vs
Manchester United vs Manchester City
Sun 13 Sept · 16:30Old Trafford, Manchester
Powered by statistical analytics hubProBettingHub · PRO Analytics hub for everyoneManchester UnitedHomeManchester CityAwayHome side first and left in every table.
In short
In short
- The call
- Manchester City by a goal. The 0.67 goals they have conceded per league game is the number United's front line has to beat.
- The one stat
- United have conceded in all three league matches, six goals in three. City have conceded two in three.
- The risk
- January's meeting here finished 2-0 to United, seven shots on target to one. This fixture ignores tables more often than most.
- The calls
- Safer: draw or Manchester City at 1.38. Riskier: Manchester City -1 at 3.67. Five more leans on the board.
The situation
Any Manchester United vs Manchester City prediction this week starts with five points and ten places. City are first with 9 points from three played (W3 D0 L0), a goal difference of plus five and two goals conceded. United are eleventh with four points from three (W1 D1 L1), a goal difference of plus one and six conceded. Both sides have scored seven.
Two new eras meet. Enzo Maresca took City in late June after Pep Guardiola resigned, and has won every league match so far. Michael Carrick, permanent at United, has a side that scores freely and defends nothing: two or more conceded in each of the first three league games. United went 5-2 past Ipswich at home, then drew 2-2 at Everton after a goalless first half.
The last meeting cuts against all of it. United beat City 2-0 at Old Trafford in January 2026, seven shots on target to one from 32% possession. Across the last eight competitive meetings the sides are level at three wins each with two draws, 2.50 goals a meeting.
The markets
Double chance
Draw or Manchester City
City have 9 points from three league matches and concede 0.67 goals per game; United have conceded in all three of theirs.
EvidenceGoals tablexG trendProbable XIs
Handicap
Manchester City -1
City have won by two or more once in three league games, and United's 2.00 goals conceded per game is the softest number on the page.
EvidenceGoals tablexG trend
Result & goals
Match result
Manchester CityPrice 2.15
2/5City create 2.23 xG per league game and allow 0.89; United allow 1.43 and have conceded twice or more in each match.
EvidenceGoals tablexG trendTeam news
Both teams to score
YesPrice 1.43
3/5Both teams scored in two of each side's three league matches, and United have conceded in every one at 2.00 per game.
EvidenceGoals tableHead-to-head
Total goals
Over 2.5Price 1.48
3/5Two of three league matches for each side have passed 2.5 goals, and both average 2.33 scored per game.
EvidenceGoals tablexG trend
Cards
Total cards
Under 4.5
2/5Michael Oliver shows 2.67 yellows per game in three league matches, below the 3.87 league average, and City's matches have averaged 3.0 cards.
EvidenceRefereeDiscipline
Corners
Total corners
Under 9.5Price 1.94
2/5United's league matches average 8.7 total corners and City's 9.3, with only one in three passing 9.5 for either side.
EvidenceCorners
Confidence is the desk's conviction, 1–5; each evidence chip jumps to the chart or table behind the lean. Analyzed markets are part of the DeepPreviews membership — free to explore during early access. Prices are indicative market averages, not an offer.
Match sheet
- Kick-off
- 16:30 Manchester time
- Sunday
- Referee
- Michael Oliver
- 2.67 yellows per game in three league matches
- Weather
- 17 °C, showers
- 58% chance of rain
- Rest
- 7 days vs 8
- Neither side played in midweek
- Table
- 11th vs 1st
- 4 points vs 9, three matches each
Part one
Analysis
Where the goals come from, why the cards should stay low, and what the corner count says about two sides that both dominate the ball.
01Goals
Both attacks are working. United and City have each scored seven in three league matches, 2.33 per game. The split is at the other end: City have conceded 0.67 per game and United 2.00, and City's expected goals against of 0.89 per game says that is not luck.
United's numbers are also loud in both directions. They created 4.56 xG beating Ipswich 5-2 and 1.81 losing 2-0 at Hull, then managed 0.69 while drawing at Everton. That is a side whose ceiling is high and whose floor is a two-goal defeat. Two of three league matches have gone over 2.5 goals for both teams, and both teams scored in two of three for each.
City's route is narrower and steadier: 2.23 xG created per league game and never fewer than 2.12 in a match. Erling Haaland has three goals in three league matches. Against a defence that has conceded in every game, one clear chance usually decides who leads.
City concede 0.67 goals per league game this season; United concede 2.00. Both sides score 2.33.
Sample: Premier League 2026/27, 3 matches each
| Metric | Manchester UnitedHome | Manchester CityAway |
|---|---|---|
| Goals scored per game | 2.33 | 2.33 |
| Goals conceded per game | 2.0 | 0.67 |
| xG created per game | 2.35 | 2.23 |
| xG conceded per game | 1.43 | 0.89 |
| Both teams scored | 67% | 67% |
| Over 2.5 goals | 67% | 67% |
Bold marks the better figure in each row.
City have created at least 2.12 xG in every league match; United's average leans on one 4.56 evening against Ipswich.
Sample: Premier League 2026/27, 3 matches each
Manchester UnitedHome
avg xG 2.35 · avg xGA 1.43
2–0L
Hull City (A)
5–2W
Ipswich (H)
2–2D
Everton (A)
Manchester CityAway
avg xG 2.23 · avg xGA 0.89
2–1W
Bournemouth (H)
1–4W
Crystal Palace (A)
1–0W
Coventry (H)
xG forxG againstOldest match on the left.
Eight competitive meetings, three wins each and two draws, 2.50 goals per meeting; both teams scored in four of the eight.
3
Manchester United
2
Draws
3
Manchester City
17 Jan 2026·Premier League
Manchester United2–0Manchester City14 Sept 2025·Premier League
Manchester City3–0Manchester United6 Apr 2025·Premier League
Manchester United0–0Manchester City15 Dec 2024·Premier League
Manchester City1–2Manchester United10 Aug 2024·Community Shield
Manchester City1–1Manchester United25 May 2024·FA Cup
Manchester City1–2Manchester United3 Mar 2024·Premier League
Manchester City3–1Manchester United29 Oct 2023·Premier League
Manchester United0–3Manchester City
Home side listed first. Winner in bold.
What this supports
Full board ↑Double chance
Draw or Manchester CitySafer
Price 1.384/5Match result
Manchester City
Price 2.152/5Handicap
Manchester City -1Riskier
Price 3.672/5Both teams to score
Yes
Price 1.433/5Total goals
Over 2.5
Price 1.483/5
02Cards
Derby weeks promise cards and this one argues the other way. Michael Oliver has shown eight yellows in three Premier League matches this season, 2.67 per game, against a league average of 3.87 cards per match. He lets play run and reaches for a card late.
The sides differ. United's three league matches have averaged 4.3 total cards, going past 3.5 in two of three, and they take more than they draw: 1.7 shown to them, 2.7 to opponents. City's league matches have averaged 3.0 and none has passed 3.5. Seven of the last eight meetings have a card count on record: two, three, one, two, one, none and four.
Oliver has shown 2.67 yellows per game across three league matches, well under the 3.87 league average.
Sample: Premier League 2026/27, 3 matches
Michael Oliver
3 matches this season
league average
United's league matches average 4.3 cards, City's 3.0, against a league average of 3.87 per match.
Sample: Premier League 2026/27, 3 matches each
| Metric | Manchester UnitedHome | Manchester CityAway |
|---|---|---|
| Total cards per match | 4.3 | 3.0 |
| Cards shown to the team | 1.7 | 1.3 |
| Cards shown to opponents | 2.7 | 1.7 |
| Over 3.5 cards in the match | 67% | 0% |
Bold marks the better figure in each row.
What this supports
Full board ↑Total cards
Under 4.5
2/5
03Corners
Corner counts follow possession, and both teams keep the ball. United's league matches have produced 8.7 corners in total, 5.7 won and 3.0 conceded; City's have produced 9.3, 6.0 won and 3.3 conceded. Only one match in three has passed 9.5 for either side.
The derby shape trims it further. United sat deep and won a single corner while beating City 2-0 in January; City took six that day. When one side gives up the ball on purpose, the total tends to land in the eights rather than the elevens, and the league average is 8.97 corners a match.
United's league games average 8.7 total corners and City's 9.3; one in three has passed 9.5 for each.
Sample: Premier League 2026/27, 3 matches each
| Metric | Manchester UnitedHome | Manchester CityAway |
|---|---|---|
| Corners won per game | 5.7 | 6.0 |
| Corners conceded per game | 3.0 | 3.3 |
| Total corners per match | 8.7 | 9.3 |
| Over 9.5 corners in the match | 33% | 33% |
Bold marks the better figure in each row.
What this supports
Full board ↑Total corners
Under 9.5
Price 1.942/5
04The numbers that matter
- 0.67
- City goals conceded per league game
- Two goals in three matches, backed by 0.89 xG against per game. United must break a defence that has barely been opened.
- 6
- goals United have conceded in three games
- They have conceded in every league match this season, and twice or more in each of them.
- 2-0
- the last derby at Old Trafford
- January 2026: United won with 32% possession and seven shots on target to one. The blueprint exists.
- 2.67
- yellows per game for Michael Oliver
- Eight cards in three league matches this season, under the 3.87 league average per match.
Three matches is a thin base, and one lopsided evening moves every average on this page. Treat the direction as firm and the decimals as provisional.
Part two
Who plays, and how
United's midfield is thinner than the table suggests, and City's shape depends on who is fit on the left.
05Team news
Rashford is a doubt after being withdrawn at Everton; City expect Nico O'Reilly back, with Jeremy Doku still uncertain.
Manchester UnitedHome
Marcus RashfordFW
knock picked up at Everton
GoalsResult
DoubtfulManuel UgarteMF
cruciate ligament, out until spring
CardsResult
OutMatthijs de LigtDF
back
Goals
OutCarlos BalebaMF
ankle
Cards
OutAmad DialloFW
knock
Goals
OutTom HeatonGK
abdominal
Goals
Out
Manchester CityAway
Jeremy DokuFW
calf since the Community Shield
GoalsCorners
DoubtfulNico O'ReillyDF
back, hurt in the warm-up against Coventry
Result
Doubtful
Dots = impact on the starting XI (1 squad depth, 3 key starter); tags name the markets the absence moves. Club and national reporting from Monday and Tuesday this week; Maresca said O'Reilly is probable and Doku uncertain.
06Expected XI
Carrick should keep the 4-2-3-1 that drew at Everton with Sesko leading it; Maresca shields his back four with Anderson alone.
Manchester UnitedHome
Michael Carrick
Manchester CityAway
Enzo Maresca
Goalkeeper first, then each line left to right. Predicted XIs published on the eve of the match.
07Shape and plan
Anderson alone at the base
Maresca's 4-1-4-1 leaves one holder between the lines. Bruno Fernandes, with three goals in three league matches, lives exactly there. If United can turn Anderson before support arrives, the 0.89 xG City concede per game stops being a wall.
Sesko against Dias and Guehi
City have conceded twice in three league matches with Ruben Dias and Marc Guehi central. United's best route is the same one that produced 4.56 xG against Ipswich: early crosses and a striker attacking the near post rather than build-up through midfield.
The January counter
United won this fixture 2-0 with 32% possession and seven shots on target from eleven. Carrick has the personnel to repeat it: Mbeumo and Cunha running at a back four pushed high, with Shaw and Dalot holding depth rather than joining.
Set pieces both ways
City win 6.0 corners a league game to United's 5.7, but neither side concedes many: 3.0 and 3.3 against. With Maguire in the box on attacking corners, United's dead balls are their most reliable equaliser.
08Three duels
Bruno Fernandes
Manchester United · MF
Three goals in three league matches, and the only United player who can turn a holding midfielder into a hole.
3 goals in 3 Premier League matches
Bryan Mbeumo
Manchester United · FW
Scored against Ipswich and Everton and carries the ball forward faster than any team-mate; the outlet for a low block.
2 goals in 3 Premier League matches
Erling Haaland
Manchester City · FW
Three goals in three, against a defence that has conceded twice or more in every league match this season.
3 goals in 3 Premier League matches
Elliot Anderson
Manchester City · MF
The single pivot in Maresca's shape: everything City build and every counter they allow runs through him.
Part three
The call
Our verdict: how the match most likely goes, the score we would write down, and what would change our mind.
09How it could go
Three ways this match goes
Chance of each, out of 100 — most likely first
- 45%chanceMost likely
City squeeze the game
City keep the ball as they did for 78% of the Coventry match, United drop into a mid-block, and one Haaland finish decides it. The total lands on two or three goals, corners in the eights, and Oliver keeps his cards in his pocket until the last twenty minutes.
- 35%chance
United counter and it opens
Old Trafford presses the game rather than the ball. Mbeumo and Cunha find space behind a high line, United score first, and City chase with a back four already stretched. Both teams score, the total passes 2.5, and the finish is frantic.
- 20%chance
January again
United defend the way they did in the 2-0, ceding possession and territory and taking their two clear chances. City create 2 xG and score none of it. The corner count stays low because United never come out.
The chances are the desk's judgement, not market prices; together they add up to 100.
Two goals conceded in three matches against six: the derby is not a contest of attacks, it is a contest between one very good defence and one that has not kept anybody out yet.
10The verdict
City's defence decides a derby United can still make loud
The market makes Manchester City 44%, Manchester United 31% and the draw 25%, and the call follows the favourite. City lead every counted line this season: 9 points from three, 0.67 goals conceded per game, 0.89 xG against, and a striker with three goals in three. United score as freely as anyone at 2.33 a game, but have conceded in every league match and twice or more in each. That points one way for the result, another for the total: City win, both score. January's 2-0 here, and a fair price that makes City only 44%, keep the confidence at two.
What would change our mind
- If Jeremy Doku is passed fit, City's left side gets sharper and the handicap lean strengthens.
- If Marcus Rashford starts, United's counter is faster and the goals leans look better than the result ones.
- If Carrick sets up as he did in January with a deep block, drop the corners lean and the goal total with it.
Quick answers
- Who is favourite for Manchester United vs Manchester City?
- Manchester City are favoured at a market average of 2.15, with Manchester United at 3.06 and the draw at 3.84. City have taken 9 points from three league matches and conceded twice; United have four points from three and have conceded six.
- What is the predicted score for the Manchester derby?
- We call Manchester United 1-2 Manchester City. City have scored 2.33 goals per league game this season and conceded 0.67; United have conceded 2.00 per game and scored in two of their three matches.
- Who is the referee for Manchester United vs Manchester City?
- Michael Oliver has been appointed for Manchester United vs Manchester City. He has shown eight yellow cards in three Premier League matches this season, 2.67 per game, below the league average of 3.87. Manchester United's three league games have averaged 4.3 cards, Manchester City's 3.0.
- What is the recent head-to-head record?
- The last eight competitive meetings are level: Manchester United three wins, Manchester City three, two draws, at 2.50 goals per meeting. Both teams scored in four of the eight. United won the most recent meeting 2-0 at Old Trafford in January 2026.
- Who is injured for Manchester United vs Manchester City?
- Manchester United are missing 5 players: Manuel Ugarte, Matthijs de Ligt, Carlos Baleba, Amad Diallo and Tom Heaton, with Marcus Rashford a doubt after being withdrawn at Everton. Manchester City expect Nico O'Reilly back and Jeremy Doku is uncertain with a calf problem, but Erling Haaland arrives with 3 goals in 3 league matches.
Written 9 Sept 2026, 01:05 · Updated 9 Sept 2026, 01:30 UTC.
AI-generated match analysis for information and entertainment. Prices are indicative market averages, never an offer. Spotted an error? Tell the desk.
More previews
vs
Superliga · Round 7
FC Midtjylland vs FC Nordsjaelland
Second hosts third in Herning with a point between them: Midtjylland unbeaten in nine, Nordsjaelland scoring far less than their chances deserve.
vs
Serie A · Round 3
Udinese vs Lazio
Two 1-0 wins and two goals scored: Lazio arrive in Udine giving nothing away, into a fixture they have not won in five attempts.
vs
La Liga · Round 4
Elche vs Real Sociedad
Elche have conceded nine in three and sold their best centre-back. Real Sociedad have won seven of eight meetings and still cannot score.