Chelsea's recorded results for 2025/2026, newest first, with the available xG behind each score. Compare outcomes with the chances created and conceded, then open a match for more detail.
Results
- Wins – draws – losses (W–D–L)
- 1 win, 1 draw, 3 losses
4 pts 5–10 goals −5 GD
Chances
- Provider xG / xGA
- 4.54 – 7.86
- 0.91 – 1.57 per match
- xGD
- -3.32
- -0.66 per match
- xG record
- 0 with the chances, 2 with even chances, 3 against them
Comparable xG: 5 of 5 completed results.
Points vs xPts Model
4 pts 4.3 xPts
−0.3 vs xPts
Matched xPts: 5 of 5 selected results.
Model · How xPts works →Match points use 3 for a win, 1 for a draw and 0 for a loss, before deductions or sanctions.
Points vs xPts uses only matches with available xPts on both sides of the comparison.
What the results say
Last 5 · Premier League · All venues
Chelsea's poor return follows chances that have gone against them at both ends. They out-created the opponent in 1 of 5 matches, for −3.32 xG difference overall on 4.54–7.86. Their points total of 4 lands on the model's 4.3 xPts, so the results have followed the chances closely.
Points minus xPts, match by match
Over these 5 results Chelsea are 0.31 points below their xPts. The gap was widest at -1.87, against Liverpool on Sat 09/05.
- Points
- 4
- xPts
- 4.31
- Gap
- -0.31
- This match
- Match gap
- Running gap
| Date | Opponent | Points | xPts | Match gap | Running gap |
|---|---|---|---|---|---|
| Tue 21/04 | Brighton & Hove Albion | 0 | 0.22 | -0.22 | -0.22 |
| Mon 04/05 | Nottingham Forest | 0 | 1.45 | -1.45 | -1.67 |
| Sat 09/05 | Liverpool | 1 | 1.20 | -0.20 | -1.87 |
| Tue 19/05 | Tottenham Hotspur | 3 | 0.77 | +2.23 | +0.36 |
| Sun 24/05 | Sunderland | 0 | 0.67 | -0.67 | -0.31 |
Scores and xG
Chelsea first
-
Match analysis →
Why this label · Sunderland, Sun 24/05
The team lost and recorded less xG than its opponent. The direction of the result aligns with the chance difference.
An xGD of at least +0.25 is positive; at most −0.25 is negative; between them is approximately even. The band uses unrounded figures and is an editorial guide, not a statistical confidence threshold. The chip on each result reads one of three ways: backed by the chances, with even chances, or against them. Counted over the selection, those three are the xG record on the Chances card — 5–1–0 reads as five results with the chances, one even and none against them.
- Points
- 0
- Post-match model xPts
- 0.66
- Points minus xPts
- -0.66
xPts is our post-match model's 3 × win probability + draw probability using the match's provider xG, not a pre-match forecast.
Learn how xG and xPts are calculated → -
Match analysis →
Why this label · Tottenham Hotspur, Tue 19/05
The team won despite recording less xG than its opponent. That contrast does not make the win impossible or establish why it happened.
An xGD of at least +0.25 is positive; at most −0.25 is negative; between them is approximately even. The band uses unrounded figures and is an editorial guide, not a statistical confidence threshold. The chip on each result reads one of three ways: backed by the chances, with even chances, or against them. Counted over the selection, those three are the xG record on the Chances card — 5–1–0 reads as five results with the chances, one even and none against them.
- Points
- 3
- Post-match model xPts
- 0.77
- Points minus xPts
- +2.23
xPts is our post-match model's 3 × win probability + draw probability using the match's provider xG, not a pre-match forecast.
Learn how xG and xPts are calculated → -
Match analysis →
Why this label · Liverpool, Sat 09/05
The match was drawn and the chance difference was approximately even. The two readings broadly align.
An xGD of at least +0.25 is positive; at most −0.25 is negative; between them is approximately even. The band uses unrounded figures and is an editorial guide, not a statistical confidence threshold. The chip on each result reads one of three ways: backed by the chances, with even chances, or against them. Counted over the selection, those three are the xG record on the Chances card — 5–1–0 reads as five results with the chances, one even and none against them.
- Points
- 1
- Post-match model xPts
- 1.20
- Points minus xPts
- -0.20
xPts is our post-match model's 3 × win probability + draw probability using the match's provider xG, not a pre-match forecast.
Learn how xG and xPts are calculated → -
Match analysis →
Why this label · Nottingham Forest, Mon 04/05
The team lost, while the chance difference was approximately even. The outcome was weaker than that difference alone suggests.
An xGD of at least +0.25 is positive; at most −0.25 is negative; between them is approximately even. The band uses unrounded figures and is an editorial guide, not a statistical confidence threshold. The chip on each result reads one of three ways: backed by the chances, with even chances, or against them. Counted over the selection, those three are the xG record on the Chances card — 5–1–0 reads as five results with the chances, one even and none against them.
- Points
- 0
- Post-match model xPts
- 1.45
- Points minus xPts
- -1.45
xPts is our post-match model's 3 × win probability + draw probability using the match's provider xG, not a pre-match forecast.
Learn how xG and xPts are calculated → -
Match analysis →
Why this label · Brighton & Hove Albion, Tue 21/04
The team lost and recorded less xG than its opponent. The direction of the result aligns with the chance difference.
An xGD of at least +0.25 is positive; at most −0.25 is negative; between them is approximately even. The band uses unrounded figures and is an editorial guide, not a statistical confidence threshold. The chip on each result reads one of three ways: backed by the chances, with even chances, or against them. Counted over the selection, those three are the xG record on the Chances card — 5–1–0 reads as five results with the chances, one even and none against them.
- Points
- 0
- Post-match model xPts
- 0.22
- Points minus xPts
- -0.22
xPts is our post-match model's 3 × win probability + draw probability using the match's provider xG, not a pre-match forecast.
Learn how xG and xPts are calculated →
Biggest overperformance
Tottenham Hotspur 2:1
· Home xG / xGA 0.63 – 1.27
- Goal difference
- +1
- xGD
- -0.64
- GD minus xGD
- +1.64
Among 5 comparable results.
Open match →
Biggest underperformance
Nottingham Forest 1:3
· Home xG / xGA 2.15 – 2.05
- Goal difference
- -2
- xGD
- +0.10
- GD minus xGD
- -2.10
Among 5 comparable results.
Open match →
Reading these results
xG is the provider's estimate of chances created; xGA is the opponent's xG. xGD is the difference. Labels compare the recorded outcome with those chances; they are interpretations, not judgments about fairness.
xPts is our post-match model's 3 × win probability + draw probability using the match's provider xG, not a pre-match forecast.
How each figure on this page is read
Every xG bar is drawn on one fixed scale: full width is 7.20 xG, the same on every card and in every window, so two cards can be compared by eye. A rarer match above that is capped with a hatched end, and the figure beside the bar still gives its value.
An xGD of at least +0.25 is positive; at most −0.25 is negative; between them is approximately even. The band uses unrounded figures and is an editorial guide, not a statistical confidence threshold. The chip on each result reads one of three ways: backed by the chances, with even chances, or against them. Counted over the selection, those three are the xG record on the Chances card — 5–1–0 reads as five results with the chances, one even and none against them.
Goal difference minus xGD says how far a scoreline finished from the chances behind it: a 4:2 won off level chances ran nearly two goals ahead of them. The two callouts under the list name the match that ran furthest each way, and a direction with no result a full goal clear of the chances gets no card rather than a card built from a rounding.
Windows count back from the most recent recorded match, and only matches attached to this club record are included. Home and away mean the provider-designated side, not a verified physical venue. Scores describe recorded play, not shootout advancement. Unverified periods receive no chance comparison.