Observational Analysis: Pressing Patterns & Transitional Play — Arsenal 1-1 Man City (21 Sep 2025)
Hi @martineastwood, thanks for the excellent #39 (Press Bypass Rate + SPTG proposal). I watched the Arsenal vs Man City fixture at the Emirates (21 Sep 2025, 1-1 draw) and wanted to share my observational data to complement the feature request discussion below.
Match Overview
| Detail |
Value |
| Date |
21 September 2025 |
| Venue |
Emirates Stadium |
| Result |
Arsenal 1-1 Man City |
| Goal Scorers |
Haaland 9' (City) — Martinelli 90+3' (Arsenal) |
| Cards |
Trossard (red, 35') |
Pressing Patterns I Observed
1. Arsenal's Aggressive 4-1-4-1 Man-to-Man High Press
Arsenal started with a 4-1-4-1 out-of-possession structure and a man-for-man high press. Ødegaard and Havertz formed the first press line, cutting off the CB-to-CB passing lanes. The back-pass trigger was decisive — City repeatedly tried to recapture the ball by playing backward within 1-2 seconds of receiving, which immediately triggered the Arsenal press. Haaland's 9' goal came directly from a goal-kick long ball that bypassed the entire press structure.
2. City's Low-Block Patience Post-60'
After 60', City dropped into a low block that completely nullified Arsenal's press. Instead of trying to break the press, they kept possession in safe zones and waited for transitions. The PPDA inflation was visible: Arsenal's press became more desperate (reaching for tackles and interceptions) rather than more effective. This is a textbook example of how elite teams neutralise high presses through patience.
3. Press Fatigue — The 60-75 Minute Window
I noticed a very clear degradation window between 60-75 minutes:
- Vertical compactness expanded from ~12-14m to ~18-20m
- PPDA jumped significantly in the second half
- Counter-press recovery dropped from ~35-40% to ~20-25%
- The adrenaline from the Trossard red card (Arsenal went down to 10 men early, which paradoxically intensified the press for the first hour) faded
- Vidal/Eze introduction briefly reignited pressing intensity, but the structural fatigue was evident
Transitional Play Observations
1. Direct Transition Vulnerability (The "Haaland Death Zone")
Haaland's 9' goal perfectly illustrated the structural vulnerability of a high press against direct long-ball transitions. The ~18-20m vertical gap between Arsenal's press line and their defensive block was completely unguarded. City played a goal-kick long ball → Doku/Palmer found the channel → Haaland finished in ~10 seconds. No Arsenal player was covering that space.
2. Decision Quality in Transition
Several moments stood out where Arsenal players in transition had multiple passing options but chose to dribble into congested areas instead of releasing teammates into space. I specifically noticed a situation where a player had a 4v4 with three clear passing options but dribbled into 3 defenders instead — this kind of "decision-quality failure" in transition is invisible to current xG models but has a massive impact on outcomes.
3. Set-Piece Transition Gap (The Martinelli Equalizer)
Martinelli's 90+3' equalizer perfectly illustrated the set-piece-to-open-play transition gap. Arsenal had won 11 corners. City defended in set-piece shape. But as they were transitioning to open-play defensive shape, Eze's lob found Martinelli unmarked. The gap between "set-piece defensive shape" and "open-play defensive readiness" was exactly where the goal was born.
4. Counter-Attack Consistency Problem
Arsenal's best transition chances came on City errors (e.g., Donnarumma's hesitation) rather than through coordinated pressing triggers. The counter-attacks were sporadic — impactful when they happened, but not sustainable as a tactical strategy.
Proposed Quantitative Metrics for penaltyblog
Building on #39, here are specific empirical observations that could be coded as analytical features:
| Metric/Feature |
What I Observed |
Potential Implementation |
| Press Bypass Rate (PBR) |
City bypassed Arsenal's press 3+ times in the first HT with long diagonals; PPDA <8 looked dominant but was easily circumvented |
Compute % of press actions followed by >25m long balls |
| Set-Piece Transition Gap (SPTG) |
Martinelli goal: 11 corners, City still transitioning from set-piece shape when Eze's lob found target |
Measure time from set-piece clearance to first open-play pass in defensive transition |
| Transition Decision Quality Index (TDQI) |
Suboptimal dribbles instead of passes to space in transition (at least 2 clear instances in 2nd HT) |
Binary flag: 'pass_to_space' vs 'dribble_under_pressure' + positional context |
| Press Fatigue Detector |
Identical 60-75' degradation window across both Arsenal fixtures; PPDA decay >25% after 60' |
Rolling-window PPDA with fatigue_index = (post_60_ppda - pre_60_ppda) / pre_60_ppda |
| Vertical Compactness Change |
Expanded from ~14m to ~18-20m during fatigue window |
Compute Y-axis spread of defensive unit between 55-65' vs 75-85' |
Why This Belongs in penaltyblog
penaltyblog already has:
- ✅ Event ingestion via MatchFlow + StatsBomb/Opta
- ✅ Statistical models (Poisson, Dixon-Coles, Bayesian)
- ✅ Pressing event extraction capabilities
What's missing is the tactical context layer — the reasons why things happen. These metrics would:
- Distinguish "effective press" from "press that was bypassed" (PBR)
- Quantify the set-piece-to-open-play vulnerability window (SPTG)
- Capture decision quality in transition, not just speed (TDQI)
- Detect press fatigue patterns automatically across match datasets
These features would complement the existing betting/odds models by adding a new dimension of tactical process analysis that could inform match predictions far more deeply than outcome-only models.
Data Sources
- Opta Stats (Arsenal 1-1 Man City, 21 Sep 2025)
- Arsenal.com Official Match Report
- BBC Sport Post-Match
- FBRef Match Center
- Coaches' Voice / Total Football Analysis video breakdowns
Happy for the community to challenge, refine, or expand on these observations. Would love to see these metrics implemented as MatchFlow filters!
Observational Analysis: Pressing Patterns & Transitional Play — Arsenal 1-1 Man City (21 Sep 2025)
Hi @martineastwood, thanks for the excellent #39 (Press Bypass Rate + SPTG proposal). I watched the Arsenal vs Man City fixture at the Emirates (21 Sep 2025, 1-1 draw) and wanted to share my observational data to complement the feature request discussion below.
Match Overview
Pressing Patterns I Observed
1. Arsenal's Aggressive 4-1-4-1 Man-to-Man High Press
Arsenal started with a 4-1-4-1 out-of-possession structure and a man-for-man high press. Ødegaard and Havertz formed the first press line, cutting off the CB-to-CB passing lanes. The back-pass trigger was decisive — City repeatedly tried to recapture the ball by playing backward within 1-2 seconds of receiving, which immediately triggered the Arsenal press. Haaland's 9' goal came directly from a goal-kick long ball that bypassed the entire press structure.
2. City's Low-Block Patience Post-60'
After 60', City dropped into a low block that completely nullified Arsenal's press. Instead of trying to break the press, they kept possession in safe zones and waited for transitions. The PPDA inflation was visible: Arsenal's press became more desperate (reaching for tackles and interceptions) rather than more effective. This is a textbook example of how elite teams neutralise high presses through patience.
3. Press Fatigue — The 60-75 Minute Window
I noticed a very clear degradation window between 60-75 minutes:
Transitional Play Observations
1. Direct Transition Vulnerability (The "Haaland Death Zone")
Haaland's 9' goal perfectly illustrated the structural vulnerability of a high press against direct long-ball transitions. The ~18-20m vertical gap between Arsenal's press line and their defensive block was completely unguarded. City played a goal-kick long ball → Doku/Palmer found the channel → Haaland finished in ~10 seconds. No Arsenal player was covering that space.
2. Decision Quality in Transition
Several moments stood out where Arsenal players in transition had multiple passing options but chose to dribble into congested areas instead of releasing teammates into space. I specifically noticed a situation where a player had a 4v4 with three clear passing options but dribbled into 3 defenders instead — this kind of "decision-quality failure" in transition is invisible to current xG models but has a massive impact on outcomes.
3. Set-Piece Transition Gap (The Martinelli Equalizer)
Martinelli's 90+3' equalizer perfectly illustrated the set-piece-to-open-play transition gap. Arsenal had won 11 corners. City defended in set-piece shape. But as they were transitioning to open-play defensive shape, Eze's lob found Martinelli unmarked. The gap between "set-piece defensive shape" and "open-play defensive readiness" was exactly where the goal was born.
4. Counter-Attack Consistency Problem
Arsenal's best transition chances came on City errors (e.g., Donnarumma's hesitation) rather than through coordinated pressing triggers. The counter-attacks were sporadic — impactful when they happened, but not sustainable as a tactical strategy.
Proposed Quantitative Metrics for penaltyblog
Building on #39, here are specific empirical observations that could be coded as analytical features:
Why This Belongs in penaltyblog
penaltyblog already has:
What's missing is the tactical context layer — the reasons why things happen. These metrics would:
These features would complement the existing betting/odds models by adding a new dimension of tactical process analysis that could inform match predictions far more deeply than outcome-only models.
Data Sources
Happy for the community to challenge, refine, or expand on these observations. Would love to see these metrics implemented as MatchFlow filters!