REF-TOWN

MANIFESTO

We measure tendencies,
not correctness.

Football has a data blind spot. It's the referee.

Every player gets analysed. Every manager gets scrutinised. Every club's performance is tracked, modelled, and debated. Referees have been the one part of football left out of that conversation.

REF-TOWN changes that.

OUR PURPOSE

To provide clarity on referee behaviour through data, context, and transparency.

WHAT WE MEASURE ✓

How often a referee's decisions deviate from what the match context would typically produce

When decisions happen — the timing and clustering of cards across 90 minutes

How consistent a referee's output is from game to game

How decision patterns shift across score states, match stakes, and fixture types

Involvement in high-attention moments — late penalties, VAR interventions, red cards

WHAT WE DON'T MEASURE ✗

Whether any individual decision was correct

Referee quality, skill, or suitability for appointment

Intent, bias, or motivation behind any decision

Anything that cannot be observed directly from match event data

❝

REF-TOWN does not exist to criticise referees.
It exists to make referee behaviour visible.

WHY IT MATTERS

Football generates more data, more opinion, and more scrutiny than almost any other sport. Managers lose their jobs over results. Players are tracked to the centimetre. Clubs invest millions in marginal gains.

REF-TOWN doesn't exist to criticise referees. It exists to give fans, analysts, and anyone who cares about football a clearer picture of how individual officials tend to operate — over a full season, across different situations.

Understanding tendencies isn't the same as assigning blame. Data helps us see the game more clearly. Not judge it.

HOW REF-TOWN WORKS

1 1 — OBSERVE

Match events are collected from official data feeds after every game.

→

2 2 — COMPARE

Referee outputs are compared against match context and historical baselines.

→

3 3 — EXPLAIN

Scores describe tendencies and patterns, not judgments or correctness.

METHODOLOGY NOTE

All data is sourced from official match event feeds. Models are statistical and rule-based — no editorial judgment enters the scoring process. Metrics are computed per referee per season and updated after each gameweek.

The RT Score and SPOTLIGHT are proprietary indicators. They measure observable tendencies and patterns, not correctness or quality. RT Score methodology →

Patterns, not judgments.