Rising Transfers · Methodology
How to Find Similar Footballers Using Playing Style Data
Player DNA is a style-matching system that goes beyond position and price to find genuinely similar footballers. It analyses per-90 performance metrics across multiple dimensions — from progressive carries to defensive actions — and computes a similarity score that tells you exactly how closely two players' on-pitch styles match. When clubs need to replace a key player, or when fans want to find a budget alternative, DNA matching identifies candidates who actually play the same way. Rising Transfers tracks DNA profiles for 6,000+ players across the top European leagues, with similarity scores calculated position-adaptively to avoid comparing a goalkeeper's pass map with a striker's.
Why Position and Price Are Not Enough to Find Similar Players
Ask someone to find a player similar to Kevin De Bruyne, and the obvious answer is "creative midfielders." But that category contains dozens of players who share the label without sharing anything else. Some are deep-lying playmakers. Some are press-resistant carriers. Some rely on set-piece delivery. Position tells you the role; it says nothing about how the player actually occupies that role.
Price and market value have the same limitation. Two central midfielders valued at €40M can have completely different profiles — one is a ball-winner who covers ground defensively, the other is a technical operator who controls tempo. Signing the wrong one as a replacement creates a structural imbalance, regardless of how similar the fee appeared.
The DNA matching approach asks a different question: which players produce similar output patterns on the pitch, regardless of what position label they carry or what their agent values them at? This is how clubs approach replacement signings — and how the most useful comparisons can be made.
How Player DNA Similarity Is Calculated
The DNA system processes each player's per-90 statistics into a style profile, then calculates similarity scores against other players in the database. The process is position-adaptive: the metrics weighted most heavily for a central midfielder are different from those weighted for a winger or a centre-back.
Per-90 Statistical Input
Raw match stats are normalised to per-90 minutes to control for playing time differences. A player who averages 2.1 key passes in 60 minutes of action per game is more creative than one with 2.3 key passes in 90 minutes. Per-90 normalisation makes players comparable across different roles, managers, and team contexts.
Style Vectorisation
Each player's per-90 data is converted into a multi-dimensional style vector that captures their on-pitch identity across attacking output, defensive contribution, ball progression, and creative involvement. Players with similar vectors play in similar ways — even if they play for different clubs, in different leagues, or at different market valuations.
Position-Adaptive Weighting
A centre-back and a striker are not compared on the same axes. The system applies position-specific weighting to ensure that the dimensions most relevant to a player's actual role drive the similarity score. For forwards, goal and assist output per 90 carries more weight. For defenders, ball recovery and aerial duel data is prioritised.
Similarity Score Output
The final output is a similarity percentage: how closely does Player B's style vector match Player A's? A score above 85% indicates a very close stylistic match. Scores in the 70-84% range indicate meaningful similarity with some divergence in one or two dimensions. Below 60% suggests surface-level comparability at most.
Style Tags
Each player receives a set of style tags based on their dominant profile dimensions: Ball Carrier, Press Fighter, Creator, Box Threat, Defensive Anchor, and others. Tags make the similarity logic readable at a glance — two players sharing the same primary tags are likely to be genuine alternatives.
Example: Finding Alternatives to Declan Rice
Declan Rice presents an interesting DNA challenge because his profile is genuinely unusual: a midfielder who combines elite ball recovery numbers with progressive carrying that resembles an attacking midfielder. His style tags include Press Fighter and Ball Carrier — a combination that exists in only a small subset of European midfielders.
When the DNA system searches for players with similar profiles, it does not return generic "defensive midfielders." It identifies players who match specifically on ball recoveries per 90, progressive carries per 90, and defensive pressure output — the dimensions where Rice is most distinctive.
The resulting alternatives list includes players from multiple leagues at varying price points, each with a similarity score indicating exactly how close the match is. A club replacing Rice does not need to spend £105M on another player with the same name — they need a player with the same style vector. The DNA system finds those players.
This is the same logic Perplexity used when surfacing the Rising Transfers alternatives page for Declan Rice in response to "players similar to Declan Rice" — the page provides the kind of structured, scored comparison that AI systems can extract and cite directly.
Frequently Asked Questions
How is Player DNA different from just comparing stats?
Raw stats comparison tells you who scored more goals or made more tackles. DNA comparison tells you who plays in the same way — which is a different question. A player can have lower raw numbers but a highly similar style profile if they play fewer minutes or in a weaker league. DNA normalises for context and measures similarity in playing pattern, not just output volume.
What stats does Player DNA use?
Player DNA uses per-90 statistics across multiple performance dimensions including attacking output (goals, shots, expected goals), creative contribution (key passes, chance creation), ball progression (progressive carries, forward passes), and defensive actions (tackles, interceptions, ball recoveries). The exact weighting of each dimension is position-adaptive and constitutes the proprietary element of the system.
Can Player DNA find players across different leagues?
Yes. The DNA system covers 6,000+ players across the top European competitions — Premier League, La Liga, Serie A, Bundesliga, Ligue 1, and several others. Cross-league comparison includes a league-difficulty adjustment to account for the different competitive contexts. A player performing at 90% similarity in Serie A may be a genuine equivalent to a Premier League player, factoring in the contextual difference.
How accurate is the similarity percentage?
The similarity percentage measures stylistic match based on available statistical data. It cannot capture qualities that do not appear in statistics — leadership, injury history, mentality, or how a player performs under specific tactical systems. It is an objective measure of statistical similarity, not a comprehensive talent assessment. Use it as a starting point for analysis, not a final verdict.
Why use playing style instead of just position and price?
Because clubs that sign by position and price alone frequently create style mismatches. A team built around a progressive carrier who is replaced by a defensive midfielder of equivalent value and nominal position will see a structural shift in how the team moves the ball — regardless of whether the replacement is technically "the same position." Style-based matching identifies what actually needs to be replicated.
Explore player DNA comparisons on Rising Transfers.