SCORE VISION (Bittensor SN44) — Full Analyst Report

SubnetAIQ Intelligence | August 9, 2026 Rating: BUY | Conviction Score: 60/100 | Momentum: NEUTRAL (-1.1) | Verified by independent AI review

EXECUTIVE SUMMARY

Score Vision is Bittensor Subnet 44, a decentralized computer vision framework targeting Game State Recognition (GSR) in football (soccer). The platform uses a miner-validator architecture to process match footage — detecting and tracking players, ball, referees, and goalkeepers in real-time — at a fraction of the cost of traditional sports analytics vendors ($10-55/min for manual annotation vs. Score's decentralized approach targeting 10-100x cost reduction).

Score currently ranks in the mid-tier of Bittensor subnets with 70,270 TAO staked (~$14.3M at $203/TAO), producing 0.1273 TAO/day in emissions. The token has appreciated ~37% from its June lows near $0.031 to the current $0.043, though momentum has flattened to NEUTRAL (-1.1). Development activity is steady but not aggressive — a small team (primarily 1-2 contributors) continues refining keypoint detection, validation stability, and containerized infrastructure.

The $600B football industry is the beachhead, with planned expansion into basketball, tennis, security surveillance, and retail analytics. The tech thesis — decentralized CV pipelines undercutting Hawk-Eye/Second Spectrum pricing — is directionally sound but unproven commercially. No disclosed revenue, customers, or formal partnerships yet. We rate SN44 a BUY for dTAO staking on strong on-chain health, reasonable valuation, and an attractive niche, but flag execution risk given the lean team and lack of commercial traction.

12-Month Base Target: ~$11.83 USD (0.058 TAO × $204 = +35% from current ~$8.77). Expected value at 12 months: 0.056 TAO (~$11.42 USD, +30%).

1. COMPANY OVERVIEW

What Score Does

Score Vision builds a decentralized pipeline for real-time sports video analysis. The core workflow:

1. Video Input: Match footage (broadcast or grassroots camera) is fed into the network
2. Miners process video frames using object detection and tracking models, generating standardized Game State Recognition outputs — bounding boxes for players, ball, referees, goalkeepers with team assignments and tracking IDs
3. Validators verify miner outputs using a two-stage lightweight validation:
4. Stage 1: Frame filtering with pitch/keypoint detection to select informative frames
5. Stage 2: Semantic verification using CLIP-based vision-language models (VLMs) to confirm object identification accuracy
6. Scoring: Performance is measured via GS-HOTA (Game State Higher Order Tracking Accuracy), combining detection accuracy with association/tracking consistency across frames

The key innovation is that validation is dramatically cheaper than mining — validators don't need to re-run full inference. They filter frames, check keypoints, and run lightweight CLIP checks, reducing validator compute costs while maintaining quality assurance.

Target Market

Score targets the $600 billion football industry, specifically the video analysis segment: - Current cost: Manual annotation runs $10-55 per minute of footage - Score's target: 10-100x cost reduction through decentralized compute - Addressable use cases: Tactical analysis, player scouting, match statistics, broadcast enhancement, grassroots coaching

Roadmap

Note: Phase 3/4 timelines appear to have slipped. As of August 2026, development activity still focuses on keypoint refinement and validation stability rather than action spotting or multi-sport expansion.

2. TEAM ANALYSIS

Known Contributors

Assessment

This is a lean team — possibly 2-3 core developers. The GitHub organization is "score-technologies" suggesting a formal company entity, and the professional email domain (wearescore.com) and polished documentation suggest this isn't a solo hobby project. However:

• No named founders/CEO identified in public materials
• No LinkedIn presence found for Score Vision
• No disclosed funding rounds or investors
• No disclosed advisors or board members

Risk: The team opacity is a yellow flag. Most successful Bittensor subnets have identifiable leadership (Chutes has John Durbin, NOVA has Micaela Bazo, etc.). Score's anonymous-ish team makes it harder to assess execution capability.

3. TECHNOLOGY DEEP DIVE

Architecture

[Video Stream] --> [Miners: Object Detection + Tracking]
                         |
                   [Standardized GSR Output]
                         |
              [Validators: Frame Filter + CLIP Verify]
                         |
                    [GS-HOTA Score]
                         |
                  [Incentive Distribution]

ML Models Used

Lightweight Validation Innovation

Score's key technical contribution is reducing validation costs through:

1. Pitch detection filtering: Only evaluating frames where the pitch is clearly visible
2. Keypoint validation: Checking consistency of player/ball positions against pitch geometry
3. CLIP-based semantic checks: Using a lightweight VLM to confirm detections without running full inference
4. Global scoring: Evaluating stability, plausibility, and reprojection error across frame sequences

This means validators can verify quality without GPU-heavy re-inference — a practical advantage for maintaining a validator fleet on modest hardware.

Recent Development Activity (GitHub)

Assessment: Development is steady but not aggressive. The team is refining existing capabilities rather than shipping major new features. The Phase 3 deliverables (action spotting, match captioning) appear delayed. Only 13 merged PRs total and 1 open issue suggest a small but focused development effort.

4. COMPETITIVE LANDSCAPE

Sports AI/CV Incumbents

Score's Positioning

Score is not competing head-on with Hawk-Eye or Second Spectrum. Those companies sell integrated hardware+software solutions to top-tier leagues. Score's play is different:

1. Cost disruption: Targeting the 99% of football that CAN'T afford Hawk-Eye ($500K-$1M+ per venue). Youth academies, lower leagues, grassroots — any footage from a single camera
2. Decentralized compute: No centralized GPU infrastructure needed; the Bittensor miner network provides processing
3. Open architecture: MIT-licensed, API-first approach vs. proprietary lock-in

Score's realistic addressable market is: - Youth/amateur football (millions of teams globally) - Lower-division professional leagues (cost-sensitive) - Individual coaches/analysts wanting cheap video breakdown - Content creators and media wanting automated highlights

This is a valid niche but unproven. The question is whether Score can bridge from "interesting Bittensor subnet" to "product people actually pay for."

5. ON-CHAIN DATA ANALYSIS

Live Metrics (August 9, 2026)

Conviction Score Breakdown (60/100 — BUY)

Standout: Risk and Price Sustainability both at maximum scores. This subnet is stable — it won't get deregistered and its price structure is sound. The weakness is in development velocity (8/20).

Momentum Analysis (-1.1 — NEUTRAL)

Interpretation: Price is modestly positive but volume is declining. The positive consistency (3 of 4 periods with positive price action) is encouraging, but falling volume suggests the recent rally may be losing steam. This is a consolidation pattern, not a momentum buy.

Harmonic Pattern Analysis

No active harmonic patterns detected for SN44. The scanner did not flag any completed or forming Butterfly, Crab, or Cypher patterns. This is neutral — no screaming technical entry or exit signal.

Early Mover Signal

SN44 was NOT flagged in the early mover cache. This subnet is established (launched Jan 2025), not an early-stage discovery play.

Price History (57-Day OHLC Analysis)

52-day range: 0.027498 — 0.046272 (68% range) Current vs. range: Trading at 80th percentile of range 30-day change: +9.2% (per conviction data) 7-day change: +16.0% 1-day change: +4.0%

The chart shows a clear V-bottom from mid-July lows (0.027-0.028) with a strong rally through August. Price is now near the upper end of its range but still below the all-time high of 0.046.

6. VALUATION MODEL

TAO-Denominated Valuation

USD-Denominated Valuation (TAO = $203)

Comparable Analysis

Score commands significantly more capital than most CV/video subnets, reflecting its real-world application and stable on-chain metrics.

Bear / Base / Bull Scenarios

Assumptions: TAO price held constant at $203 for token price targets. USD values scale linearly with TAO price.

BEAR CASE (25% probability)

• Team remains anonymous and small; development velocity doesn't accelerate
• Phase 3/4 roadmap continues to slip; no multi-sport expansion
• No commercial customers or revenue by mid-2027
• Larger competitors (Hawk-Eye, Stats Perform) launch cheap API-first products
• Capital outflow as stakers rotate to higher-performing subnets

BASE CASE (50% probability)

• Team maintains steady development; validator containerization completed
• Action spotting and match captioning ship by Q1 2027
• 1-2 pilot partnerships with lower-division football leagues or analytics companies
• No material revenue but growing ecosystem credibility
• Staking inflows keep pace with emission dilution

BULL CASE (25% probability)

• Named team emerges with credible sports tech backgrounds
• Major partnership announced (e.g., FIFA Grassroots, a national football association, or a sports data provider)
• Multi-sport expansion (basketball, tennis) ships successfully
• Revenue-generating API business launched ($100K+ ARR)
• Bittensor ecosystem tailwind — TAO price appreciation amplifies returns

EXPECTED VALUE (probability-weighted)

7. RISK FACTORS

HIGH RISK

1. Team Opacity: No named founders, no disclosed funding, no LinkedIn presence. If the team walks away, there's no one to hold accountable. This is the #1 risk factor.

2. Zero Commercial Traction: No disclosed customers, revenue, partnerships, or pilot programs. The technology exists in a Bittensor vacuum — miners process video for TAO rewards, but nobody is paying USD for the output yet.

3. Roadmap Slippage: Phase 3 (action spotting, captioning) was targeted for Q2-Q3 2025 and appears undelivered as of August 2026. Phase 4 (multi-sport, APIs) is similarly delayed.

MEDIUM RISK

1. Small Development Team: 1-2 active GitHub contributors creates bus-factor risk. If DataAndMike stops contributing, development effectively halts.

2. Competitive Response: Sports AI is a hot market. If Hawk-Eye or Stats Perform launches a low-cost API tier, Score's cost-disruption thesis gets challenged by incumbents with existing relationships and data.

3. Single-Sport Concentration: 100% football focus. If the football analytics market proves smaller or more fragmented than expected, growth is capped.

4. Bittensor-Specific Risk: All value accrues through TAO emissions. If Bittensor's dTAO model changes unfavorably or TAO price crashes, Score's economic model breaks regardless of technology quality.

LOW RISK

1. Deregistration Risk: Conviction score gives 15/15 for risk — this subnet is stable and not at risk of being deregistered.

2. Price Sustainability: Perfect 5/5 score — the price structure is sound with balanced buy/sell flow (238 buys vs 210 sells in 24h).

8. SENTIMENT & SOCIAL ANALYSIS

Twitter/X (@webuildscore)

Unable to access Score Vision's X feed directly due to platform restrictions. The account exists and is referenced in GitHub documentation, suggesting active social presence.

Community Signals

• Discord: Active community channel linked from GitHub
• GitHub Issues: Only 1 open issue (migration-related), suggesting either a clean codebase or low community engagement
• GitHub Stars: 27 — modest community interest, not viral
• Forks: 21 — indicates some miner/validator operator engagement

YouTube/Media Coverage

No Score Vision-specific YouTube interviews or deep-dive content found. This is a gap — most successful subnets have founder interviews on crypto/Bittensor YouTube channels.

Commercial Activity Update (August 9, 2026)

Additional research surfaced traction beyond the initial report: - 124 NVIDIA Jetson deployments — edge-compute hardware deployed in the field, indicating real physical infrastructure beyond software-only - PwC France/Maghreb alliance — partnership with PricewaterhouseCoopers France and North Africa region; adds enterprise credibility and potential distribution into professional sports organizations - World Cup fantasy app — a consumer-facing application built on Score's computer vision output, demonstrating a path from B2B analytics to B2C product - GitHub remains active — continued commits confirm development has not stalled

9. ANALYST THESIS

Bull Case in One Sentence

Score Vision is building the "AWS of sports video analytics" — a decentralized compute layer that makes real-time football analysis accessible to the 99% of teams that can't afford Hawk-Eye, with a sound technical approach (CLIP-validated lightweight inference) and rock-solid on-chain fundamentals.

Bear Case in One Sentence

An anonymous 2-person team processing football video for TAO rewards with zero paying customers, a slipping roadmap, and no evidence that anyone outside the Bittensor ecosystem wants this product.

Our View

BUY for dTAO staking allocation. Score occupies an attractive niche at the intersection of computer vision and sports — a massive, underserved market. The on-chain metrics are strong (70K TAO staked, 21% APY, perfect risk/sustainability scores), and the 37% rally from July lows shows market confidence. The 60/100 conviction score places it in the upper half of Bittensor subnets.

However, this is a position-size-with-caution BUY: - Keep allocation modest (2-5% of dTAO portfolio) until team identity is disclosed - Watch for Phase 3 deliverables (action spotting) as the key near-term catalyst - Monitor GitHub commit frequency — any sustained drop signals abandonment risk - Re-evaluate if momentum drops below -20 (currently -1.1)

Position Sizing for $10K Challenge

Risk/Reward ratio: 1.8x — acceptable but not compelling. The asymmetry improves significantly if the team doxxes or announces a partnership.

10. MONITORING CHECKLIST

• [ ] GitHub commit frequency (weekly check)
• [ ] Team doxxing / founder interviews
• [ ] Partnership announcements (football leagues, analytics companies)
• [ ] Phase 3 delivery: action spotting, match captioning
• [ ] Multi-sport expansion announcements
• [ ] TAO staked trend (currently 70,270 — watch for outflows)
• [ ] Momentum score changes (currently NEUTRAL at -1.1)
• [ ] Competitor moves from Hawk-Eye, Stats Perform, Statsbomb

Report prepared by SubnetAIQ Intelligence Engine. On-chain data as of block 8,809,678 (August 9, 2026). Not financial advice. Always verify on-chain data independently before making staking decisions.