Reconstruct the Contest

Establish the match frame

Record the competition, date, venue, teams and match phase before collecting statistics. Specify whether the figures cover regulation time, stoppage time, extra time or a penalty shoot-out. These distinctions matter because a 120-minute match offers more exposure than a 90-minute match, while shoot-out attempts are not ordinary match shots.

Use a documented event source, and retain the source's definitions. A shot on target, a blocked attempt and a dangerous attack are not interchangeable categories. If you combine an official match report with a separate event dataset, check that goals, substitutions and dismissals align before interpreting more detailed patterns.

Build a timeline before a verdict

Start with events that alter the conditions of play: goals, dismissals, injuries, substitutions and major shape changes. Then place shots, entries and possession spells around them. The goal is not to declare every visible incident a turning point, but to identify windows in which the tactical question changed.

1. 0–25 minutes · Level scoreCompare starting roles and the locations of early chances.
2. 26 minutes · Home team takes the leadBegin a new score-state segment rather than blending it with the opening period.
3. 60 minutes · Away team changes personnelRecord the role change and examine the following phase, including unsuccessful possessions.
4. Final phase · Away team chases the gameSeparate increased attacking volume from the quality and risk of those attacks.

Invented teaching sequence. Times illustrate segmentation; no actual match data is implied.

Choose the windows consistently. Five minutes after a substitution may contain only a few meaningful possessions. A longer segment offers more events but introduces more competing explanations. State the trade-off instead of presenting an arbitrary window as decisive proof.

Read possession through score state

A trailing side often has a reason to take more risks; a leading side may prefer space for counterattacks over sustained possession. Therefore, a late possession advantage can be an effect of the score rather than evidence that the trailing team was superior throughout. Aggregate possession hides those incentives.

Split the match into periods when the teams were level, leading or trailing. Within those segments, examine field location and chance creation. Possession in safe areas answers a different question from controlled entries into the penalty area. Neither should be casually renamed “control” without explaining what control means.

Dismissals require particular care. Eleven-against-ten statistics are not directly comparable to eleven-against-eleven statistics. Separate the periods, report their duration and avoid attributing the entire change to one tactical adjustment when the number of players also changed.

Inspect chances, not only totals

A team can accumulate several low-probability attempts while conceding one excellent chance. Shot totals reveal activity; chance-quality estimates help describe the opportunities. An expected-goals total adds modelled probabilities, not guaranteed goals, and does not tell us exactly how often that specific match “should” have been won without further modelling.

Sequence also matters. A rebound attempt can follow a saved shot within the same attack. Treating every probability as an independent opportunity may oversimplify the possession. If you use a model that adjusts for linked chances, name it; if you do not, describe the limitation.

A cumulative chance chart should use a common time axis and clearly labelled teams. Mark goals and dismissals without relying on colour alone. Do not smooth a stepwise event series into a curve that suggests continuous measurement between recorded attempts.

Distinguish association from intervention

A goal after a substitution is compatible with the substitution helping. It does not prove that the substitute caused the goal or that the same goal would not have happened otherwise. Opposition fatigue, a score-state change, a set piece or ordinary variation may also explain the sequence.

Stronger analysis specifies a mechanism: a new player occupied a different zone, the team accessed that zone more often, and the resulting possessions created better chances. Even then, a single match supplies limited evidence. Compare similar phases across multiple fixtures before promoting an observation into a general tactical rule.

Close the report in three sentences if necessary: what happened, what may explain it, and what remains uncertain. That structure preserves the excitement of the contest without turning a plausible story into an unsupported certainty.