goals and generated 582.28 xG. Away teams scored 465 and generated 473.63 xG. Those totals show a home-side advantage in both actual goals and estimated chance quality across this particular season. They do not, by themselves, identify the causes of home advantage or tell us what will happen in any single future fixture.

For context on venue splits and sample size, read Home vs Away Football Statistics. Our xG vs xGA guide explains the difference between the two measures and how to interpret xG difference.

For a closer look at the team behind the strongest defensive xGA, read our Arsenal 2025/26 analysis. To see how the league-wide scoring picture changed from the previous season, explore the 2024/25 vs 2025/26 goals and xG comparison.

Methodology and data coverage

This review uses SmartMatchData's stored 2025/26 Premier League match records, with dates running from August 15, 2025 to May 24, 2026. The sample contains 380 completed regular league matches, no administrative results, no missing final scores and no missing home or away xG values. Every club appears in 38 matches. Team xG adds its home xG when playing at home and away xG when playing away; xGA adds the opponent's corresponding xG. Totals are rounded to two decimal places for display.

Expected goals estimates depend on the underlying provider and modelling definitions, so they should not be treated as universally interchangeable with another source's numbers. They measure estimated shot quality rather than guaranteed goals. Explore the current Premier League page for league information, or start with Expected Goals Explained for the statistical foundations.