BALLER

Beyond individual statistics lies collective intelligence.

BALLER is a foundational AI that represents footballers in their true context: teammates, opponents, roles, tactics, playing style. Trained on team lineups from hundreds of thousands of matches, it captures the collective identity of every player.

Our mission: to put team context at the core of every football statistic — and unlock a new era of smarter metrics, scouting, and data-driven insights.

🚀

Build with BALLER

Test our models & API →

⚽

goal Contribution

The new way to quantify decisiveness in football →

🔍

SquadGen

From a few names to a coherent squad, instantly →

🔄

Transfer Portal

Simulate player performance in new teams →

Football is deeply collective — analytics are deeply individual


Watch any football match — from Barcelona's Tiki-Taka era to Luis Enrique's PSG in 2025 — and it's impossible to ignore how deeply collective football is. Players move as units. Systems flow. Roles shift.

But once we look at the data, football becomes oddly individual. Every player is reduced to a row of stats: passes, shots, touches — detached from the teammates who made them possible, the opponents they faced, or the tactical context that shaped their actions.

With BALLER, we're closing this gap. Our models bring back the context — putting systematically teammates, opponents, systems, roles and playing style into the heart of every metric, to reflect the true nature of the game.

Tactical Collective Play

From context to identity: learning what truly defines a player


If football is collective on the pitch, it should be in the data too. Instead of treating players as isolated numbers, we model their identity through the lens of context — who they play with, against, in which role, and over time.

Inspired by language models, we treat matches as sequences and players as contextual tokens. Trained only on team lineups — no stats — our model learns a deep representation of each player, shaped entirely by the collective structures they evolve within.

The result? A context-rich identity vector for every player — ready to power the next generation of context-aware analytics and scouting tools.

Transformer Football Animation

From identity to transfer: simulating how performance moves across contexts


Once a player's identity is shaped by the collective structures around them, a new capability emerges: the ability to simulate how that identity behaves when the context changes. This is the core of the performance transfer problem — a layer football analytics has always lacked.

With BALLER-Transfer Portal, we generate game contexts at scale and model how a player's output shifts when placed in a different team, role, or tactical environment. Instead of treating performances as fixed attributes, we treat them as context-dependent behaviours that evolve with teammates, opponents, and match conditions. The result is a model that finally captures the true dynamics of performance transfer — predicting, with unprecedented granularity, how a player will perform when moving from one environment to another.

In a world where clubs spend tens of millions on players without ever truly knowing how their performance will translate, BALLER brings a foundational, identity-aware simulation engine built precisely for that question. Predict the shift. Explore scenarios. Anticipate fit. This is the missing layer that turns contextual identity into actionable transfer intelligence.

BALLER Transfer Portal

Collecting a new generation of context-aware metrics


Even though we're now able to model the context around players with unprecedented accuracy, we're still convinced that the way football data is collected today does not reflect the collective DNA of the game. And when the data fails to capture the collective, it breaks the storytelling.

Everything is built around the same flagship metric — goals and assists (G/A). But G/A only highlight the finishers. In reality, 35% of the value behind a goal doesn't come from the scorer or the final passer. Entire pieces of the action, and entire types of players, simply disappear from the narrative.

That's why we built goal Contribution (gC): a metric that captures every player involved in a goal-scoring sequence and measures their true impact. gC is context-aware by design, built directly from sequences of play, and it serves as the foundation for a whole new family of contextual metrics — each of them essential to our ambition of building the most context-native foundational model in football analytics.

Goals and assists are the currency of football — the way the world measures decisiveness — and most of the sport's storytelling and analytics revolves around them. But G/A miss 35% of the economy. With gC, our better currency, we're fixing that gap and reshaping football at every level, from the narrative to the analytics.

Messi Goal Bernabeu

Who is BALLER for?

A foundational layer for next-gen football intelligence — plug it into your stack, build on top, or power your predictions.

📡 Football Analytics Companies

Integrate player contextual embeddings directly into all your metrics and analytics models. Build on top of our API

🧠 Developers, Analysts & Researchers

Use our embeddings and models to build new tools and run unprecedented experiments.

📱 Medias, LiveScore Apps & Content Creators

Power your football content and narratives with our goal Contribution (gC) data and derived football insights.

🧍 Clubs, Scouts, Agents & Fans

Whether you're scouting, negotiating, or dreaming — SquadGen builds the most balanced squad possible, tailored to your exact criteria. Start with a few players, and uncover perfect fits in seconds.

Want to join the adventure?


Join us in shaping the next frontier of AI-driven football analytics. Early access & partnerships now open.