Premier League Round 10 2026-11-08 · 14:00 UTC finished
Sunderland (home)
VS Stadium of Light, Sunderland
Chelsea (away)
Pre-match preview · AI Match Analysis

The prediction model provides the probabilities. AI explains the evidence behind them.

Sunderland Look to Halt Slide Against In-Form Chelsea

Bottom line

Sunderland sit bottom with 17 goals conceded and a manager under mounting pressure after consecutive defeats; Chelsea sit fourth with superior conversion rates and in-form attackers in Cole Palmer and Joao Pedro. Eight rest days and a 390km trip south are manageable. Chelsea hold a clear data advantage across the board.

Sunderland sit bottom with 0.2 points per game in their last five, coming off a Europa League defeat; Chelsea sit fourth with stable away form and a positive xG differential. The conversion gap is stark—Chelsea convert at 15.1% versus Sunderland's 6.2%. The key variable to watch is whether Sunderland's backline can hold under pressure at home.

What the stats show

Sunderland may see more of the ball at home but lack the efficiency to threaten; Chelsea are likely to dominate territory and pressure the penalty area, while Sunderland need better shot accuracy and rely on Chelsea's conversion falling below their season average.

Core data comparison

MetricSunderland ChelseaReading
League position20/20 4/20Sunderland are propping up the table, 16 places below Chelsea.
Last 5 matches points per game0.20 1.80Sunderland earned just one point from five matches; Chelsea average 1.8 points over the same stretch.
Expected goals per match (xG)1.47 1.49Both sides average roughly 1.47-1.49 expected goals per match.
Expected goals against per match (xGA)1.75 1.40Sunderland face 1.75 expected goals against per match, above Chelsea's 1.40.

Derived metrics

MetricSunderland ChelseaReading
Shot accuracy rate30.2% 42.1%Chelsea hit the target at 42.1% compared to Sunderland's 30.2%.
Conversion rate (goals per shot)6.2% 15.1%Chelsea score from 15.1% of their shots, more than double Sunderland's 6.2%.
Last 5 vs previous 5 xG change-0.15 -0.24Both teams see declining xG trends recently, with Chelsea's drop slightly larger but still at a high absolute level.

Sunderland review

  • Season baselineSunderland have taken just 5 points from 9 games (1-2-6) and sit 20th, averaging 0.78 goals scored and 1.89 conceded at home.
  • Possession and passingHome possession sits at 49.6% and Granit Xhaka averages 466 passes per game—solid control but limited attacking output.
  • Attacking approachSunderland create 2.6 big chances per match but convert at only 6.2%, indicating poor finishing rather than chance creation.
  • Defensive structureConceding 12.6 shots per match with 1.75 xGA, Sunderland's defensive structure breaks down under sustained pressure.
  • Duels and set piecesAerial and ground duel win rate is moderate at 52 per match, yet goals conceded suggest poor positioning in duels.
  • Pressing intensityPressing intensity of 11.9 is lower than Chelsea's 13.7, suggesting limited high-press engagement.
  • Risk factors and counter-evidenceFour consecutive league defeats and a Europa League loss to Dinamo Zagreb have manager Regis Le Bris under pressure, though Sunderland beat Chelsea in both previous meetings this season.

Chelsea review

  • Defensive structureChelsea average 4.3 saves per match with a +0.08 xG differential; their defence is solid overall but seven of 15 goals conceded came away from home.
  • Defensive line depthAllowing 14 shots and 5.9 on target per match indicates the defence can be stretched by high-tempo attacks.
  • Counter-attack and long ballsCole Palmer has 3 goals and 3 assists with 6 big chances created; Joao Pedro has 5 goals and 3 assists from 6 appearances.
  • Aerial prowessAerial and ground duels won at 51.6 per match, roughly matching Sunderland's output.
  • Zone-entry abilityChelsea enter the penalty area 28.2 times per match, well above Sunderland's 22.9.
  • Goalkeeper performanceEmiliano Martinez has made 25 saves from 6 appearances with a strong save percentage, though his fitness is in doubt for this match.
  • ConcernsChelsea's away record of 2-0-2 shows no clear home-away split; European fixtures and a seven-player injury list, including key starters, add uncertainty.

Generated 2026-11-06 08:46 (UTC)

Prediction
36% Sunderland (home)
21% Draw
44% Chelsea (away)
Expected goals 1.36 — 1.58 Over 2.5 goals: 56% Both teams to score: 59%

Model: blend/v3-w1460-r150. Probabilities are produced by our own model from public match data. They are not odds and not betting advice.

Why — recent form

"—" means the number is not available in our data; we never fill gaps with zeros. Recent form covers only matches with expected-goals data, which may be fewer than the team's actual recent fixtures.

Metric Sunderland Chelsea
Recent xG (attack, per-match avg) 1.491.38
Recent xGA (defence, per-match avg) 1.961.29
Sample size (matches with xG data) 55
Rest days since last match 38
Season pressing (PPDA, lower = more intense) 11.9 13.7
Matches in the last 7 days 31
Matches in the next 7 days 11
Travel to this match (km) 0390

What could go against this prediction

The model's top probability for this match is below 50% (backtest hit rate for this band is 42.7%%).
no head-to-head data available
Recent team-state reports (conflict, manager pressure or board unrest) add uncertainty; treat the probability as a baseline, not a verdict.

Injuries & suspensions

Latest pre-match snapshot from our data source; availability varies by league.

Sunderland

  • injured Daniel Ballard return: Early November 2026
  • injured Habib Diarra return: Late November 2026
  • injured Romaine Mundle return: Early January 2027
  • suspended Omar Alderete return: Early November 2026

Chelsea

  • injured Aaron Anselmino return: Mid November 2026
  • injured Jamie Gittens return: Back in training
  • injured João Pedro return: Late November 2026
  • injured Jorrel Hato return: Mid November 2026
  • injured Maxence Lacroix return: Early January 2027
  • injured Reece James return: Mid January 2027
How predictions are made

Probabilities are generated by our own statistical model from public match data. AI is used only to explain the evidence and analysis.

Data coverage

Model probabilities: available Recent xG form: available Injuries & suspensions: available Head to head: unavailable Pressing baseline: available Season context: available