PSG's attack makes an early statement
Paris Saint-Germain have started the 2026/27 Champions League campaign with three wins from three matches and 14 goals scored. Their attacking numbers are just as eye-catching: 10.13 expected goals (xG), or 3.38 per match. PSG have created a high volume of dangerous chances while also finishing more of them than the model estimates.
That does not mean they have the best attack in Europe. Three matches are too small a sample for a firm ranking, and opponents differ in strength. But the numbers give us a clear starting point for looking beyond results.
SmartMatchData has 68 completed Champions League matches in its 2026/27 dataset, including qualifying ties. Full-quality xG is available for 44 of them. The comparisons here focus on selected clubs with at least two qualifying xG matches, rather than treating every team as if it had the same data coverage.
Champions League xG: selected team comparison
PSG have averaged 3.38 xG in their three recorded matches. Bayern Munich follow in this selected group at 2.95 xG per game across two matches, while Arsenal have produced 2.88 across two. Manchester City have averaged 2.60 over three matches.
The defensive numbers tell a different story. Arsenal have allowed just 0.52 xGA across their two qualifying matches, or 0.26 per game. Bayern have conceded 1.02 xGA across two matches. PSG, despite their strong attack, have allowed 5.58 xGA in three games.
These are per-match averages, not full-tournament rankings. The sample sizes differ, and recent fixtures without complete xG have been excluded.
PSG have scored fourteen goals in three games
PSG opened with a 6-1 win over Slovan Bratislava, creating 4.57 xG and allowing 0.76 xGA. They then beat Manchester City 4-2 away, with 1.83 xG and 2.63 xGA. A further 4-2 victory against Barcelona produced 3.73 xG and 2.19 xGA.
The three games show both sides of PSG's early campaign. They have created chances consistently and scored heavily, but their opponents have also found opportunities. A 3.38 xG attacking average is impressive. The 1.86 xGA average suggests their defence has faced meaningful pressure as well.
PSG's actual goals are 3.87 above their total xG. That gap is worth following, but it should not be presented as proof of superior finishing over an entire season. Three matches cannot separate repeatable skill from short-term variation.
Arsenal's defensive xGA stands out
Arsenal's two matches with full-quality xG have produced a combined 5.75 xG and only 0.52 xGA. They won 1-0 away to Napoli and 3-0 at home against Lille. In the Napoli match, Arsenal created 4.05 xG while allowing 0.22. Against Lille, they created 1.70 and allowed 0.30.
Those numbers point to two strong defensive performances. They also show that Arsenal have created more chances than their four actual goals in the qualifying xG sample.
However, Arsenal's 2-2 draw away to Bayern on 21 October does not yet have full-quality xG in the database. Their current xG and xGA averages therefore cover two matches, while their record of two wins and one draw covers three. Mixing those samples would give readers the wrong impression.
Bayern create chances and restrict opponents
Bayern's two matches with complete xG data produced 5.90 xG and 1.02 xGA. They beat Bodø/Glimt 5-0 at home, creating 2.61 xG and allowing 0.30. A 3-3 draw away to Viking FK brought 3.29 xG and 0.72 xGA.
The Viking result is a useful reminder that xGA is not the same as goals conceded. Bayern allowed three actual goals despite a much lower xGA estimate. The gap may involve finishing, goalkeeping and the limits of the chance-quality model. The figures alone cannot explain the exact cause.
Bayern's third match, a 2-2 draw with Arsenal, is excluded from the xG sample because full-quality values are not yet available. That leaves only two matches for the underlying comparison.
Manchester City: good chance creation, mixed results
Manchester City have won two of their first three Champions League games and lost one. Their three matches produced 7.79 xG and 2.83 xGA, averages of 2.60 and 0.94 per game.
They beat Porto 2-0 away, lost 4-2 at home to PSG and beat AEK Athens 5-1. The defeat against PSG came despite City recording 2.63 xG to PSG's 1.83. That is not a contradiction: xG describes the estimated quality of chances, while the score records the goals actually scored.
City's early numbers show strong chance creation, but also why a single match should not be used to judge an entire campaign.
Why xG and xGA need context
Expected goals estimates the likelihood of shots becoming goals. Expected goals against (xGA) applies the same idea to opponents' chances. Used together, they can show whether a team is creating more dangerous opportunities than it allows.
But xG is not a substitute for the final score. It also does not account for every part of a football match. Red cards, game state, opponent strength and the timing of chances can change how a performance should be interpreted.
This early Champions League sample is especially uneven. Some clubs have two full-quality xG matches, others have three, and the database also contains qualifying-round games. The averages should be treated as useful observations rather than a definitive ranking of Europe's best teams.
For competition results and match archives, visit the Champions League page and its match analysis archive. For the basics, read Expected Goals Explained and xG vs xGA Explained.
Data and methodology
The analysis uses SmartMatchData's UEFA Champions League 2026/27 records through 21 October 2026. There are 68 finished matches in the dataset, including qualifying fixtures. Forty-four have full-quality xG for both teams. Team xG and xGA totals use only those full-quality observations. Per-match figures divide by the number of qualifying observations, not all matches played. Scores and results may include fixtures without xG. Figures are rounded to two decimals. This is an early-season descriptive study, not a forecast or a complete ranking of all participating clubs.