Premier League · 2025/26
Chelsea
Over 38 matches, Chelsea create most danger through the middle (1.08× the league norm), and concede most through the middle (1.05×).
58 scored, 14.2 below xG
52 conceded, 6.1 below xGA
Where the shots come from
Shots taken
every shot, or counted by zoneChelsea, 2025/26 · 509 shots in the final 35m, 57 scored. Shape carries the outcome rather than colour, so it reads in print and to anyone who cannot separate the hues.
vs league average · 1.00 marked · share of xG
Shots taken
every shot, or counted by zoneChelsea, 2025/26 · 509 shots in the final 35m, 57 scored. Shape carries the outcome rather than colour, so it reads in print and to anyone who cannot separate the hues.
vs league average · 1.00 marked · share of xG
Shots faced
every shot, or counted by zoneChelsea, 2025/26 · 411 shots in the final 35m, 51 scored. Shape carries the outcome rather than colour, so it reads in print and to anyone who cannot separate the hues.
vs league average · 1.00 marked · share of xG
Flanks and routes
Danger conceded by flank
left against right · open play onlyNo meaningful side bias. Chelsea concede about evenly on both flanks, so neither is worth targeting over the other.
Set pieces are excluded, since a corner has no flank. Sides within ten points of an even split are treated as level, since a season is short enough that smaller gaps are usually noise.
Crosses or through balls
open play onlyChelsea take 14% of their open-play danger from crosses and 15% from balls played through the middle.
Measured from the action immediately before each shot. This shows whether a side works the ball wide or through the middle; which flank they progress down cannot be derived, because no free source publishes pass locations.
Who does the work
Territory and pressing
1.00 = league averageChelsea move the ball about as the league does.
- Territory won1.06×
7.2 deep completions a match
- Territory allowed0.98×
6.6 conceded a match
- Pressing1.17×
10.6 opponent passes per defensive action
- Freedom to build1.39×
17.2 of their own passes allowed
- Danger per arrival1.14×
0.242 npxG per deep completion
A deep completion is a pass finished within about twenty yards of goal, and PPDA counts an opponent’s passes per defensive action. A lower figure is a harder press, so the bar inverts it to keep every row reading the same way. This describes where the ball ends up rather than how it travelled: no pass or carry record exists in this data, so progressive passes and carries are genuinely out of reach rather than merely absent.
Share of attacking output
non-penalty xG plus expected assistsChelsea’s attack is effectively 10.9 players. A side spread across fourteen absorbs an absence that one built on nine does not.
5.6% of their attack is doubtful or out. Moisés Caicedo, Trevoh Chalobah.
- João Pedro16.7%
- Enzo Fernández16.3%
- Pedro Neto11.1%
- Cole Palmer6.6%
- Estêvão5.7%
Share of the squad’s non-penalty xG plus expected assists, so it counts chances taken and chances given alike. Penalties are left out because they describe who the club nominates rather than who creates. Contributors under 2% are not listed, though they still count towards the spread above.
Share of defensive work
tackles, blocks, clearances, recoveriesChelsea’s defensive work is effectively 12.4 players. A side spread across fourteen absorbs an absence that one built on nine does not.
29.4% of their defensive work is doubtful or out. Moisés Caicedo, Trevoh Chalobah, Benoit Badiashile Mukinayi.
- Moisés Caicedo75% chance15.8%
- Trevoh ChalobahUnavailable11.5%
- Enzo Fernández10.6%
- Wesley Fofana8.9%
- João Pedro6.0%
- Cole Palmer5.8%
Share of the squad’s defensive contribution: tackles, blocks, clearances, interceptions and recoveries combined. Goalkeepers record none of it, so a keeper never appears here however much the side depends on him. This statistic begins in 2025/26, so there is no earlier season to compare against.
Creator to finisher
shots and xG per pairing- Pedro Neto→João Pedro2.4 xG10 shots · 1g
- Enzo Fernández→João Pedro2.3 xG13 shots · 3g
- Malo Gusto→João Pedro1.8 xG5 shots · 1g
- Enzo Fernández→Pedro Neto1.5 xG9 shots · 0g
- Enzo Fernández→Andrey Santos1.3 xG2 shots · 1g
- Marc Cucurella→João Pedro1.3 xG6 shots · 2g
- Reece James→Enzo Fernández1.2 xG4 shots · 0g
- Alejandro Garnacho→João Pedro1.2 xG3 shots · 1g
The pass immediately before the shot, so a player who starts moves without playing the final ball will not appear here. Roughly seven in ten shots name their creator.
How the season splits
Danger by scoreline
leading, level or behindChelsea produce most of their danger while ahead (37% of their xG).
- Ahead37%
- Level36%
- Behind26%
- Ahead
- 37%
- 27.5 xG · 138 shots
- Level
- 36%
- 26.8 xG · 224 shots
- Behind
- 26%
- 19.3 xG · 148 shots
The scoreline at the moment each shot was struck, rebuilt from goal times. Sides that trail often will look busier in attack than they really are.
Home and away
same team, two recordsChelsea travel well. Their home and away numbers are close.
Shots faced and shots stopped
two different jobsChelsea allow 1.57 xG a match, and concede 0.7 fewer goals than the league converts that into.
- Chance prevention
- 1.57
- Shot stopping
- +0.7
xG allowed per match. Mostly the outfield shape, and it tends to repeat.
Goals prevented against the league rate. Largely the keeper, and it swings year to year.
Measured against the league’s own conversion rate rather than raw xG. Blocked and off-target shots carry xG that can never be scored, so comparing goals to raw xG would mark every side in the division as leaky. A high block share means bodies in the way rather than chances denied.