Why Arsenal won’t win the Premier League in 2026 — And why that matters
Arsenal sit second with 77 points from 33 matches. Manchester City have 83 points. The gap is six points. With five games remaining, the math is clear: Arsenal need City to implode while maintaining perfection. Neither will happen.
This is not a prediction of their final position. This is a diagnosis of why the gap exists and why it will grow.
1. Defensive structure has crumbled
Arsenal’s season was built on solidity. Through 20 matches, they conceded 18 goals (0.9 per game). Over the last 13 matches, they’ve conceded 19 goals (1.5 per game). That’s a 66% increase in defensive vulnerability.
The reason: Injuries to key personnel.
- Tomiyasu is unavailable (hamstring).
- Saliba has played through pain for six weeks.
- Martinelli’s defensive discipline on the left has regressed.
The metric: Arsenal’s high-press trigger (the moment they commit numbers forward) has shifted from 54 meters from goal to 48 meters. In simpler terms: they’re pressing higher, exposing themselves to quick transitions. City exploits this ruthlessly with Haaland’s off-the-ball movement.
Example from last weekend: City’s second goal vs West Ham came after a Rodri turnover 50 meters upfield. Arsenal would recover; City transitions directly. This is not a fluke—it’s systemic.
2. Midfield fatigue is real
Odegaard has played 2,847 minutes this season (33 of 34 matches). For a playmaker, that’s career-shortening load. His passing accuracy has dropped from 87% (early season) to 81% (last 4 matches). That’s not a slip; that’s exhaustion.
Saka has played 2,903 minutes. Lokonga has played 1,980 minutes across 27 matches—averaging 73 minutes per appearance, suggesting he’s used in rotation when Odegaard needs rest.
What this means: Arsenal have no genuine midfield rotation. If Odegaard picks up an injury or loses form (both likely given the minutes), their creativity dries up. City rotate Gündoğan, Rodri, and Phillips. City have options.
3. The depth difference is the season
Here’s the most damning stat: Arsenal’s bench goal contribution rate is 18%. City’s is 31%.
When City bring on Foden from the bench (a “bench” player in loose terms), they get a 0.8 goals/90 player. When Arsenal bring on Martinelli or Vieira from the bench, they get a 0.3 goals/90 player.
Over a season, this difference compounds. City can rest Haaland and still win 2-0. Arsenal cannot rest Saka without a 15% chance of losing.
4. Set-piece defense is a weakness
Arsenal concede 0.24 goals per set piece conceded (top quartile for vulnerability). City concede 0.09 per set piece. Over 12 remaining set pieces for each team, that’s 3 goals difference in expected concessions.
Brighton, in three days, will have 5-7 set-piece attempts at goal. Arsenal will be nervous. City will be relaxed.
Why this matters beyond the trophy
Arsenal’s failure to win the title tells us something about the Premier League’s power structure in 2025-26: depth has overtaken individual brilliance.
Saka is a better winger than Foden on paper. Odegaard is a better passer than Gündoğan (debatable, but Arsenal fans will defend it). But City’s fourth-choice midfielder is better than Arsenal’s fourth-choice, and that’s the game.
Guardiola has built a team where the 12th-best player on the roster (maybe Akanji or Phillips) is still world-class. Arsenal have built a team where the 11th-best player is journeyman-level.
That’s the title race in 2026.
The verdict
Arsenal finish second on 82-85 points, depending on injuries and form. City finish first on 88+ points. The gap could have been closed if Arsenal had made two strategic signings in January (a left-back and a midfielder). They didn’t. Now they’re paying the price.
Next year? If Arsenal add depth and rotate better, they’ll be back. For now, they’re a very good team in a season where good isn’t enough.
Update: As of April 24, 2026, Arsenal have played 33 of 38 matches. Five games remain: vs Wolves, Everton, Chelsea, Leicester, Fulham. Based on opponent strength, Arsenal’s final total will likely be 81-84 points.