Arsenal3-0Manchester City

The FA Community Shield opens the 2026 season when Arsenal and Manchester City meet at Wembley Stadium. Both sides are playing their first competitive match since May: Arsenal last appeared on 30 May, City on 24 May, making this a true season opener for each.

The model gives Arsenal a 67.6% probability of winning, the most likely outcome. That projection is built on Arsenal’s home record—16 wins, 2 draws and 2 losses in their last 20 at home—and City’s comparatively mixed away form: 10 wins, 5 draws and 5 losses in their last 20 on the road.

With Arsenal designated as the home side at a neutral Wembley, the model’s projected scoreline is 1-0. The next most likely exact scores are 2-0 and 1-1.

Arsenal’s underlying numbers are strong. In their last 20 home matches they won 16, drew 2 and lost 2, an 80% win rate, while averaging 1.95 goals scored and 0.65 conceded. That defensive average is the key pillar of the model’s outlook.

Across their final five matches of the previous season, Arsenal went DWWWW (most recent first), averaging 1.2 scored and 0.4 conceded. That recent run is less prolific in attack than their longer home sample, but the defensive solidity is consistent.

Those figures are all from the previous campaign; the Community Shield is the first test of the new season.

Manchester City’s away record over the last 20 matches shows 10 wins, 5 draws and 5 losses—a 50% win rate. They averaged 1.55 goals scored and 1.25 conceded, a profile that is respectable but significantly weaker than Arsenal’s home output.

In their last five matches overall, City produced LDWWD (most recent first), with an average of 2.2 goals scored and 1.2 conceded. The sequence includes one defeat and two draws, which is why their recent form is not as clean as the attacking average might suggest.

Those numbers also come from the previous season, offering the only recent sample available for a side that has not played a competitive match since 24 May.

The four most recent meetings between these sides are evenly split in outcomes: each has one win and two draws. Arsenal outscore City 9-6 across those games, a combined 15 goals at an average of 3.75 per meeting.

The full results are: Manchester City 2-1 Arsenal on 19 April 2026; Arsenal 1-1 Manchester City on 21 September 2025; Arsenal 5-1 Manchester City on 2 February 2025; and Manchester City 2-2 Arsenal on 22 September 2024.

One of the four meetings finished with two or fewer total goals. The sample is small, so the head-to-head alone provides limited guidance, but it is at least consistent with the model’s expectation of a moderate-scoring game.

Wembley Stadium is the host venue, with Arsenal listed as the home side. Because Wembley is a neutral ground, the home/away splits should be applied with caution; they still give a sense of each team’s recent tendencies.

Arsenal’s 16-2-2 home record and 0.65 goals-conceded average is the standout split. Manchester City’s 10-5-5 away record and 1.25 goals-conceded average is the clear gap in this fixture.

The difference between those two profiles helps explain why the model assigns such a high probability to an Arsenal win, even though the match is not being played at Arsenal’s usual home ground.

Both teams enter this match after an extended off-season break. Arsenal have 78 days between matches, City 84. Neither has played a competitive fixture since late May, and this is the 2026 season opener for both.

There is no league table data provided for either side, so the analysis rests on the home/away splits and recent form from the previous season. The model’s confidence is rated medium, a sign that the projection carries more uncertainty than a typical mid-season match.

The long layoff affects both sides equally, but the lack of recent competitive action means early-season form is hard to project from data alone.

The model, based on a Poisson goal distribution, gives Arsenal a 67.6% win probability, a 20.5% draw probability, and a 12% away win probability. The single most likely exact scoreline is 1-0 at 14.2%, followed by 2-0 at 13.8% and 1-1 at 9.5%.

Expected goals are 1.94 for Arsenal and 0.67 for City, which produces a projected 1-0 scoreline. The model expects about 2.61 total goals, with a 73.5% chance of at least two goals, 48.5% of at least three, and 26.7% of four or more. The probability of each side finding the net is 41.8%.

From the home side’s perspective, the model sees a 41.8% chance of Arsenal winning by two or more, a 25.8% chance of a one-goal home win, a 20.5% chance of a draw, an 8.9% chance of a one-goal away win, and a 3.1% chance of City winning by two or more. The distribution clusters around a home win, with a meaningful tail toward a draw.

The data collectively points to an Arsenal win. Their home form is dominant, the model favors them at 67.6%, and the projected scoreline is 1-0. The probability of a City win, at 12%, is low enough that the most likely outcome is clear.

The main caveat is the season-opening context. Both teams are coming off a long break, and the model’s medium confidence rating flags that such matches can deviate from the underlying numbers. City’s away record, while weaker than Arsenal’s home record, still includes a 50% win rate.

If the model is wrong, the likely reason would be City’s recent attacking form—they averaged 2.2 goals in their last five—outstripping the defensive projection. But the model still assigns them just a 12% win probability, and the balance of evidence favors Arsenal.