Each line is generated from a finding that cleared its evidence bar; nothing here is a generic coaching point. The grey line says what it rests on; the button opens the sequences.
In your own build-up area, play on one or two touches: don't try to dribble past their press, and keep first touches clean.
37% of their high regains come straight after the opponent tries to dribble past someone or takes a bad touch (league median 29%); only 26% come from forward passes.
Copy the sides that escaped it: Everton got out 64% (long on 36%).
Average escape rate against them is 40%.
When you win it, go straight through the middle immediately.
43% of the xG they concede after losing the ball comes straight through the middle (22 attacks, 8 shots), against 19% of their own attacking xG through that channel.
Avoid conceding corners; mark Jorrel Hato first.
20% of their corners end in a shot; first contacts: Jorrel Hato (2), João Pedro (2), Pedro Neto (1).
Where each player takes the ball in each phase of their possession, and who passes to whom. Event data has no off-ball positions, so this is the shape of their touches — the standard event-data proxy. A phase is defined by where the ball is, so the dots sit inside that phase's third (shaded): read the order and the width, not the depth of each line. Chelsea attack left to right.
attacking → · shaded = the build-up zone
428 build-up phases. Dot = the median spot of a player's touches in this phase; size = share of touches; lines = consecutive touches between two players.
attacking → · shaded = the progression zone
604 progression phases. Dot = the median spot of a player's touches in this phase; size = share of touches; lines = consecutive touches between two players.
attacking → · shaded = the final third zone
397 final third phases. Dot = the median spot of a player's touches in this phase; size = share of touches; lines = consecutive touches between two players.
Tackles, interceptions, recoveries, blocks, challenges and fouls in the opponent's final 40%.
159 in the window. The press on a build-up = the first two rows: 32 (4.0 a match).
| source | times | share | per match | shot | |
|---|---|---|---|---|---|
| press: settled build-up taken off them | 31 | 19% | 3.9 | 13% | |
| press: their restart forced into an error before they settled | 1 | 1% | 0.1 | 0% | |
| counter-press: won straight back after losing it | 8 | 5% | 1.0 | 0% | |
| second ball: the opponent touched it but never controlled it | 119 | 75% | 14.9 | 32% | |
| ↳ rebound after their own shot (saved or blocked) | 20 | 13% | 2.5 | 45% | |
| ↳ their cross / ball into the box, cleared | 49 | 31% | 6.1 | 29% | |
| ↳ their long ball, only headed or touched away | 13 | 8% | 1.6 | 23% | |
| ↳ after their own throw-in / set piece | 6 | 4% | 0.8 | 17% | |
| ↳ loose ball in their combination play (pass blocked or tackled) | 31 | 19% | 3.9 | 35% |
| opponent was | share | league | rank | |
|---|---|---|---|---|
| dribbling or a loose touch | 37% | 29% | 3 / 20 | |
| passing forward | 26% | 25% | 9 / 20 | |
| passing sideways | 11% | 6% | 2 / 20 | |
| passing to a full-back | 11% | 14% | 14 / 20 | |
| going long | 6% | 7% | 14 / 20 | |
| passing back | 3% | 3% | 8 / 20 | |
| passing to the keeper | 3% | 0% | 6 / 20 | |
| taking a throw-in | 3% | 3% | 11 / 20 | |
| a loose ball | 0% | 0% | 1 / 20 |
The opponents who went long most (38% of build-ups) got out 39% of the time; those who mostly played short (23% long) got out 43%. That is a lean, not proven at 20 matches (p=0.7584).
Best escapes: Everton 64%; Brighton 54%; Brentford 44%. One dot per match; hover for the opponent.
Most often started by Cole Palmer (11), Maxence Lacroix (9), Jordan Henderson (7), Roméo Lavia (6).
| lost in | losses | won back within 5s | led to a shot |
|---|---|---|---|
| own third | 31 | 35% | 13% |
| midfield | 160 | 21% | 6% |
| final third | 242 | 25% | 1% |
13.2 opponent attacks per match after winning the ball in open play that broke forward fast or reached the final third. Shading and % = share of the xG they produced, by the channel they used. Chelsea's goal is on the left. Central attacks carry more xG for every team, so compare with Chelsea's own attacking xG in the same kind of channel (centre, or the average flank): centre 43% conceded vs 19% own, left 39% conceded vs 40% own, right 18% conceded vs 40% own.
Every real turnover, named by the situation it came from. "Fast" = the side with the ball reached the final third within 10 seconds. Only real turnovers appear here: for everything the press wins high, including restarts it forced into an error, see "Every ball won in the opponent’s third" in the press section.
| type | times | share | fast | shot | xG | |
|---|---|---|---|---|---|---|
| Won high from a settled build-up (press) | 34 | 6%lg 4% | 35% | 9%lg 11% | 0.2 | |
| Won after their long possession (10+ passes) | 71 | 12%lg 9% | 14% | 3%lg 3% | 0.2 | |
| Won straight back after losing it (counter-press) | 64 | 11%lg 13% | 25% | 6%lg 7% | 0.5 | |
| Won by stopping their counter | 12 | 2%lg 4% | 0% | 0%lg 0% | 0.0 | |
| Won in midfield | 127 | 22%lg 25% | 34% | 9%lg 6% | 1.2 | |
| Won deep in own third | 244 | 42%lg 38% | 11% | 3%lg 3% | 0.9 | |
| Won from their set piece | 31 | 5%lg 5% | 6% | 3%lg 3% | 0.1 |
| type | times | share | opponent fast | shot against | xG | |
|---|---|---|---|---|---|---|
| Lost in own build-up under pressure | 23 | 5%lg 4% | 22% | 4%lg 13% | 0.0 | |
| Lost after own long possession (10+ passes) | 58 | 12%lg 9% | 12% | 7%lg 3% | 0.5 | |
| Lost straight after winning it | 52 | 10%lg 13% | 38% | 6%lg 8% | 0.3 | |
| Lost during own counter | 16 | 3%lg 4% | 19% | 6%lg 0% | 0.1 | |
| Lost in midfield | 139 | 28%lg 25% | 29% | 8%lg 6% | 0.9 | |
| Lost in the final third (attack broke down) | 185 | 37%lg 39% | 10% | 3%lg 3% | 0.7 | |
| Lost from own set piece | 24 | 5%lg 5% | 17% | 12%lg 2% | 0.2 |
Every controlled possession is typed by how it was played, and every real turnover (a possession that was established and then lost) by what the losing side was doing, how it lost the ball and what the winner did next. "Settled" = sustained or worked possessions; "unsettled" = short, second-ball, counter and direct possessions.
| type | times | share | ended in a shot | avg passes | |
|---|---|---|---|---|---|
| sustained circulation (10+ passes) | 81 | 13%lg 11% | 10%lg 16% | 14.5 | |
| worked possession (4-9 passes) | 164 | 26%lg 26% | 12%lg 10% | 5.8 | |
| direct attack | 85 | 13%lg 14% | 5%lg 2% | 1.5 | |
| short possession (0-3 passes) | 126 | 20%lg 22% | 3% | 1.4 |
| the side losing it was in… | times | fast attack | secured | lost straight back | can’t keep it | shot | xG | how it was lost | |
|---|---|---|---|---|---|---|---|---|---|
| sustained circulation (10+ passes) | 58 | 12% | 12% | 10% | 45% | 7% | 0.5 | pass cut out 37, bad touch 9 | |
| worked possession (4-9 passes) | 132 | 17% | 9% | 12% | 54% | 6% | 0.6 | pass cut out 67, long ball lost 27 | |
| direct attack | 63 | 25% | 17% | 8% | 38% | 3% | 0.1 | long ball lost 26, pass cut out 25 | |
| short possession (0-3 passes) | 97 | 29% | 12% | 5% | 48% | 6% | 0.6 | pass cut out 38, bad touch 22 | |
| second-ball possession | 100 | 16% | 19% | 5% | 52% | 5% | 0.5 | pass cut out 49, long ball lost 22 | |
| counter / fast transition | 23 | 22% | 4% | 4% | 48% | 4% | 0.1 | pass cut out 12, bad touch 7 | |
| set piece | 24 | 17% | 21% | 12% | 50% | 12% | 0.2 | pass cut out 15, long ball lost 6 |
| mechanism | times | share | led to a shot against | |
|---|---|---|---|---|
| pass cut out | 243 | 49%lg 52% | 5% | |
| long ball lost | 111 | 22%lg 24% | 7% | |
| bad touch | 81 | 16%lg 14% | 11% | |
| tackled / dispossessed | 30 | 6%lg 6% | 0% | |
| failed dribble | 30 | 6%lg 6% | 3% |
| the opponent was in… | times | fast attack | secured | lost straight back | can’t keep it | shot | xG | how it was lost | |
|---|---|---|---|---|---|---|---|---|---|
| sustained circulation (10+ passes) | 73 | 15% | 14% | 10% | 53% | 4% | 0.3 | pass cut out 40, long ball lost 19 | |
| worked possession (4-9 passes) | 168 | 20% | 12% | 6% | 55% | 4% | 0.9 | pass cut out 90, long ball lost 45 | |
| direct attack | 55 | 27% | 24% | 7% | 40% | 9% | 0.5 | long ball lost 24, pass cut out 17 | |
| short possession (0-3 passes) | 100 | 25% | 12% | 2% | 57% | 8% | 0.9 | pass cut out 45, long ball lost 25 | |
| second-ball possession | 136 | 18% | 10% | 7% | 60% | 4% | 0.4 | pass cut out 78, long ball lost 20 | |
| counter / fast transition | 20 | 5% | 20% | 10% | 60% | 0% | 0.0 | pass cut out 15, bad touch 3 | |
| set piece | 31 | 6% | 16% | 10% | 61% | 3% | 0.1 | pass cut out 16, long ball lost 9 |
Fast attack = their final third or box reached within 10s. Secured = kept 10s+ or 4+ passes. Lost straight back = controlled, then lost within 5s. Can’t keep it = won the ball but never got it under control (cleared it, headed it away, blocked it or lost it at once).
Each phase split into the situations it is made of, every row with its league median and clips. Shots are their own category: a possession that ends in a shot has finished, it has not lost the ball.
| type | times | share | got out | reached the box | lost in build-up | shot | |
|---|---|---|---|---|---|---|---|
| from the keeper / goal kick, played out short | 23 | 16%lg 17% | 65% | 0% | 43% | 0% | |
| from the keeper / goal kick, went long (bypass) | 21 | 14%lg 13% | 33% | 19% | 57% | 14% | |
| from the keeper / goal kick, stuck / recycled | 11 | 8% | 0% | 0% | 55% | 0% | |
| from the back line, played out short | 9 | 6%lg 8% | 78% | 11% | 11% | 11% | |
| from the back line, went long (bypass) | 11 | 8%lg 5% | 27% | 27% | 73% | 18% | |
| from the back line, stuck / recycled | 6 | 4% | 0% | 17% | 50% | 0% | |
| straight after winning it, played out short | 26 | 18%lg 16% | 77% | 38% | 19% | 4% | |
| straight after winning it, went long (bypass) | 17 | 12%lg 11% | 18% | 18% | 76% | 6% | |
| straight after winning it, stuck / recycled | 14 | 10% | 0% | 7% | 7% | 7% |
| outcome | times | share | opponent shot | |
|---|---|---|---|---|
| won it high | 18 | 10%lg 10% | 0% | |
| forced long, then won it back | 46 | 25%lg 28% | 0% | |
| forced long, they kept it | 8 | 4%lg 5% | 25% | |
| they played through short | 54 | 30%lg 30% | 9% | |
| kept them stuck (recycled back) | 55 | 30%lg 24% | 0% |
Won high = the ball won in the opponent's own third. Forced long = the opponent left its third with a long ball (then either lost it in that possession, or kept it).
| created by | shots | share of open play | xG | xG per shot | goals | |
|---|---|---|---|---|---|---|
| cross | 2 | 3%lg 12% | 0.3 | 0.15lg 0.11 | 1 | |
| cutback | 10 | 13%lg 9% | 1.4 | 0.14lg 0.11 | 2 | |
| through ball | 1 | 1%lg 3% | 0.1 | 0.13lg 0.17 | 0 | |
| short pass / combination | 46 | 60%lg 52% | 3.5 | 0.08lg 0.08 | 7 | |
| long ball | 5 | 6%lg 4% | 0.5 | 0.09lg 0.07 | 0 | |
| individual (dribble / self-made) | 6 | 8%lg 9% | 0.6 | 0.09lg 0.08 | 1 | |
| rebound / second ball | 7 | 9%lg 9% | 0.5 | 0.07lg 0.13 | 0 | |
| corner (set piece) | 14 | – | 1.6 | 0.11 | 1 | |
| free kick (indirect) (set piece) | 6 | – | 0.3 | 0.05 | 2 | |
| direct free kick (set piece) | 7 | – | 0.2 | 0.02 | 1 | |
| throw-in (set piece) | 3 | – | 0.4 | 0.14 | 2 |
| created by | shots | share | xG | xG per shot | goals | |
|---|---|---|---|---|---|---|
| cross | 8 | 11%lg 12% | 0.9 | 0.12 | 1 | |
| cutback | 3 | 4%lg 9% | 0.4 | 0.14 | 0 | |
| through ball | 3 | 4%lg 3% | 0.4 | 0.15 | 0 | |
| short pass / combination | 32 | 44%lg 54% | 2.4 | 0.08 | 3 | |
| long ball | 6 | 8%lg 4% | 0.6 | 0.09 | 0 | |
| individual (dribble / self-made) | 9 | 12%lg 10% | 0.5 | 0.06 | 1 | |
| rebound / second ball | 12 | 16%lg 9% | 1.5 | 0.13 | 2 | |
| corner (set piece) | 12 | – | 2.4 | 0.20 | 3 | |
| free kick (indirect) (set piece) | 6 | – | 0.6 | 0.11 | 0 | |
| direct free kick (set piece) | 2 | – | 0.0 | 0.02 | 0 | |
| throw-in (set piece) | 8 | – | 0.7 | 0.09 | 3 |
| ending | shots | share | |
|---|---|---|---|
| goal | 17 | 16%lg 9% | |
| keeper held it | 10 | 9%lg 9% | |
| keeper parried it | 14 | 13%lg 11% | |
| blocked | 34 | 32%lg 32% | |
| off target | 30 | 28%lg 39% | |
| hit the woodwork | 2 | 2% |
Penalties excluded. A "save" by an outfield player is a block.
After their keeper held a shot (13), median release 10s; league share turned into a fast attack 14%:
| their keeper held it, then… | times | share | |
|---|---|---|---|
| keeper released fast -> fast attack | 0 | 0% | |
| keeper released quickly, no fast attack | 3 | 23% | |
| keeper slowed the game | 10 | 77% |
After their own shot was held by the opponent keeper (10); league share broken against fast 14%:
| the opponent keeper held it, then… | times | share | |
|---|---|---|---|
| keeper released fast -> fast attack | 3 | 30% | |
| keeper released quickly, no fast attack | 3 | 30% | |
| keeper slowed the game | 4 | 40% |
Fast = the keeper's side reached its final third within 15s of the save. Quick = first pass within 6s.
The other half of each phase above. Their own phases break down when they lose the ball (real turnovers only); their press and block break down when the opponent gets through. Each card: how often, how, where, who is involved, and what it costs them.
123 balls lost in 428 build-up phases (29%)
giveaways per 10 build-up phases: 2.5/10 · league 2.7/10
What it costs: 28% turned into a fast attack · 50% the opponent can’t keep · 7% led to a shot · 0.5 xG
178 balls lost in 604 middle-third phases (29%)
giveaways per 10 middle-third phases: 2.5/10 · league 2.8/10
What it costs: 24% turned into a fast attack · 39% the opponent can’t keep · 8% led to a shot · 1.3 xG
171 balls lost in 397 final-third phases (43%)
deliveries into the box that fail to arrive: 55% · league 60%
What it costs: 9% turned into a fast attack · 60% the opponent can’t keep · 3% led to a shot · 0.5 xG
50 deliveries into the box did not arrive
23 of 40 counter-attacks lost (57%)
attacking transitions that produce a shot: 18% · league 18%
What it costs: 22% turned into a fast attack · 48% the opponent can’t keep · 4% led to a shot · 0.1 xG
40 counters: 11 ended in a shot, 23 were lost, 6 slowed down or went out
66 of 163 opponent build-ups got out (40%)
opponent build-ups that get out cleanly: 40% · league 46%
What it costs: 24% reached the final third · 8% led to a shot · 0.3 xG
Sides are their own: "down the left" = down their left.
118 box entries in 463 opponent final-third phases (25%)
opponent final-third phases that create: 44% · league 51%
What it costs: 43% led to a shot · 6.6 xG
Sides are their own.
24 set pieces lost in open play
What it costs: 17% turned into a fast attack · 50% the opponent can’t keep · 12% led to a shot · 0.2 xG
Their corners and free kicks that were lost in open play, and what the opponent did next.
Takers: Jordan Henderson (15), Pedro Neto (13), Reece James (10)
First contact in the box: Jorrel Hato (2), João Pedro (2), Pedro Neto (1), Cole Palmer (1), Valentín Barco (1)
What each player does in the phases, counted over the window. Only lines that clear a minimum count are shown.
The comparison behind the report: each behaviour as a rate per opportunity, ranked among the 20 Premier League clubs' full seasons, with how many of the 8 matches sat on the same side of the league median. "Style" is never a verdict.
| behaviour | Chelsea | league median | rank | matches | ||
|---|---|---|---|---|---|---|
| style | turnovers won by stopping a counter | 2% | 4% | 20 / 20 | 6/8 | |
| style | final-third entries down the right | 52% | 42% | 1 / 20 | 5/8 | |
| style | final-third entries down the left | 35% | 45% | 20 / 20 | 6/8 | |
| strength | opponent final-third phases that created | 44% | 51% | 19 / 20 | 6/8 | |
| strength | shots off target | 28% | 39% | 20 / 20 | 8/8 | |
| weakness | balls they won but can't keep (cleared, blocked, lost loose) | 55% | 49% | 1 / 20 | 6/8 | |
| strength | shots scored | 16% | 9% | 2 / 20 | 5/8 | |
| style | run in behind per 100 final third phases | 4.3/100 | 6.3/100 | 20 / 20 | 7/8 | |
| style | build-up exits down the right | 29% | 39% | 20 / 20 | 7/8 | |
| style | open-play shots conceded from rebounds / second balls | 16% | 9% | 1 / 20 | 6/8 | |
| style | open-play shots created by crosses | 3% | 12% | 20 / 20 | 7/8 |
| phase | route | times | shot | their average |
|---|
| split | behaviour | a | b | p | |
|---|---|---|---|---|---|
| against the more aggressive pressers vs the deeper / passive sides | build-up exits played by the goalkeeper | 48% | 24% | 0.0135 | |
| against the more aggressive pressers vs the deeper / passive sides | final-third entries by carrying it in | 34% | 17% | 0.0283 | |
| when leading vs when level | build-up giveaways that led to an opponent shot within 15s | 0% | 8% | 0.0318 | |
| when trailing vs when level | chance-creation phases that produced a shot | 13% | 39% | 0.001 | |
| when trailing vs when level | final-third phases that created (box entry or key pass) | 62% | 44% | 0.0053 | |
| when trailing vs when level | middle-third phases that progressed into the final third | 88% | 66% | 0.0134 |
| date | opponent | result | npxG | opponent press | what happened |
|---|---|---|---|---|---|
| 2026-08-24 | at Fulham | W 3-2 | 1.9–1.2 | pressed aggressively | Won 3-2, chances 1.85 v 1.23 xG. Fulham pressed harder than a typical Premier League side does (strict PPDA 11.5, league median 14.4). Biggest departure from their norm: counter-attacks 6.0/match, far above their usual 4.5/match. |
| 2026-08-30 | vs Brighton | W 4-3 | 2.6–1.0 | pressed aggressively | Won 4-3, chances 2.63 v 1.04 xG. Brighton pressed harder than a typical Premier League side does (strict PPDA 8.1, league median 14.4). Biggest departure from their norm: PPDA, strict (opponent passes in their own 60% per committed ball-winning action) 31.7, far above their usual 18.1. |
| 2026-09-06 | at Arsenal | L 1-2 | 0.6–1.9 | pressed aggressively | Lost 1-2 and the chances agreed (0.59 v 1.94 xG). Arsenal pressed harder than a typical Premier League side does (strict PPDA 11.9, league median 14.4) and went long on 55% of their own build-ups. Biggest departure from their norm: ball given away during build-up 8.0/match, far below their usual 13.4/match. |
| 2026-09-12 | vs Hull City | D 2-2 | 0.8–1.5 | mid-intensity press | Drew 2-2 (0.78 v 1.51 xG). Hull City pressed at roughly league-average intensity (strict PPDA 17.7, league median 14.4) and went long on 75% of their own build-ups. Biggest departure from their norm: opponent build-ups that got out cleanly 14%, far below their usual 41%. |
| 2026-09-18 | at Brentford | L 0-3 | 0.8–1.9 | sat deep / passive | Lost 0-3 and the chances agreed (0.80 v 1.94 xG). Brentford defended deep and let them have the ball (strict PPDA 20.7, league median 14.4). Biggest departure from their norm: build-up exits down the left 83%, far above their usual 48%. |
| 2026-10-10 | vs Bournemouth | W 5-1 | 0.8–1.7 | mid-intensity press | Won 5-1 on fewer chances (0.77 v 1.69 xG). Bournemouth pressed at roughly league-average intensity (strict PPDA 16.4, league median 14.4). Biggest departure from their norm: giveaways regained within 5s (counter-press) 12%, far below their usual 24%. |
| 2026-10-17 | at Everton | W 2-0 | 1.4–0.5 | sat deep / passive | Won 2-0, chances 1.39 v 0.49 xG. Everton defended deep and let them have the ball (strict PPDA 20.9, league median 14.4). Biggest departure from their norm: opponent build-ups that got out cleanly 64%, far above their usual 41%. |
| 2026-10-24 | vs Tottenham | L 0-2 | 0.5–0.7 | pressed aggressively | Lost 0-2 and the chances agreed (0.51 v 0.74 xG). Tottenham pressed harder than a typical Premier League side does (strict PPDA 11.4, league median 14.4). Biggest departure from their norm: box entries with a cross 0%, far below their usual 7%. |