← KICK90 board
Premier LeagueT-7dlineups pending

Manchester City vs Sunderland

Sunday 20 Sep, 13:00 UTC · Etihad Stadium

Referee not appointedWeather pending
Most probable outcome
MANCHESTER CITY WIN

Outcome distribution

Calibrated 1X2 probabilities

72.7%
Manchester City
17.7%
Draw
9.6%
Sunderland
Home xG
2.77
Away xG
0.74
Likely score
2–0
11.5%
Completeness
76.3%
input coverage

Why the model says this

Ranked contributing factors

  • Manchester City attacking output+64

    2.23 xG and 2.23 goals per match over the last 28.

  • Sunderland defensive record+1

    1.57 xG and 1.57 goals conceded per match.

  • Manchester City home record+100

    38.7 scored / 8.7 conceded across 12 home matches.

  • Sunderland away record+24

    9.2 scored / 13.1 conceded across 8 away matches.

  • Squad availability+7

    Manchester City: Jérémy Doku  (out) · Sunderland: Alan Browne (out), Romaine Mundle (out), Aji Alese (out), Habib Diarra (out)

  • Squad depth-1

    Manchester City: 30 registered, 0 regular starters, fit cover 3GK/11DEF/7MID/6FWD, 0% of squad minutes unavailable · Sunderland: 35 registered, 0 regular starters, fit cover 3GK/10DEF/6MID/8FWD, 0% of squad minutes unavailable

  • Recent transfer activity-2

    Manchester City: 12 in (0 on loan), 20 out including 0 regulars — Claudio Echeverri to Benfica, Iliman Ndiaye from Everton, Jeremy Monga to Swansea City · Sunderland: 8 in (1 on loan), 26 out including 0 regulars — Kevin Danso from Tottenham Hotspur, Luís Semedo to Clermont, Juan Riquelme Angulo Caiced from Independiente del Valle

  • Manchester City last 5 played+88

    W W W W W — 2.23 scored / 0.48 conceded per match, 3.00 points per game (recency weighted).

  • Sunderland last 5 played-16

    W D W L W — 1.00 scored / 0.67 conceded per match, 1.87 points per game (recency weighted).

  • Model agreementneutral

    Elo 75% · form 77% · goal model 78% · results history 74% for the home win.

  • Expected attacking output+68

    Goal-expectancy model: Manchester City 2.77 xG, Sunderland 0.74 xG.

Most likely outcomes

Ranked by model probability · confidence is per outcome, from the raw sample behind it

Betting parameters

Full-time 1X2 and derived result combinations.

Probabilities are model estimates from the same score distribution as the 1X2 line. KICK90 publishes likelihoods, not tips.

Team comparison

Last-10 window

MetricManchester CitySunderland
Elo17601441
Record20W 6D 2L4W 5D 4L
Goals for / against62.50645297579875 / 18.99846686219074613.898721149945443 / 22.000668529777823
xG for / againstDATA UNAVAILABLEDATA UNAVAILABLE
Possession60%43%
Rest daysDATA UNAVAILABLEDATA UNAVAILABLE
Venue split12 matches, 38.68546574517253–8.7218904663598278 matches, 9.207258869119345–13.071517974410815

AI briefing

The model explains the numbers — it does not produce them

Generate a written analyst briefing that separates verified data from interpretation.

Confidence

Separation · completeness · model agreement

72High

Confidence reflects how strongly the evidence points one way — not how attractive the outcome is. It rises as lineups confirm and data completeness improves.

Probability evolution

How the answer moved toward kickoff

  • T-7dcompleteness 76.3%
    72.7% / 17.7% / 9.6%

Squad availability

Weighted by importance and output

Manchester City
  • Jérémy Doku  Attackerout
Sunderland
  • Alan Browne Midfielderout
  • Romaine Mundle Attackerout
  • Aji Alese Defenderout
  • Habib Diarra Midfielderout

Availability impact: 1.7% home attack, 6.1% away attack.