What Player DNA is
“Descriptive, not predictive: nothing here is a forecast.”
Player DNA ranks a player's rates against the other qualified players in the same competition and season. The shipped metrics are grouped as:
| Group | Metrics |
|---|---|
| Scoring | Goals per 90, shots per 90, shots-on-target rate, goals per shot |
| Creation | Assists per 90, key passes per 90, goals + assists per 90 |
| Role | Start rate, minutes per appearance |
| Discipline | Cards per 90 (lower is better) |
Families such as ball progression, defensive actions, aerial ability and set-piece contribution are part of the product's design, but they are not shipped metrics today. Where the source cannot support a metric — line-breaking passes need player positions, carries need carry events, pressure resistance needs a pressure flag — the methodology marks it blocked rather than approximating it.
Per-90 rates and the minutes gate
rate per 90 = count × 90 / minutes played qualified = at least 450 nominal minutes in the competition-season
- Illustrative: a substitute scores 2 goals in 110 minutes → 1.64 goals per 90.
- A starter scores 9 in 1,800 minutes → 0.45 per 90.
- The substitute's rate is higher — on a twentieth of the sample, often against tired defences late in matches.
Minutes are nominal — built from lineups and substitutions — not a provider's playing-time figure. Below 450 minutes the profile says how many of the required minutes the player has and shows complete counts with no ranks. A metric without a rank shows "no rank"; nothing is imputed, and the radar draws only ranked metrics.
What it does not prove
- Not a forecast. An 85th-percentile shots rate says the player shot more than most qualified peers — not how many shots the player will take next match.
- Role and team shape it. A forward on a dominant team gets more shots; the percentile does not separate player from system.
- Cohorts differ. A percentile in one league is not comparable with a percentile in another.
- Finishing is noisy. Goals per shot over a season is mostly variance — see Sample size.
Cohorts are the whole competition-season rather than position groups, so defenders and forwards are ranked together on attacking rates. Read the metric with the player's position in mind.