The short version
A percentile describes position within a group—not overall quality.
Player percentiles are powerful because they translate many different football metrics onto a common 0–100 scale. But every result depends on who is included, which metric is measured and the context that produced the raw number.
What a football percentile means
If a player ranks in the 80th percentile for progressive passes, their raw value is equal to or higher than 80% of players with valid data in the active comparison cohort. It does not mean they complete 80% of their progressive passes, and it does not mean they are “80 out of 100” as a footballer.
Higher than or equal to approximately 8 in every 10 valid players in this specific cohort.
Percentile is a rank
Percentiles describe relative standing. They do not show the size of the gap between players. A player at the 80th percentile might be only fractionally ahead of the 70th-percentile player, or substantially ahead, depending on the distribution.
How player percentiles are calculated
Talent DNA takes the player’s raw metric value, compares it with every valid value in the filtered cohort and calculates the proportion at or below it. The result is rounded for display.
values at or below the player ÷ valid cohort values × 100With small cohorts, ranks move in larger steps. In a cohort of ten valid players, each player represents roughly ten percentile points. In a cohort of one hundred, the ranking is more granular.
Missing values matter
Only valid numeric values should enter the calculation. A player can therefore have a different effective sample size for different metrics. Always check data completeness before comparing profiles.
Why the cohort changes everything
The same raw value can produce very different percentiles in different groups. A centre-back’s progressive passing may look exceptional among centre-backs but ordinary in a cohort that also includes creative midfielders.
Build groups around sufficiently similar roles and responsibilities.
Reduce volatile small samples with a defensible playing-time threshold.
Match the genuine recruitment window when development stage matters.
Account for league strength, team dominance and tactical environment.
A good cohort reflects the recruitment question. There is no universally correct cohort, so record the filters used and test whether the conclusion survives sensible changes.
Read percentiles beside raw values
A percentile makes metrics comparable across different units, but the raw value preserves the football meaning. “92nd percentile” tells you relative rank; “8.4 progressive passes per 90” tells you the actual output being ranked.
Ask three questions
- Rank: Where does the player sit in this cohort?
- Magnitude: How large is the raw difference from other candidates?
- Meaning: Does this metric reflect the action required by the role?
High is not automatically better for every metric. Direction must reflect the meaning of the data, and volume should often be considered alongside efficiency.
How to read a football radar chart
Radar charts plot several percentile ranks around a common shape. They are effective profile summaries: you can quickly see whether a player’s output leans toward progression, creation, defending or another trait group.
- Check that every axis uses an appropriate direction.
- Read the cohort definition before comparing shapes.
- Use role-relevant metrics rather than filling every available axis.
- Check raw values where apparent strengths or weaknesses affect the decision.
- Avoid judging overall quality from the area of the shape.
Two similar radar shapes can hide different raw outputs, roles or tactical contexts. Use the chart to decide where to investigate, not where to stop.
A responsible percentile workflow
- Define the role and recruitment constraints.
- Select a coherent comparison cohort.
- Choose a small set of role-relevant metrics.
- Review percentile ranks and raw values together.
- Test the result with alternative sensible filters.
- Compare credible candidates using identical conditions.
- Validate the statistical signal through video and live scouting.
Talent DNA recalculates percentiles whenever the minutes, age or position filters change. That makes the relationship between cohort and result visible rather than hiding it behind a fixed score.
Build a percentile profile free →