Player
Train with your Personal AI Coach, build your Player Twin and track development.
Open SSOT →
Global AI Football Platform
Discover Your Super-Self,
Discover Your Super-Self,
Develop Your Super-Talent.
Your Personal AI Coach helps you understand and improve. Your AI-augmented Human Coach turns intelligence into better football development. One SSOT identity connects the journey.
6 role experiences1 shared SSOT identityDigital Twin intelligenceHuman validated
AI-AUGMENTED HUMAN COACH
Coach. Improve. Grow.
AI helps you learn faster, coach better, improve player outcomes, build reputation and grow recurring coaching income.
Coach LevelL4
Players23
Rating4.8★
AI COACHING PARTNER · CAREER FLYWHEEL
SuperStars of Tomorrow
FIFA Basic Certified · Single Camera Broadcast Tracking System
View details →
CHOOSE YOUR EXPERIENCE
One platform. Six simple ways in.
Coach
Become a better coach, build better players, grow your reputation and earn more with AI.
Start Coach Journey →Parent / Guardian
Protect, approve and support a young player’s development and opportunities.
Club / Academy
Connect teams, coaches, capture, player development and academy intelligence.
Scout / Agent
Discover consented talent through verified evidence and comparable player profiles.
League / Federation
Connect grassroots development, programs, research and interoperable services.
AI-AUGMENTED COACH LIFECYCLE
Become a better coach. Build better players. Grow your career.
One intelligent system connects onboarding, AI-coaching skill development, player outcomes, reputation, discovery and earnings. The coach does not just use AI — the coach continuously becomes better with AI.
Joinverify identity
LearnAI coaching
Coachhuman + AI
My Playersplayer outcomes
Growreputation
Earnsubscriptions
Level Up ↻L0 → L9
Active players23+7 this month
Player retention82%+5 pts
Coach rating4.8★top quartile
Monthly earnings$612+$94
COACH DIGITAL TWIN
Level 4 · Trusted Developing Coach
Your AI Coaching Partner continuously recommends the shortest path to the next level.
L446 / 100
v0.17 · TWIN MATCHING ENGINE
Player Digital Twin ↔ Coach Digital Twin
The platform should not only rank players and coaches separately. It should learn which coach is most likely to help a particular player improve next — by age, development need, coaching specialty, evidence of outcomes and trust.
PLAYER DIGITAL TWIN
Who is this player becoming?
Skill Passport · age band · position · development trajectory · recent weaknesses · learning response · goals.
⇄AI MATCH + LEARN
COACH DIGITAL TWIN
Who can help this player next?
Credentials · experience · AI-coaching expertise · player outcomes · reputation · specialties · availability.
FIFA QUALITY PROGRAMME · BROADCAST TRACKING SYSTEM
Officially Certified for Single Camera Broadcast Tracking System
SuperStars of Tomorrow is officially certified under the FIFA Quality Programme as a Broadcast Tracking System. The certification is awarded specifically for a Single Camera Broadcast Tracking System (TM Broadcast EPTS | 2025), validating tracking intelligence directly from single-viewpoint match broadcasts.
CERTIFIED PRODUCT
SuperStars of Tomorrow
Officially certified product under the FIFA Quality Programme for EPTS Broadcast Tracking Systems.
CERTIFIED DATE
11 Aug 2026
Official certification start date under FIFA Quality Programme.
SYSTEM SPECIFICATION
View on FIFA website →
Broadcast System
Single Camera Broadcast Tracking System
Certified for single camera broadcast video tracking under FIFA EPTS test standards.
v0.18 · AI TALENT & DEVELOPMENT GRAPH
From one great match to an entire football opportunity network.
The next layer connects each Player and Coach Digital Twin to academies, clubs, scouts and verified opportunities. The platform learns not only who can help a player improve, but also what opportunity may fit next — while keeping evidence, consent and human judgment visible.
Develop: connect player needs to the right coach and training path.
Prove: carry verified Skill Passport evidence and progress history forward.
Discover: surface academy, club and scout interest based on fit — not hype alone.
Advance: turn trusted development evidence into the next football opportunity.
AI Coachdevelopment fit · outcomes
Academyprogram fit · age band
PLAYER TWINskills · trajectory · goals
Clubteam pathway · trial
Scoutverified discovery
Opportunitytrial · camp · pathway
Trusted Nodes1,248
Live Matches286
Opportunities47
v0.19 · OPPORTUNITY LIFECYCLE
Turn discovery into a trusted pathway — not just a lead.
The Talent Graph can surface a possible next step. The Opportunity Lifecycle manages what happens after that: eligibility, consent, application, coach endorsement, academy/scout review, trial and outcome — with evidence and human decisions visible at every stage.
Qualify: check age, geography, skill evidence and program requirements.
Prepare: assemble a verified Player Passport, clips and coach endorsement.
Progress: track review, invitation, trial and outcome without losing context.
Learn: feed outcomes back into Player Twin, Coach Twin and future matching.
GABRIEL · U15 TECHNICAL TRIAL
One opportunity. Seven managed stages.
1 · Discover87% explainable fit surfaced
2 · Eligibleage + role + evidence checked
3 · Applyguardian approval + dossier
4 · Endorsecoach evidence & note
5 · Reviewacademy / scout decision
6 · Trial3-day evaluation
7 · Outcomedecision + learning loop
Application readiness82%
Verified evidence9 items
Next actionGuardian approve
v0.33 · PLAYER CAREER PATHWAY MANAGER
Manage many opportunities as one development career.
A player should not chase every attractive trial. v0.33 compares simultaneous opportunities against the Player Digital Twin, development priorities, timing, evidence, family constraints and long-term goals — then turns them into a coherent 1-, 3- and 5-year pathway.
1 Year · develop & prove3 Years · advance pathway5 Years · build optionality
GABRIEL · CAREER OPTIONS
Prioritize the next step without losing the long view.
U15 Technical Trialbest immediate development fit
87%
1Y
Barcelona Youth Labtechnical camp · high learning value
82%
1Y
European Academy Pathwayprogression option after evidence growth
76%
3Y
Primary goalU15 → elite academy
Current priorityDribbling + match evidence
AI recommendationTrial first, camp second
ONE SHARED FOUNDATION
Different experiences. One football intelligence layer.
Keep complexity underneath the product. Users only need to know what happens next; the platform handles identity, evidence, validation, permissions and Digital Twins across roles.
Real Footballtrain · play · capture
Personal AI Coachanalyze · learn · recommend
Human Coachvalidate · interpret · develop
Trusted Ecosystemshare · discover · progress
v0.33 · SUPER-TALENT DEVELOPMENT SCIENCE
Discover → Hypothesize → Intervene → Measure → Learn
Turn the Player Digital Twin into a longitudinal development model. AI proposes an explainable development hypothesis; the human coach validates or modifies the intervention; evidence updates both the Player Twin and Coach Twin.
Super-Self Digital Twin
Current capability is not destiny. Track state, development velocity, and uncertainty over time.
Current Development Hypothesis
Bottleneck: First touch under speed & pressure
Confidence: 72%
3-week intervention: directional first touch + pressure receiving + weak-foot repetitions
Development Response Function
The scientific question is not only “How good is the player today?” but “How rapidly and reliably does this player improve under a specific intervention and coach?”
Current Capability: 72Development Velocity: +4.0/cycleEvidence Cycles: 3Model Confidence: 72%
v0.33 · Global Football Intelligence Network
One Football Intelligence. Ten Languages.
Players and coaches can work in their own language while the underlying football evidence stays consistent. SSOT separates the language layer from the football-intelligence ontology so development evidence can travel across borders without changing its meaning.
Language-independent football ontology
Same evidence. Local language.
A pass remains the same football event whether the player sees Portuguese, the coach works in Spanish, and a scout reviews the evidence in English.
PASSINGScan → recognize options → decide → pass → receiver response → team state
FCIAwareness, decision quality, anticipation and adaptation
FSITeammate awareness, cooperation, combination and team contribution
FEIObservable response to pressure, mistakes and feedback
PLAYER TWINCapability, development velocity, evidence confidence and history
Global learning loop
Local training → shared evidence model → multilingual AI interpretation → human-coach validation → player response → updated intelligence network.
Cross-language AI coaching demo
Coach globally. Learn locally.
The platform can preserve the football meaning while adapting the explanation to each user's language.
PLAYER · PORTUGUESE
“Quero melhorar meu primeiro toque e passar mais rápido sob pressão.”
→
COACH · SPANISH
“Tu primer toque es estable. El siguiente objetivo es escanear antes de recibir y reducir el tiempo hasta el pase.”
AI EVIDENCE · SHARED
First-touch control 74 · scan-before-receive 61 · release time 1.42s · evidence confidence 82%
→
PLAYER FEEDBACK · LOCALIZED
“Ponto forte: controle. Próximo foco: olhar antes de receber e soltar a bola mais cedo.”
Player Twinlocal evidence
Coach Twinvalidated expertise
Football Ontologyshared meaning
AI Translationlocalized interaction
Evidence Graphcross-border learning
Opportunity Networkglobal pathways
Localization changes presentation, not the underlying evidence. Production deployment should use football-specific terminology review, native-language QA, safeguarding review for youth content, and locale-aware consent/privacy text.
PLAYER · AI COACH · HUMAN COACH
Three learners. One continuous development cycle.
The AI does not replace the coach. The Player, AI Coach and AI-Augmented Human Coach learn from the same evidence and outcomes.
PLAYERGoal · response · engagement · development
↔
AI COACHMeasure · remember · recommend · learn
↔
HUMAN COACHContext · judgment · motivate · ACCEPT / MODIFY / REJECT
Context → AI Proposal → Human Judgment → Intervention → Player Response → Outcome → Lesson → Better next cycle
One Player journey — clustered in one place.
Discover → Train → Measure → Develop → Be Seen
AI Personal Coach, five essential drills, 1/2/3-camera evidence, Football Intelligence, Development Twin, Player Passport and opportunity pathways become one coherent story.
AI Personal Coach5 DrillsMulti-CameraPlayer TwinPlayer PassportOpportunities
Customer value
Know yourself, see progress, understand why, and know what to do next.
Business valueAcquisition · engagement · IAP · coach demand · evidence density · marketplace demand.
All Coach capabilities — one coherent system.
Coach better → prove impact → grow professionally
Coach Agent, My Players, Triadic Learning, Training & Evidence, Development Impact, Coach Passport, Founding Coach OS, Marketplace, Network and Trust now belong together.
Coach AgentMy PlayersTriadic LearningEvidenceImpactCoach PassportFounding Coach OSMarketplace
Customer value
Combine human judgment with AI memory, evidence and recommendations.
Business valueSubscription · player acquisition · AI validation · marketplace supply · regional growth.
Four concise gateways — not four long homepage stories.
PARENT / GUARDIAN
Protect, understand and support the player's development.
CLUB / ACADEMY
Develop players and coaches as one evidence-based system.
SCOUT / AGENT
Discover trajectory and permissioned evidence, not only today's performance.
LEAGUE / FEDERATION
Improve the ecosystem through governed aggregate intelligence.
GLOBAL NETWORK
Five regions · Ten languages · One intelligence network.
Localization is developed with native-speaking coaches, while identity, evidence definitions and trust remain globally coherent.
AFRICALATAMSWITZERLAND / EUROPENORTH AMERICAASIA
NOW: FOUNDING COACH PILOT
BECOME A FOUNDING COACH8–10 intensive pilot coaches for Triadic Learning validation; a wider 20–30 localization cohort can cover all 10 languages with native reviewers.
IDENTITY & CONSENT
Role relationships, guardian consent and revocation.
EVIDENCE & PROVENANCE
Confidence, source and validation remain visible.
PASSPORTS & PERMISSIONS
One canonical Passport per role; multiple authorized views.
GLOBAL TRUST LAYER
Stakeholder → Agent → permissions → Twin/Data → Passport.
v0.33 · Community-Built Network Infrastructure
Build Locally. Connect Globally.
Create and lead your own AI-Augmented Football Network — powered by the Global AI Football Platform. Local leaders organize their communities; SSOT provides shared Passports, consent, Digital Twins, football ontology, scientific validation, multilingual intelligence and global interoperability.
Network Builder Lifecycle
1 · Define your networkCommunity, academy group, regional development network or league.
2 · Verify organization & leadershipOrganization Passport, administrators, safeguarding and authority.
3 · Invite the ecosystemClubs, academies, coaches, players, parents and scouts.
4 · Operate locallyLocal language, programs, permissions and development workflows.
5 · Validate evidenceCommon protocols, confidence and longitudinal quality.
6 · Scale regionallyAdd cities/countries and institutional partners.
7 · Connect globallyPortable Passports and permissioned benchmarks interoperate.
Create Your Network
My AI-Augmented Football Network
Regional Development Network · Africa · Grassroots & Player Development
Governance: community and development networks may be self-created. Official league, federation, confederation or governing-body status requires separate verification before it is presented as official.
Five strategic pilot regions
Five Regional Networks. Five Growth Laboratories. One Global Standard.
The Global AI-Augmented Football Network now starts with five strategic regional pilots: Africa, LATAM, Europe, North America and Asia. Each region has a different entry strategy while sharing the same Player/Coach/Organization Passports, guardian consent, football ontology, scientific validation and interoperability layer.
🇧🇷 🇦🇷 🇲🇽
LATAM Net
Brazil, Argentina and Mexico as the first connected national nodes for player/coach scale and football talent discovery.
🌍
Africa Net
Enable credible local football leaders to build country and regional networks with scalable institutional pathways.
🇨🇭 🇪🇺
Europe Net
Use Switzerland as the initial hub for validation, sports innovation and wider European football partnerships.
🇺🇸 🌎
NA Net
Anchor the network around AIBrain’s Silicon Valley global HQ and ongoing U.S. football/soccer partnership development.
🇯🇵 🇰🇷 🇨🇳 🌏
Asia Net
Tier 1: Japan + Korea · Strategic Scale: China · Expansion: Greater Asia.
Selected strategic pilot
LATAM AI-Augmented Football Network
Brazil · Argentina · Mexico
3Initial nodes
PT-BR · ES · ENLanguages
L1Network maturity
90dFounding pilot
Hypothesis: Mass adoption + multilingual player/coach growth
Entry: Players and coaches first; academies, scouts and clubs next
Regional autonomy + global interoperability: each network can develop its own partnerships, languages and operating priorities while retaining common SSOT trust, evidence and portability standards.
Regional Organization Passport
LATAM Net
Organization typeStrategic Regional Network
Regional leadershipVerified Network Builder(s)
Safeguarding / consentRequired
Official federation statusOnly if independently verified
Global interoperabilitySSOT standard
Benchmark participationPermissioned + validated
UNIFIED STAKEHOLDER WORKSPACES
Different roles. Clear ownership. One trusted football network.
Every stakeholder gets a coherent workspace, AI Agent, evidence view and permission boundary. A person with more than one role switches workspaces without creating another human identity.
COACH WorkspaceOwned by the AI-Augmented Human Coach
COMMON WORKSPACE GRAMMAR
Home → Agent → Activities → Evidence → Goals/KPIs → Passport/Dashboard → Network → Trust
The functions differ by stakeholder, but the navigation logic stays familiar. This lets a player understand what belongs to the coach, a parent understand what the coach may see, and an academy or federation consume only authorized aggregate evidence.
Trust rule: Stakeholder → Workspace → AI Agent → Permissions → Twin/Data → Passport. Private Player Twin data never becomes public Coach Passport content simply because it contributed to aggregate coach evidence.
ONE PASSPORT DESIGN SYSTEM
Same credential family. Different stakeholder identity.
SSOT mobile and the Global AI Football Platform render the same underlying Passport object. Color identifies the role; geometry, verification and credential grammar stay consistent.
PLAYER PASSPORT · SSOT / GLOBAL
⚽
GABRIEL FERREIRA
PLAYER TWINDRI49
JUG99
PAS83
TOTAL82
COACH PASSPORT · SSOT / GLOBAL
🧑🏫
LUCAS OLIVEIRA
FOUNDING COACHAICLL3
EC78
EBB2
PDI+6.8
One Coach Passport
SSOT uses a compact mobile view; the Global Platform uses an expanded professional view. Passport ID, verification, evidence and status remain the same.
Coach headline evidence
Human Qualification · Primary Role · AICL · Evidence Confidence (EC) · Evidence Breadth (EB) · Player Development Impact (PDI) · Specialty.
Permissioned expansion
Overview · Development Impact · Human–AI Judgment · Players & Cohorts · Qualifications & Verification · Marketplace.
No star-player halo
Player volume or pre-existing player quality never becomes the coach's headline quality score. Development impact and evidence confidence stay separately visible.
PASSPORT COLOR LANGUAGE
Recognize the stakeholder instantly.
Player and Coach are the first canonical implementations; the remaining colors are reserved for the wider credential family.
PLAYER · REDPlayer Twin · development evidence
COACH · INDIGO / BLUECoach Twin · AICL · EC · EB · PDI
SCOUT / AGENT · TEALPermissioned discovery and opportunity
CLUB / ACADEMY · GREENOrganization identity and development system
LEAGUE / FEDERATION · NAVYGovernance · aggregate ecosystem intelligence
v0.33 · Global AI-Augmented Football Network
Connect the People Who Develop Football.
The Global AI Football Platform powers a human-centered network: players develop, parents control permissions, coaches augment expertise, scouts and agents discover permissioned evidence, academies organize development, and leagues/federations strengthen the ecosystem. AI expands capability; people remain responsible for football decisions.
Connected network
Six roles. One trusted football journey.
⚽ Playerdevelop · prove · progress
🛡 Parent / Guardianprotect · approve · support
🧠 AI-Augmented Coachcoach · validate · grow
🔎 Scout / Agentdiscover · evaluate · advance
🏟 Club / Academydevelop · organize · validate
🌐 League / Federationconnect · develop · benchmark
Network principle
Technology is AI-powered. Human coaches, scouts and the football ecosystem are AI-augmented.
Four partnership programs
Join where you create value.
AI-Augmented Coach PartnerLearn better · coach better · build reputation · earn more
AI-Powered Club & Academy PartnerPlayers · coaches · teams · development intelligence
AI-Augmented Scout & Agent PartnerDiscover more · see evidence · make better human decisions
AI-Powered League & Federation PartnerConnect clubs · develop talent · strengthen ecosystems
Global AI-Augmented Football Network
Player Passport + Coach Passport + Guardian Consent + Digital Twins + Scientific Validation + Multilingual Football Intelligence.
Global Football Development Benchmark · simulation
Scale the Evidence — Then Validate It.
A larger membership can increase statistical power and enable finer reference cohorts, but scale alone does not guarantee precision. SSOT should benchmark only standardized, representative, longitudinal and scientifically validated evidence.
How the benchmark could mature
Complementary to academies: SSOT can provide broad standardized longitudinal evidence; professional academies provide deep physical, tactical, medical and expert observation. Connecting the two can improve discovery and validation.
Example · U14 midfielder
1.8kComparable cohort
68%Benchmark confidence
2Validated regions
4 moLongitudinal history
Passing execution76 · 82nd percentile
FCI · scanning71 · 78th percentile
FSI · combination play74 · 84th percentile
Development velocity+5.1 · 88th percentile
Suggested actionHuman scout review — not selection
Scientific guardrail: all percentile values here are simulated UI examples, not validated SSOT population benchmarks.
v0.33 · Scientific Validation Layer
Measure the Measurement.
Before SSOT uses FCI, FSI, FEI or Super-Talent Development metrics for meaningful decisions, the platform should validate whether the measurements are repeatable, consistent across human coaches, stable across age/context differences, and predictive of observed improvement rather than merely descriptive.
Validation framework
Six scientific checks before trust
Each metric receives its own evidence-quality profile. A score can be visible while still carrying uncertainty and validation status.
1. Repeatability
Does the same player produce similar results under comparable conditions?
Does the same player produce similar results under comparable conditions?
TEST
2. Inter-rater agreement
Do qualified coaches agree on the same observable behavior?
Do qualified coaches agree on the same observable behavior?
TEST
3. Age / context normalization
Are comparisons fair across age, position, drill and game context?
Are comparisons fair across age, position, drill and game context?
CALIBRATE
4. Construct validity
Does the metric actually represent the football behavior we claim?
Does the metric actually represent the football behavior we claim?
VALIDATE
5. Longitudinal sensitivity
Can it detect meaningful development over weeks/months?
Can it detect meaningful development over weeks/months?
TRACK
6. Predictive usefulness
Does change in the metric relate to future observed development?
Does change in the metric relate to future observed development?
LATER
Interactive validation example
FCI · Scan-before-receive
Test whether scan-before-receive can be observed reliably across repeated sessions and coaches before using it in development recommendations.
0.84Repeatability
0.79Coach agreement
72%Context coverage
81%Evidence confidence
Validation interpretation
Promising, not yet a universal benchmark.
Use this metric for within-player development tracking first. Avoid cross-player ranking until age, position, camera setup and drill-context normalization are validated.
Scientific rule: a low-confidence or poorly validated metric should not silently become a high-stakes recommendation. SSOT should expose confidence, provenance and validation status alongside the result.
v0.33 · AICL Scoring Lab
Measure AI-Augmented Coaching Without Replacing Human Coaching Judgment.
AICL is the validated AI-Augmented Coaching Level (L0–L9). AICS is the underlying 0–100 score. The model separates human coaching qualification from demonstrated AI-augmented capability and requires score, evidence volume, dimension floors and trust gates before a level is validated.
COACH PASSPORT · PROTOTYPE
Alex Morgan · Founding Coach #017
Human Qualification: National B Licence · Football / Futsal · English / Spanish
L2
Current ValidatedAICS 69Evidence Confidence 61%Capability projection L5Highest achieved L2
A · AI Coaching Competence
68
J · Human Coaching Judgment
72
D · Player Development Impact
58
E · Evidence & Scientific Practice
64
T · Trust & Safeguarding
96
AICS = 20%A + 25%J + 25%D + 20%E + 10%T. Human Judgment + Player Development carry 50% of the score so software usage alone cannot create a high level.
Validated level logic
Level = minimum of four gates.
1 · Score thresholdAICS determines the numerical capability band.
2 · Evidence thresholdPlayers, sessions and longitudinal duration limit validated level.
3 · Dimension floorHigh AI literacy cannot compensate for weak player impact or judgment.
4 · Trust gateMaterial safeguarding or integrity failure freezes/suspends progression.
Why this coach is L2, not L5
AICS 69 numerically projects to L5, but only 6 players, 12 AI-assisted sessions and 70 days of evidence have been observed. Evidence currently validates only L2.
- Next evidence gate: 8 players
- Next session gate: 15 validated sessions
- Next duration gate: 90 days
- Human review required for L3
L0 → L9 validation ladder
Progressively harder—not ten equal bands.
| Level | AICS | Min players | Min sessions | Evidence duration | Validation | Dimension floor |
|---|---|---|---|---|---|---|
| L0 | Insufficient evidence | 0 | 0 | None | Baseline | — |
| L1 | 30–39 | 3 | 3 | 30d | Gate check | — |
| L2 | 40–49* | 5 | 8 | 60d | Sampled | — |
| L3 | 50–59 | 8 | 15 | 90d | Human review | All ≥40 |
| L4 | 60–67 | 12 | 30 | 6mo | Human review | All ≥45 |
| L5 | 68–75 | 20 | 60 | 9mo | Panel/sample | All ≥55 |
| L6 | 76–82 | 30 | 100 | 12mo | Panel | All ≥60 |
| L7 | 83–88 | 40 | 175 | 18mo | Expert validation | All ≥70 |
| L8 | 89–94 | 60 | 300 | 24mo | Independent component | All ≥78 |
| L9 | 95–100 | 100+ | 500+ | 36mo+ | Independent board | All ≥85 |
*A coach may have an AICS above the nominal score band yet remain at a lower validated level because evidence or other gates constrain progression. Thresholds are provisional research values for the First 100 Validation Cohort.
Coach Contribution Record
AI recommends. Human decides. Outcome teaches the Twin.
Passing Session #12 · AI: increase one-touch combination drillCoach modified: reduced speed, added scanning cue for two developing players. Outcome: completion +7%, scanning events +11%.
Finishing Session #9 · AI: repeat left-foot patternCoach rejected: player fatigue and confidence signal indicated overload. Outcome: recovery session; no injury concern; next-session accuracy +5%.
Decision Session #7 · AI: narrow-space rondoCoach accepted with contextual adjustment. Outcome: faster release time and better off-ball movement.
Judgment is not “AI acceptance rate.” Appropriate acceptance, modification and rejection are all positive when subsequent evidence supports the decision.
Next-level readiness
Tell the coach exactly what is missing.
6 / 8Players
12 / 15Sessions
70 / 90Days
61%Evidence confidence
Ready
AICS exceeds L3 threshold. All five dimensions are above 40. Trust gate is green.
Still required
2 more players, 3 sessions, 20 days of longitudinal evidence and human validation.
Current vs. highest achieved
L2Current Validated
L2Highest Achieved
Future versions can allow a current level to decline when evidence becomes stale while preserving the highest historical achievement in Coach Passport history.
Research status: AICL v0.1 is a provisional scoring model. The First 100 Validation Cohort should be used to calibrate weights, evidence minimums, inter-rater reliability, score stability, predictive validity, regional/language effects and football/futsal differences before any claim of scientific validation.
v0.33 · Live Marketplace Pilot
Turn Evidence-Based Matching Into Real Player–Coach Opportunity.
The marketplace now connects Player Agents, Guardian Agents, Coach Agents and Career/Scout Agents to a permissioned transaction workflow. Matching uses demonstrated development evidence, AICL, Evidence Confidence, Evidence Breadth, player fit, reputation and availability — not popularity alone.
Find My Coach · Interactive Pilot
Coach Kim · L5 · EC 86% · B2 Focused
U13 · Passing 88 · FCI 81 · Football + Futsal · Korean/English
Coach Sofia · L4 · EC 82% · B3 Established
U11–U15 · Technical 91 · FCI 78 · Football · Spanish/English
Coach Daniel · L6 · EC 91% · B4 Diverse
U15–U18 · FSI 90 · Team Development · French/English
BEST CURRENT MATCH
92%
Coach Kim
Strongest evidence fit for U13 Passing Intelligence, with high relevant development scores, adequate comparator quality and strong evidence confidence.
Marketplace signals: AICL · player-goal fit · development specialty · EC/EB · reputation · retention · availability.
Agent-Mediated Transaction Flow
Match → Consent → Offer → Subscription → Development Twin.
1 · Player Agent defines development needGoal, age, language, context and evidence profile.
2 · Marketplace Agent ranks coachesEvidence-based match; no single score decides.
3 · Guardian Agent approves request if minorCoach profile, price, data use and communication scope explained.
4 · Coach Agent reviews fit / capacityCoach may accept, decline or propose a different development plan.
5 · Subscription / trial activatesCommercial terms recorded; Founding economics applied where eligible.
6 · Development Twin createdGoals, baseline, sessions, evidence, consent and outcomes become the shared operating record.
Step 1 · Define player need
The Player Agent translates the player's goal into a structured matching request without exposing unnecessary private data.
Primary agentPlayer Agent
Human approvalPlayer / Guardian context
Data scopeMinimum necessary development fields
Audit eventMatching request created
Marketplace rule: AI may rank and recommend, but it cannot silently enroll a player, bypass guardian consent, or force a coach–player match.
Commercial Pilot · Illustrative Only
Test Value Before Optimizing Price.
Starter Trial
30 days1 coach · baseline + 3 AI-assisted sessions · Development Twin · player report.
Monthly Development
$29–49Illustrative monthly player subscription range for recurring AI-augmented coaching. Final pricing requires market testing.
Premium Specialist
$79–149Illustrative premium range for specialist/high-touch coaching, depending on market and coach profile.
No live purchase: these are prototype commercial scenarios. The pilot should validate willingness to pay, activation and retention before final pricing.
Founding Coach Economics & Marketplace Health
Reward Development, Retention and Trust.
Illustrative Revenue Split
Founding 300 · first 12 months80% Coach / 20% Platform
Planned standard marketplace70% / 30%
Referral rewardsOnly after qualified activation
Marketplace rankingEvidence + fit + trust
Commercial principle: avoid incentives that reward player volume alone. Better player development and retention should matter more than aggressive acquisition.
Marketplace Pilot KPIs
Match request → coach response≥70%
Coach response → trial start≥40%
Trial → paid conversion≥25%
30-day player retention≥70%
Coach retention≥80%
Guardian consent completion≥90%
Material safeguarding failures0
Decision rule: do not optimize revenue until the marketplace shows healthy activation, development evidence, repeat usage and trust.
v0.33 · Multi-Agent Football Network
Every Stakeholder Gets an Authorized AI Agent — Connected Through Trusted Digital Twins.
The network now moves from coach-centric automation to coordinated stakeholder automation. Each agent represents a defined human or organization, operates inside a permission envelope, communicates through shared Digital Twins, and leaves an auditable trail. Football remains the flagship sport; Futsal can use the same architecture when relevant without creating a separate platform.
Multi-Agent Network Map
Global Network AgentStandards · network health · scale gateGLOBAL
Regional Network AgentsAfrica · LATAM · Europe · NA · AsiaREGIONAL
League / Federation AgentPrograms · benchmarks · governanceINSTITUTION
↓
Academy / Club AgentTeams · coaches · development programsORG
Coach AgentPlan · review · develop · evidenceCOACH
Scout / Career AgentDiscover · match · request accessCAREER
↓
Guardian AgentConsent · privacy · opportunity approvalGUARDIAN
Player AgentGoals · development · feedback · opportunityPLAYER
Development TwinPlayer ↔ Coach relationship memoryDIGITAL TWIN
Architecture rule: agents do not own data or authority. They act on behalf of identified stakeholders under explicit roles, permissions, consent and audit rules.
COACH AGENT
Coach Agent
Plans sessions, reviews Player Twins, supports human judgment, builds Contribution Records and manages coach development.
READYSelect a task to see which agents and Digital Twins participate.
Identity · Authority · Consent · Audit
The Trust Layer Governs Agent-to-Agent Interaction.
Permission Matrix · Simplified Prototype
| Agent | Own Twin | Player Evidence | Contact Minor | Change AICL | Publish / Share |
|---|---|---|---|---|---|
| Player Agent | ✓ | Own | — | No | Consent |
| Guardian Agent | Relationship | Authorized | ✓ | No | Approve |
| Coach Agent | ✓ | Assigned players | Via guardian | Recommend only | Permissioned |
| Scout Agent | ✓ | Visible scope | No direct | No | Request |
| Academy Agent | Org | Org scope | Policy | No | Policy |
| Network Agent | Aggregate | De-identified | No | No | Governed |
Agent Permission Envelope
IdentityWho does the agent represent?
RolePlayer, guardian, coach, scout, academy, federation, regional or global network.
Data ScopeWhich Digital Twins and evidence objects may be read?
Action ScopeMay the agent explain, recommend, request, schedule, share or execute?
Consent / Legal BasisWhat permission authorizes the action, especially for minors?
Human Approval ThresholdWhich actions require explicit human validation?
Audit TrailWhat did the agent see, decide, recommend and route?
Expiry / RevocationWhen does authority end or get withdrawn?
Protected Multi-Agent Scenario
A Scout Discovers a 13-Year-Old Player — Without Bypassing the Guardian.
1 · Scout Agent searches permissioned profilesOnly fields exposed by Player/Guardian visibility policy are searchable.
2 · Matching Agent identifies potential fitDevelopment evidence supports a possible academy opportunity.
3 · Scout Agent requests expanded accessNo direct contact with the minor.
4 · Guardian Agent receives requestExplains who is requesting access, why and which fields would be shared.
5 · Guardian approves / limits / rejectsHuman guardian retains authority.
6 · Permissioned channel opensOnly approved information and communication paths become available.
Step 1 · Permissioned discovery
The Scout Agent can discover a Player Passport only within the visibility scope already authorized by the player/guardian policy.
ActorScout Agent
Data scopePublic / permissioned Passport fields
Human approvalNot yet required
Audit eventSearch + profile view logged
No direct minor contact: the system routes opportunity access through the Guardian Agent rather than allowing a Scout Agent to initiate unrestricted communication.
Multi-Agent Network Operations
One Network. Different Agents. Shared Evidence and Rules.
8Primary Agent RolesPlayer through Global Network.
5Regional AgentsAfrica · LATAM · Europe · NA · Asia.
10Language LayerShared ontology, localized interaction.
1Trust FrameworkIdentity + authority + consent + audit.
PlayerDigital TwinDevelopment evidence and goals.
CoachDigital TwinAICL + judgment + contribution.
OrgDigital TwinAcademy / club programs and resources.
NetworkDigital TwinRegional/global health and learning.
v0.33 principle: AI agents coordinate the football network, but rights and accountability remain attached to real stakeholders and organizations. High-impact actions cross explicit human approval gates.
v0.33 · Founding 100 Control Tower & Global Network Agent
Operate Five Regions as One AI-Augmented Global Coaching Network.
The Control Tower turns the First 100 Validation Cohort into an observable operating system. It tracks the full funnel across five regions and ten languages, while the Global Network Agent identifies bottlenecks and recommends interventions. Human teams retain authority over qualification, safeguarding, policy and consequential scale decisions.
First 100 Validation Funnel
147Applications
100Accepted
82Verified
64Curriculum Complete
513+ Players
383+ AI Sessions
31Evidence Ready
27Founding Active
76%30d Retention
0.56Qualified K-factor
Five Regional Networks
Asia25 target · 18 accepted · 9 active
LATAM25 target · 22 accepted · 7 active
North America20 target · 19 accepted · 5 active
Europe20 target · 24 accepted · 4 active
Africa10 target · 17 accepted · 2 active
ENESPT-BRFRDEITARZHJAKO
GLOBAL NETWORK AGENT ACTIVE
Asia · Player activation is the main bottleneck.
Verification and curriculum completion are healthy, but only 43% of qualified coaches have activated three players. Guardian/player invitation handoff is the largest visible friction.
Recommended interventionSimplify local invitation copy, shorten Guardian Consent handoff, and test a coach-guided invite flow in Korean and Japanese.
Policy boundaryThe Agent can recommend experiments, but cannot waive safeguarding, change qualification thresholds, or open Phase II seats without human approval.
Region → Language → Coach → Development Twin
Drill Down Without Losing the Global View.
| Region | Language | Accepted | Verified | 3+ Players | Founding Active | Main Friction |
|---|---|---|---|---|---|---|
| Asia | KO/JA/ZH/EN | 18 | 16 | 8 | 9 | Guardian/player activation |
| LATAM | ES/PT-BR | 22 | 18 | 14 | 7 | Evidence consistency |
| North America | EN/ES | 19 | 16 | 10 | 5 | Coach time / session cadence |
| Europe | EN/FR/DE/IT/ES | 24 | 20 | 11 | 4 | Verification heterogeneity |
| Africa | EN/FR/AR | 17 | 12 | 8 | 2 | Credential + connectivity |
Asia Network
Strong verification and curriculum completion; activation friction is concentrated in player invitations and consent handoff rather than coach interest.
89%Verified / accepted
43%3+ players / qualified
78%30d retention
0.42Qualified K-factor
71%Evidence completeness
0Safeguarding failures
Control-Tower principle: the objective is not to maximize raw registrations. Optimize for Active Founding Coaches generating real player-development evidence with healthy trust and retention.
Founding 300 Scale Gate
Should We Open the Next 200 Seats?
GREEN · Open Phase II
≥50 Active Founding Coaches; ≥70% retention; ≥5 active players/coach; ≥60% longitudinal evidence; no material safeguarding failures; evidence quality stable across regions.
AMBER · Improve First
30–49 Active Coaches or meaningful regional/language bottlenecks. Improve onboarding, evidence quality, player activation or coach value before scaling.
RED · Do Not Scale
<30 Active Coaches, poor repeat usage, weak evidence generation, serious trust issues or unstable AICL/attribution behavior.
Current prototype status: AMBER. The cohort shows promising retention and early referral behavior, but only 27 Founding Coaches are currently active in this demonstration. The Global Network Agent recommends improving activation and evidence consistency before opening all remaining 200 seats.
Multi-Agent Global Network Architecture
Global Agent → Regional Agents → Ambassadors → Coach Agents → Development Twins.
Global Network AgentCross-region health, scale gate, standards, anomaly detection.
Regional Network AgentsAfrica · LATAM · Europe · North America · Asia.
Country / Language AmbassadorsLocal terminology, coach recruitment, partnerships, cultural feedback.
Coach AgentsNext best action, sessions, player activation, AICL development.
Development TwinsPlayer ↔ Coach goals, sessions, evidence, outcomes and consent.
Player / Guardian AgentsDevelopment support, permissions, explanations and protected workflows.
Human Authority Layer
AI can observe, summarize, recommend and route work. Humans remain accountable for high-impact decisions.
QualificationHuman/trust review at meaningful AICL and Founding thresholds.
SafeguardingNo agent may waive guardian consent or youth-protection rules.
Marketplace policyRanking weights and commercial rules require governance approval.
Scale decisionsOpening the next 200 seats requires human approval at the Scale Gate.
Scientific claimsAttribution language must match comparator tier and evidence confidence.
Operating principle: automate coordination aggressively; automate authority cautiously.
v0.33 · Founding Coach Operating System
Coach Agent + 30-Day Founding Coach Journey.
This version turns the Founding Coach architecture into an operational workspace. The coach no longer returns to a marketing page after joining; the Coach Agent manages the next best action across identity, learning, player activation, AI-assisted sessions, evidence, AICL progression, Founding status and marketplace readiness.
My Founding Coach Journey
DAY
12 / 30
12 / 30
ACTIVATE
Invite players and complete first AI-assisted sessions.
Invite players and complete first AI-assisted sessions.
L1AICL
36%Evidence Confidence
B1Evidence Breadth
3Active Players
Next best action
Complete Session #3 with Min-jun and record your accept / modify / reject decision on the AI recommendation.
Complete Session #3 with Min-jun and record your accept / modify / reject decision on the AI recommendation.
Days 1–3 · IDENTITYVerify identity/experience, safeguarding and Coach Passport baseline.
Days 4–7 · LEARNComplete micro-learning curriculum and first AI-assisted plan.
Days 8–14 · ACTIVATEInvite ≥3 players, Guardian Consent, Player Twins, Session 1.
Days 15–21 · COACHComplete Sessions 2–3 and document human judgment.
Days 22–27 · PROVEGenerate longitudinal evidence and Coach Contribution Records.
Days 28–30 · FOUNDINGQualification review, initial AICL, Founding activation and marketplace readiness.
Coach Agent · Daily Operating Partner
COACH AGENT ACTIVE
Prepare the next session with evidence.
Three U13 players show a similar late-scan pattern before receiving under pressure. I recommend a 25-minute scanning + pressured-passing sequence. Review it, then accept, modify or reject it before the session.
No decision recorded yet. Human review is required before this recommendation becomes part of the Development Twin.
Atomic Operating Unit
Player ↔ Development Twin ↔ Coach.
Development Twin
The platform's core unit is the player–coach development relationship. It stores the goal, baseline, plan, AI recommendations, human decisions, sessions, guardian permissions, evidence and outcomes.
Player TwinBaseline · goals · evidence
↔
Development TwinPlan · sessions · decisions · outcomes
↔
Coach TwinAICL · judgment · contribution
PlayerMin-jun · U13 · Midfielder
Development GoalPassing Intelligence + scan-before-receive behavior
BaselinePassing Intelligence 58 · FCI 54 · Comparator Tier C
Plan3-session pressured receiving + decision sequence
Guardian PermissionActive · development analysis + Coach Passport contribution evidence allowed
Coach Contribution Record · Live Example
AI observationLate scan before receiving under pressure.
AI recommendationConstrained scanning + one-touch passing sequence.
Human decisionMODIFY — slower first progression + shoulder-check cue.
Player action3 progressive sessions planned.
OutcomePending post-session evidence.
Why this matters: Human Coaching Judgment becomes observable without rewarding blind AI compliance. The record updates both Player Development Attribution and the Coach Twin.
Founding 100 Control Tower
Operate the Cohort — Not Just the Website.
| Coach | Region | Language | Day | Stage | AICL | Players | Status |
|---|---|---|---|---|---|---|---|
| Coach Kim | Asia | KO/EN | 12 | Activate | L1 | 3 | On track |
| Coach Sofia | LATAM | ES | 19 | Coach | L3 | 7 | On track |
| Coach Daniel | Europe | FR/EN | 24 | Prove | L4 | 11 | Needs evidence review |
| Coach Amina | Africa | AR/FR | 9 | Activate | L1 | 2 | Player invite bottleneck |
| Coach Alex | NA | EN/ES | 28 | Founding | L4 | 9 | Review ready |
25LATAM target
25Asia target
20NA target
20Europe target
10Africa target
Founding Network Agent
Where is the cohort getting stuck?
The Network Agent summarizes funnel health by region, language, coach type and stage, and proposes interventions without silently changing policy or qualification rules.
Asia · Player ActivationCurriculum completion is strong, but Guardian/Player invitation conversion is weaker than the cohort median. Recommendation: simplify local invitation copy and consent handoff.
Africa · VerificationCredential formats vary. Recommendation: add an experience-based verification pathway rather than requiring one licensing pattern.
LATAM · ReferralEarly coach-to-coach referral behavior is strong. Recommendation: test Founding Connector rewards after qualified activation, not raw invitations.
v0.33 · AICL Fairness Revision
Quality Determines Level. Quantity Determines Confidence and Breadth.
A part-time or specialist coach should not be held at a low AICL merely because they coach fewer players. Once minimum evidence eligibility is crossed, coaching capability, human judgment, player-development impact, evidence practice and trust determine AICL; sample size primarily determines how confidently and broadly that capability can be generalized.
Three independent professional signals
CAPABILITYAICL L5AICS 69
CONFIDENCEEC 68%Moderate
BREADTHB2Focused
Revised rule: AICL = capability subject to dimension floors, trust and minimum evidence eligibility. Evidence Confidence (EC) expresses certainty. Evidence Breadth (EB) expresses how widely the evidence generalizes.
Evidence Breadth · B0–B5
B0Insufficient
B1Initial
B2Focused
B3Established
B4Diverse
B5Extensive
Breadth considers players, sessions, duration, age groups, skill levels, positions and contexts — not player count alone. A focused specialist can therefore achieve a high AICL while honestly showing narrower breadth.
Fairness Example
Part-Time Excellence Should Not Lose to Full-Time Volume.
Coach A · Part-Time Specialist
Six U13 midfielders · high-quality longitudinal development · strong human judgment.
L6AICL
68%Confidence
B2Focused breadth
82AICS
6Players
18Sessions
High capability with moderate confidence. The Passport should describe the specialization rather than penalize the coach for working part-time.
Coach B · Full-Time Generalist
Forty-five players · extensive evidence volume · moderate demonstrated development impact.
L4AICL
94%Confidence
B5Extensive breadth
66AICS
45Players
120Sessions
More evidence makes the assessment more certain and generalizable, but volume alone does not make the coach more capable.
Revised L0–L9 Interpretation
Require breadth mainly when the claim itself becomes broader.
L1–L2 · Practice
Safe, competent AI-assisted workflow. Small but real minimum evidence is sufficient.
L3–L4 · Demonstrated
Visible player-development evidence and repeatable human judgment. Quantity has diminishing influence after eligibility.
L5–L6 · Advanced / Specialist Excellence
Part-time specialists can reach these levels through strong repeated outcomes, even with focused B2/B3 breadth.
L7 · Expert
Requires evidence that capability generalizes beyond a narrow player or context sample.
L8 · Leader
Adds institutional, mentoring or network contribution and broader validation.
L9 · Master
Exceptional sustained player development, methodology contribution, leadership and independent validation.
Minimum evidence still matters: one player and two sessions cannot establish advanced capability. But after a sensible eligibility threshold, player volume should show confidence/generalizability rather than function as a crude quality score.
v0.33 · Evidence-Based Coach Passport & Marketplace
Make Coaching Value Visible — Through Player Development Evidence.
Connect the Coach Digital Twin to economic opportunity. Human qualification stays visible, while AICL, longitudinal player-development impact, evidence confidence, comparator quality, human judgment and reputation explain why a coach may be a strong match.
Evidence-Based Coach Passport
CK
Coach Kim · Founding Cohort 01
Seoul · Korean / English · Youth · Football + FutsalL5VALIDATED AICL
HUMAN QUALIFICATIONA LicenceIndependent from AICL
AICS74Evidence confidence 86%
ACTIVE PLAYERS3411 months evidence
REPUTATION4.8 / 5Separate from capability
PLAYER DEVELOPMENT IMPACT
+12%
+12%
Comparator Tier B · matched longitudinal peers · strongest evidence: Passing Intelligence 88 · FCI 81 · FSI 84.
Why L5? Advanced AICS and strong development evidence meet score/dimension floors. Current L5 reflects demonstrated capability; evidence volume is reported separately through Evidence Confidence and Breadth.
Evidence-Based Coach Matching
Find the right coach for this player's development need.
Coach Kim · L5 · 92% match
Passing 88 · FCI 81 · Evidence 86% · Tier B · strong U13 development evidence.
Coach Sofia · L4 · 86% match
Technical 91 · FCI 78 · Evidence 82% · Tier B · strong youth retention.
Coach Daniel · L6 · 83% match
FSI 90 · Mentor · Evidence 91% · Tier A/B · advanced team-development evidence.
SELECTED COACH
Coach Kim · 92%
Coach Kim · 92%
Best fit for U13 Passing Intelligence because relevant development evidence, comparator quality and evidence confidence align with the player's need.
Matching rule: no single score decides. Combine player need, demonstrated specialty, AICL, evidence confidence, comparator tier, age/context fit, reputation, retention and availability.
Evidence → Economic Opportunity Flywheel
Reward demonstrated development — not visibility alone.
1 · Coach develops playersReal sessions + human judgment + AI augmentation.
2 · Player Twins generate evidenceLongitudinal change + comparator + confidence.
3 · Coach Passport strengthensAICL + specialties + contribution history.
4 · Better-fit players discover coachEvidence-based marketplace matching.
5 · Reputation and earning opportunity growSubscriptions, premium services, mentoring, network roles.
6 · More diverse evidence improves the TwinSubject to anti-gaming and scientific validation.
Marketplace Ranking · Provisional
Capability ≠ popularity.
AICL / demonstrated capability35%
Player / development-goal fit20%
Reputation quality15%
Evidence confidence + comparator15%
Retention / continuity10%
Availability / context5%
Anti-gaming: do not reward cherry-picking elite players, easy drills, excessive AI usage, review manipulation or raw player volume. Audit rankings for regional, language, age-group, resource and socioeconomic effects.
v0.33 · Player Development Attribution Engine
Measure Player Improvement. Estimate What the Coach Actually Contributed.
Dimension D in AICL should not reward a coach simply for having talented players. The attribution engine separates observed development from expected development and then estimates the evidence-supported contribution associated with human coaching, AI augmentation and context. It is a provisional research model — not a causal claim engine.
Core attribution model
Observed ≠ Attributed.
Observed Development
− Expected Development without the intervention
= Development Above Expectation
Then estimate:
Human Coach + AI Augmentation + Player Effort + Environment + Unexplained
− Expected Development without the intervention
= Development Above Expectation
Then estimate:
Human Coach + AI Augmentation + Player Effort + Environment + Unexplained
Scientific guardrail: SSOT should initially say “estimated contribution associated with the coaching cycle,” not “the coach caused X% of improvement.” Causal attribution requires stronger experimental or quasi-experimental validation.
Five evidence components
Build D from development evidence, not popularity.
1
BaselineStarting Player Twin, age, experience, position, prior trend.
BEFORE2
Expected trajectoryMatched historical peers / within-player prior trend / benchmark model.
COUNTERFACTUAL3
Observed changeTechnical + FCI + FSI + FEI + physical/context evidence.
AFTER4
Contribution recordAI recommendation → coach judgment → player action → outcome.
TRACE5
ConfidenceVolume, duration, reliability, comparator quality and missing-data penalties.
TRUSTInteractive Attribution Lab
Change the Player Twin and See What Can Be Credibly Attributed.
The sliders demonstrate architecture only. Production estimates should be learned/calibrated from longitudinal cohorts and should preserve uncertainty rather than forcing every outcome into a precise percentage.
Player Development Attribution Preview
+14Observed gain
+9Above expectation
64Provisional D score
Expected / natural development5 pts
Coach-associated signal4.7 pts
AI-associated signal2.1 pts
Player / environment / unexplained2.2 pts
Moderate-confidence attribution
The player improved 14 points; 5 were expected from the comparison trajectory. The remaining 9-point above-expectation change is apportioned provisionally using traceable coaching, AI and engagement evidence, then discounted by evidence confidence.
Coach Contribution Record
Make human judgment observable.
1 · AI observationPlayer scans late before receiving under pressure.
2 · AI recommendationAdd constrained scanning + one-touch passing sequence.
3 · Human coach judgmentMODIFY — reduce speed first; add shoulder-check cue and teammate trigger.
4 · Player actionCompletes three sessions with progressive pressure.
5 · OutcomeDecision latency and successful pressured-pass evidence improve.
Attribution hierarchy
Use the strongest comparator available.
Tier A · Controlled / quasi-experimentalStrongest research setting; matched controls or designed intervention.
Tier B · Matched longitudinal peersComparable age, baseline, role, training exposure and context.
Tier C · Within-player trajectoryCompare intervention period against stable prior development trend.
Tier D · Benchmark expectationPopulation benchmark adjusted for age/experience/context.
Tier E · Descriptive onlyNo defensible comparator: report improvement but do not attribute it.
First 100 Validation Cohort · Research Agenda
Use the founding cohort to learn whether Dimension D deserves trust.
ReliabilityDo repeated measures and reviewers produce stable development estimates?
Counterfactual qualityWhich peer/trajectory model best predicts expected development?
Coach attributionDo contribution records predict outcomes beyond baseline/player factors?
AI contributionWhen does AI improve human decisions versus add no measurable value?
HeterogeneityAge, level, region, language, position and coaching context.
Football / FutsalShared framework, sport-specific benchmarks where evidence requires them.
Bias / gamingDetect cherry-picked players, easy drills, missing sessions and score inflation.
Predictive validityDoes D predict sustained future development rather than short-term test gains?
Recommended production rule: Dimension D should display three things together — Development Impact Score, Attribution Confidence, and Comparator Tier. Never show a precise coach-contribution percentage without sufficient evidence.
Interactive AICL Scoring Lab
Change the coach twin. See what validates.
PROVISIONAL SCORE
69
Validated L2CAPABILITY PROJECTION
L5
before evidence constraintsScore gate · L5Based on weighted AICS.
Evidence gate · L2Players + sessions + duration.
Dimension floor · L5Weakest dimension can cap level.
Why L2?
Evidence volume is the binding constraint.
START WITH YOUR ROLE
One SSOT. One Football Journey.
Begin as a player, coach or club — then connect the rest of the ecosystem when you need it.