Score
SN44Turns any camera into a smart camera using AI-powered computer vision
Decentralized computer vision for sports. Miners process live football match footage to track players, detect the ball, analyze formations, and generate automated highlights. The same technology that billion-dollar sports analytics firms sell for premium prices, built through open competition.
// Making every camera intelligent.
Score is a computer vision network built specifically for sports analytics. Miners process video feeds from football matches and extract structured data: player positions, ball movement, formation patterns, and gameplay statistics. Validators verify the accuracy of this tracking against ground truth data.
The simple version: Imagine AI watching a football match and generating real-time stats on every player's position, speed, distance covered, and tactical movements. That's Score: turning raw video into structured sports intelligence.
Centralized equivalent: Think Hawk-Eye (Premier League), Second Spectrum (NBA), or Stats Perform, but built through decentralized competition rather than exclusive corporate contracts.
How it works:
- Miners deploy computer vision models to detect and track players, analyze ball movement, and extract gameplay statistics from video feeds. They handle various scenarios including player occlusion, fast movements, and different camera perspectives.
- Validators distribute video segments, evaluate tracking precision against ground truth data, and score miners based on accuracy, processing speed, and ability to handle complex sports scenarios.
- The problem it solves: Professional sports analytics costs millions per season. Only the top leagues can afford tracking systems like Hawk-Eye. Amateur leagues, lower divisions, and individual coaches have no access to this technology.
- The opportunity: The global sports analytics market is projected to reach $8 billion by 2030. Most sports content is still unanalyzed video. Automated analysis unlocks value at every level.
- The Bittensor advantage: Decentralized processing means any camera can become intelligent. Score's roadmap includes mobile integration, meaning a smartphone could analyze an amateur match using the same technology as the Premier League.
- Traction signals: Co-founded by Max Sebti (CEO), Tim Kalic (CTO), Nigel Grant (CBO), and Dr. Peter Cotton. 4,757 holders (well-distributed with Gini 0.551 and HHI 0.015, the lowest concentration in our coverage). 249 commits, 21 GitHub stars.
Category: Sports Analytics and Prediction | Centralized Competitor: Hawk-Eye (Sony), Second Spectrum (Genius Sports), Stats Perform, Catapult Sports
Score targets one of the few real-world markets where computer vision has proven commercial value. Sports tracking is a solved problem technically, but an unsolved problem economically: it's too expensive for most leagues. Score's decentralized approach could change the economics.
Mechanism:
Validators distribute video segments to miners and evaluate the results against ground truth data. Scoring accounts for tracking precision across complex scenarios: player occlusion (when bodies overlap), fast movements (shots, sprints), and varying camera angles. This multi-scenario evaluation prevents miners from overfitting to easy cases.
The founding team brings genuine sports tech experience. Max Sebti (CEO), Tim Kalic (CTO), Nigel Grant (CBO), and Dr. Peter Cotton represent a mix of business and technical depth unusual for Bittensor subnets.
The codebase has 249 commits across 4 contributors, though development has been inactive since December 2025 (last commit was a validator hotfix). This gap is concerning but may reflect a transition to internal development or a product-focused phase.
Market position is notable. At 136,285 TAO market cap, Score is one of the larger subnets. The holder distribution is the most egalitarian we've covered: Gini 0.551 and HHI 0.015 indicate extremely distributed ownership across 4,757 holders. Root proportion of 0.171 confirms organic demand.
Net 7-day outflow of -745 TAO suggests some profit-taking after the 68% 90-day run. The 30-day decline of 12% indicates a cooling period. Unrealized gains likely remain significant across the broad holder base.
The roadmap targets multi-sport expansion (basketball, tennis), tactical analysis and formation recognition, real-time broadcasting overlays, and mobile integration for amateur sports. The mobile angle is the killer app: turning any smartphone into a professional-grade sports analytics tool.
- Development inactive: No commits since December 2025. Four months without code changes is concerning for any software project.
- Current outflows: -745 TAO net 7-day flow and -12% 30-day price decline suggest cooling interest.
- Sports-specific risk: Revenue depends on sports broadcasting and analytics adoption. Seasonal patterns (off-season) could affect usage.
- Competitive moat: Hawk-Eye and Second Spectrum have exclusive contracts with major leagues. Score needs to win at the amateur and semi-professional level first.
