How to Choose Football Metrics for a Player Role

The best metric set starts with a football role, not with the most popular numbers in a dropdown.

Position can define an initial population, but “midfielder” or “forward” is not a complete job description. The same position label can contain different responsibilities, zones and team contexts.

Write a Role Hypothesis

Describe what the player is expected to do in the intended team.

For example:

Receive under pressure, progress possession into advanced areas, support chance creation and protect central spaces after turnovers.

This is still too broad for a chart, but it provides responsibilities that can be mapped to evidence.

Build a Responsibility Table

For every responsibility, record:

Use the Metrics Glossary for current names and definitions. Do not assume that similarly named provider metrics are interchangeable.

Choose a Small Core

Start with one or two metrics per essential responsibility. More metrics do not automatically make a profile more complete.

Remove a metric when:

• it duplicates another measure;
• its unit does not fit the question;
• the role rationale is unclear;
• the population cannot support a fair comparison;
• the definition is not understood;
• it mostly reflects a responsibility outside the role.

The radar density guide provides a practical readability test once the candidate set exists.

Separate Volume, Efficiency and Outcome

Metrics can describe different parts of a process:

• volume: how often an action occurred;
• efficiency: how often attempts succeeded;
• outcome: what followed;
• involvement: how often the player entered the situation.

Combining one from each category can be useful, but only when the role hypothesis needs them. Do not interpret an efficiency percentage without the attempt volume behind it.

Decide on Units

Totals, per-90 rates, percentages and contribution shares answer different questions. A role profile may contain multiple unit types, but each axis must be labelled and interpreted independently.

Use per-90 metrics when playing-time normalization fits the question. Retain minutes and matches so a high rate from limited exposure is visible.

Check the Population

Metric relevance and comparison population are connected. A defensive action percentile among all players may mostly reproduce position differences.

Define:

• position or role group;
• competition scope;
• period;
• exposure rule;
• unit;
• age or other eligibility criteria.

The fair comparison guide turns those choices into a reusable contract.

Record What the Metrics Miss

Public event metrics may not capture:

• off-ball positioning;
• tactical instruction;
• communication;
• decision difficulty;
• opponent behaviour;
• physical or medical context;
• transfer suitability.

Write these limitations beside the chart. They are not a reason to abandon data; they define the next evidence required.

Choose the Chart After the Set

Use:

• a radar for a compact multidimensional profile;
• a player bar when metric rows and values need more space;
• a scatter for a relationship between two selected concepts;
• a swarm for the distribution of individual metrics.

The chart selection guide explains that decision.

Why the Standard Preset Is a Starting Point, Not a Role Model

The Standard FBPlot preset deliberately spans attacking, passing, possession, defending and duels. That breadth is useful for orientation, but a centre-back recruitment question should not inherit every metric simply because it is available. Alessandro Bastoni's pizza profile shows the more deliberate alternative: the Central defender preset concentrates the scan on distribution, duels and defensive actions, so the shape reflects a role hypothesis rather than the entire catalog.

Suppose the brief asks for a defender who can protect space, win contact and progress possession under pressure. The metric conversation can now become concrete. Interceptions may support anticipation, but they also reflect team structure. Aerial-duel percentage describes success rate, but not how often the player was challenged. Completed passes show volume, but not whether those passes broke a line. Each selection needs a sentence connecting it to the responsibility.

That sentence is the best defence against decorative metric collecting. If the analyst cannot explain what a slice tests, it should leave the final chart. The omitted dimensions do not disappear; they move into video review, qualitative notes or a second figure. A compact chart is not claiming that the role is simple. It is being honest about what this particular visual can support.

Build the Role Profile

Write the role hypothesis, select a small core from the current FBPlot glossary and explain every inclusion in one sentence. Remove any metric whose purpose cannot be defended. The resulting chart should be a transparent description of selected evidence—not a universal template for the position.