Arsenal's points per game from August to March under Mikel Arteta stands at a title-winning 2.38. From April onwards, that number plummets to 1.65.

It is a statistical cliff edge. The recent defeat to Manchester City, which severely dented their Premier League aspirations, was not an anomaly. It was the latest data point in an established, measurable trend.

We often reach for psychological explanations when a team stumbles at the finish line. As The Mirror reported following the match, the familiar claims of a bottling mentality have resurfaced, amplified by Erling Haaland's tactical mind games. It is an easy narrative to sell. It is also entirely insufficient.

Psychology is notoriously difficult to quantify. Pass completion under pressure, defensive line height, and expected goals against (xGA) are not. The underlying metrics reveal a much darker reality for Arsenal. The numbers tell a story of a system that becomes completely rigid when the margins shrink.

The structural collapse of the press

Arsenal’s modern identity is built on manufacturing high turnovers. When they are at their fluid best, their Passes Per Defensive Action (PPDA) hovers around 8.5. They choke opponents high up the pitch, turning defensive actions into immediate attacking transitions.

Against City, that defining intensity vanished. The pressing triggers were consistently delayed. Arsenal’s PPDA ballooned to 14.2, marking their second-most passive defensive display of the entire campaign.

This drop-off is rarely about a lack of physical effort. It is entirely about risk aversion. Martin Ødegaard and Declan Rice found themselves dropping five to ten yards deeper out of possession than they routinely did in January.

The result is a fundamentally detached midfield. The front line still attempts to press the initial phase of build-up, but the gap between the forwards and the double pivot expands significantly. Elite teams play through that expanded space effortlessly. City did exactly that, bypassing the first line of pressure with single, vertical passes.

When a team stops winning the ball high, they have to defend their own penalty area for longer stretches. Arsenal are competent in a low block, but it invites an unsustainable volume of pressure.

The false comfort of possession

Here is the counterintuitive element of Arsenal’s late-season struggles. They rarely lose control of the ball entirely. In fact, they often monopolize it.

In the matches they have dropped points in since the start of April over the last three seasons, Arsenal average 61% possession. They are not being battered into submission by superior midfields dictating the tempo.

The central issue is what they actually do with that possession. Their ball circulation speed drops dramatically. Progressive passes into the penalty area fall by nearly 30% compared to their rolling season average.

They move the ball in a predictable U-shape around the perimeter of the opposition's defensive block. It looks like territorial dominance to the naked eye, but the data shows it is entirely sterile. The opponent rests while Arsenal pass sideways.

We can measure this sterility through field tilt. Field tilt calculates the share of final third passes a team makes compared to their opponent. It is a pure metric of where the game is actually being played.

Against mid-table opposition earlier in the season, Arsenal routinely register a field tilt above 65%. Against City, and indeed in crucial away fixtures during this current run-in, that number rarely breaks 45%.

They are surviving, not dictating. Against a side that averages over two expected goals per game, relying on survival rather than dictation is mathematically doomed to fail over 90 minutes.

The individual data deficit

Systemic failures inevitably drag down individual metrics. The data shows clear drop-offs for key personnel when the structure fractures.

Consider Bukayo Saka's isolation. When Arsenal are flowing, Saka receives the ball on the move, usually following an underlapping run from Ben White that drags a defender away. In the final weeks of the season, the nature of his touches changes completely.

He frequently receives the ball static, backed into a set double team. Consequently, his successful take-on rate plummets from 48% before April to just 31% during the run-in.

This is not a world-class player suddenly losing his technical ability. It is a failure of the offensive structure to provide him with favorable isolation scenarios. He is being asked to solve complex defensive blocks entirely on his own.

Declan Rice suffers a similar statistical decline. Rice was purchased specifically for these high-leverage fixtures. His ball-carrying numbers are elite. Yet, when the team drops deep out of possession, his average starting position shifts back by nearly eight yards.

Instead of intercepting passes in the middle third and launching immediate transitions, he is forced into last-ditch blocks on the edge of his own six-yard box. His progressive carries per 90 minutes drop from an elite 3.2 down to an average of 1.4 in these specific matches.

The Haaland efficiency problem

Then there is the highly specific problem of defending the current iteration of Manchester City. Erling Haaland does not require constant involvement to completely dictate the geometry of a game.

Against Arsenal, Haaland often registered fewer than 20 touches across 90 minutes. Yet, his mere positioning forces William Saliba and Gabriel Magalhães into a constant state of tactical compromise.

If the defensive line pushes up to compress the midfield space and support the press, Haaland immediately exploits the channel in behind. If they drop deeper to manage the threat of his pace, City’s technically gifted midfielders receive the ball under zero pressure.

This creates a cascading failure across the pitch. Because the center-backs cannot step aggressively into midfield without risking a fatal through-ball, Rodri and Kevin De Bruyne are afforded an extra half-second on the ball. At this level of the sport, a half-second is an eternity.

Arsenal's defense chose the passive route. They dropped off, minimized Haaland's direct threat, and subsequently allowed the midfield to be overrun by City's engine room. It was a calculated risk, and the data shows it failed.

The math of the run-in

City have fundamentally broken the historical curve for what is required to win a Premier League title. Since Pep Guardiola's arrival, their win percentage in the final ten games of a season sits remarkably above 85%.

You cannot simply be good to beat them in April; you must be flawless. Arsenal’s current trajectory, factoring in their historical drop-off, suggests a final points tally hovering around the 86-point mark.

Historically, that constitutes a highly commendable season. In the Guardiola era, it is usually only good enough for a comfortable second place. The mathematical reality is incredibly harsh.

The tactical rigidity shown in recent weeks highlights a critical flaw. Arteta and his team still do not possess a reliable B-plan for when their primary system is neutralized.

They consistently try to execute plan A with slightly more intensity. But against a team that manipulates space as intelligently as City, running harder without making structural adjustments is entirely useless.

The prevailing narrative will undoubtedly focus on mental fragility and the heavy weight of expectation. The underlying data suggests a far more difficult problem to solve: a predictable offensive structure and a high press that inevitably fractures against elite distribution.