Player
Train with your Personal AI Coach, build your Player Twin and track development.
Open SSOT →
SUPERSTARS OF TOMORROW: 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
OFFICIALLY CERTIFIED PRODUCT
SuperStars of Tomorrow
AI FOOTBALL PLATFORM FIFA Quality Programme for EPTS Broadcast Tracking Systems.
View on FIFA website →
AI FOOTBALL PLATFORM FIFA Quality Programme for EPTS Broadcast Tracking Systems.
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.
FOUNDING COACH PROGRAM
Start small. Prove value. Evolve together.
Today's invitation is a manageable first step — not a commitment to the entire 18–24 month roadmap. We begin with evidence, human judgment and a small number of real player-development cycles.
1 · Apply · ~10 minCoaching background, player-development experience,
AI/evidence mindset and readiness.
2 · 14-Day ChallengeExperience SSOT and practice ACCEPT / MODIFY / REJECT
decisions over AI recommendations.
3 · 30-Day PilotSelected coaches run real Player × Coach × AI learning cycles
and measure outcomes.
4 · Founding NetworkStrong contributors may continue into Coach Passport,
network and future multi-level programs.
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
Improveplayer 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.
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
v1.0 · PLAYER CAREER PATHWAY MANAGER
Manage many opportunities as one development career.
A player should not chase every attractive trial. v1.0 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
v1.0 · CO-EVOLUTION ENGINE
PLAYER × COACH × AI · FROM PERSONAL LEARNING TO GLOBAL COLLECTIVE INTELLIGENCE
Every session becomes a learning episode. Every validated episode can improve the network.
v1.0 connects the existing Player Twin, Coach Twin, AI-Augmented Coaching, evidence, Passport, marketplace and trust layers into one progressive learning architecture. Start small, prove measurable value, then increase the number of players, coaches, academies and AI policies only when each level passes its graduation gate.
KILLER-APPLICATION LADDER
Product value grows before algorithmic complexity.
The platform does not require deep MARL on day one. Each stage is a useful product and creates the structured trajectories, reward signals and trust records needed for the next.
Development TwinPredict the next best intervention for one player and
verify whether it worked.
Coach-of-the-CoachLearn which human coaching decisions work for which
player profiles.
Dynamic Coach–Player MatchingMatch by evidence-based development fit
rather than popularity.
Training Group ComposerOptimize groups, coaches and goals across an
academy.
AI Coaching ChallengeCompare competing AI/human intervention policies
using real outcomes.
Football Intelligence LeagueTurn talent development itself into a
measurable competitive learning ecosystem.
WORKING v1.0 PROTOTYPE
Co-Evolution Lab
Operate L1 and L2 as measurable learning loops; simulate L3 Academy MARL before production-scale distributed learning.
PLAYER DEVELOPMENT TWIN · GABRIEL FERREIRA · U14 FORWARD
What intervention should we try next?
Baseline state combines drill evidence, coach observation, workload and previous response. This prototype keeps the recommendation bounded and human-reviewable.
Finishing68baseline
Weak Foot54priority
Consistency61moderate
Load72%safe range
BOUNDED INTERVENTION CANDIDATES
AI hypothesis: weak-foot finishing under moderate pressure offers the
best development value this cycle.
Choose an intervention. v1.0 will simulate a re-test,
attribute a provisional Δ and update the Player Twin.
L1 graduation gate: repeatable, evidence-backed improvement over
baseline with acceptable safety, confidence and coach validation.
TRIADIC CO-EVOLUTION · LEARNING EPISODE #002843
AI proposes. Human decides. Player responds. All three learn.
PLAYERGoal · effort · response · outcome
↔
AI COACHMeasure · propose · predict · remember
↔
HUMAN COACHContext · motivate · ACCEPT / MODIFY / REJECT
AI PROPOSAL
25 weak-foot finishes under progressive pressure
Predicted benefit +4.6 technical points. Confidence 78%. Constraint: fatigue increases after rep 20 in prior sessions.
Human decision has not yet been recorded.
TRAJECTORY / PROVENANCE
Observe
AI Proposal
Human Decision
Intervene
Re-test
Outcome
Learn
Player Δ—
Coach Lesson—
AI Update—
CEI—
SMART ACADEMY · 20 PLAYERS × 4 COACHES × 5 AI ROLES
Optimize today's learning configuration.
Prototype objective balances individual development, group development, coach fit, workload/safety and scarce coaching resources. In production this becomes a MARL environment with bounded action spaces and human override.
ACADEMY OBJECTIVE SCORE78.4
Weighted multi-objective score — simulation only
Player Development35
Coach Fit20
Group Learning20
Safety / Load15
Efficiency10
L3 graduation gate: academy-level policy should improve outcomes
versus isolated triads without degrading safety, fairness or coach autonomy.
L4 · COHORT + POPULATION LEARNING
Promote lessons only after evidence thresholds are met.
Fast learning remains personal. Cohort learning aggregates comparable experiences. Global knowledge changes slowly after validation, preventing one noisy session from retraining everyone.
Personal Episode→Cohort
Evidence→Replication→Scientific
Validation→Network Policy
Promotion rule: Evidence → Hypothesis → Experiment → Replication →
Validation → Network Knowledge.
NETWORK SNAPSHOT · SIMULATED
Active Triads1,248+84 / month
Validated Episodes18.6Kprovenance intact
Cohorts42age/role/context
Promoted Lessons137network-level
| Candidate lesson | Cohort | Evidence | Status |
|---|---|---|---|
| Weak-foot pressure progression | U13–U15 forwards | B1 · 184 episodes | Replication |
| Scanning-before-receive cue | U14–U17 midfield | A2 · 412 episodes | Promoted |
| High-rep blocked finishing | U11–U13 | C2 · 73 episodes | Hold |
FOOTBALL
INTELLIGENCE LEAGUE · FIL
COMPETE TO DEVELOP TALENT BETTER · COOPERATE TO ADVANCE FOOTBALL INTELLIGENCE
From academy challenges to a Global Football Intelligence League.
FIL evaluates development systems—not only match results. Players, human coaches, academies and AI policies can participate in standardized learning challenges. Verified outcomes feed the same Co-Evolution Engine.
Player GrowthCoach GrowthAI
ImprovementTransferSafetyEvidence
GLOBAL FIL
CONTINENTAL FIL
NATIONAL FIL
REGIONAL FIL
ACADEMY / LOCAL CHALLENGE
LEAGUE SCOPE
Competition principle: standardized challenge + context adjustment +
evidence confidence + human safeguards. FIL should never reduce young players to a single opaque ranking.
REGIONAL FIL · DEVELOPMENT CHALLENGE
Seoul U14 Technical Development Challenge
8-week standardized challenge: improve first touch and decision speed while preserving player load and coach autonomy.
| # | Human + AI Team | Player Δ | Coach Δ | Evidence | CEI |
|---|
Co-Evolution Intelligence Index (CEI)
Player learning + Coach learning + AI improvement + collaboration + transfer + evidence/safety adjustment. CEI is a transparent decision-support metric, not a permanent human label.
Player learning + Coach learning + AI improvement + collaboration + transfer + evidence/safety adjustment. CEI is a transparent decision-support metric, not a permanent human label.
AI COACHING CHALLENGE · INSIDE FIL
Which AI helps humans improve—not merely which model sounds smartest?
Different AI policies can receive comparable, permissioned development scenarios. Human coaches retain final authority; real-world outcomes determine whether a policy advances.
AIBrain Coach87.3
Academy Policy A84.8
Human-only Baseline78.6
Best Transfer91
Safety98
EvidenceA2
TRUSTED CO-EVOLUTION FABRIC
MARL may optimize recommendations. Authority remains explicit.
v1.0 extends the existing trust model across Player, Coach, Academy and AI Agent identities. Learning never silently expands an agent's permissions.
Player PassportIdentity · consent · development evidence ·
rights
Coach PassportCredentials · capability · outcomes · authority
AI Agent PassportPrincipal · policy/model · purpose · scope ·
expiry
Academy IdentityOrganization · roles · safeguarding ·
delegations
FIL ProvenanceChallenge rules · evidence · context adjustment ·
audit
IDENTITY→ROLE /
RELATIONSHIP→AUTHORITY→ACTION→EVIDENCE→OUTCOME→LEARNING
v1.1 · RL / MARL IMPLEMENTATION BRIDGE
Product learning first. RL complexity only when the evidence supports it.
The latest production network remains the product and data foundation. Reinforcement learning is added behind bounded environments, evidence gates and human authority — not as an uncontrolled replacement for coaching logic.
L1 · Personal adaptation
Contextual bandits / offline policy evaluation over bounded drill choices. Reward comes from measured player delta, confidence, retention and safety.
L2 · Triadic learning
Player state + AI proposal + human ACCEPT/MODIFY/REJECT + outcome becomes a provenance-preserved trajectory for incremental policy improvement.
L3 · Smart Academy MARL
PettingZoo-style multi-agent environment contract; RLlib-compatible policy training for player grouping, coach allocation and specialized AI roles.
L4–L7 · Population + FIL
Cohort learning, policy leagues, self-play/league evaluation and evidence-gated promotion from regional to national and global intelligence.
OBSERVATIONPlayer Twin · Coach Twin · Development Twin · session/match
evidence · consent/authority context
ACTION SPACEBounded drills · difficulty · repetitions · grouping · coach
assignment · feedback strategy · escalation
REWARD / VALUEPerformance delta · human judgment · evidence quality ·
transfer · retention · safety · fairness · attribution confidence
ENVIRONMENTSingle-player → triadic → academy MARL → regional FIL →
national/global policy league
TRUST BOUNDARYRL may optimize recommendations; identity, consent,
authorization, safeguarding and hard policy remain deterministic and auditable.
v0.39 continuity: its measurable coaching-value concept is retained as
part of the reward/evaluation layer. Diagnosis, evidence quality, intervention design, human judgment,
execution, performance delta, attribution, adaptation and reflection remain candidate reward/value components
rather than being discarded.
v1.0 · 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%
v1.0 · 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.
v1.0 · Super-Talent Development Science
Three Football Intelligence Engine
Develop football intelligence as observable, longitudinal evidence—not as a fixed label. The engine separates Football Cognitive Intelligence (FCI), Football Social Intelligence (FSI), and Football Emotional Intelligence (FEI), then reconnects them inside real football actions. Passing is the first integration laboratory because it links perception, another player, team coordination and technical execution.
FCI · COGNITIVE
See. Decide. Adapt.
Scanning, awareness, option recognition, decision quality, anticipation and tactical adaptation.
FSI · SOCIAL / TEAM
Connect. Combine. Contribute.
Teammate awareness, timing, cooperation, support movement, combination play and team contribution.
FEI · EMOTIONAL
Regulate. Recover. Respond.
Observable response to mistakes, pressure, feedback and repeated challenge—supported by coach/player evidence.
Passing Intelligence Ladder
From technique to team intelligence
Higher levels progressively introduce other intelligent agents, pressure and real match context.
L0–L2Wall passing · individual technical
baselineTECHNICAL
L3–L4Two-player passing ·
synchronizationSOCIAL
L5–L6Triangle passing · awareness +
supportCOMBINE
L7Rondo · decisions under pressurePRESSURE
L8Small-sided game · dynamic team
intelligenceGAME
L9Match context · real team contributionMATCH
FSI Evidence Model
Wall Passing · L0–L2
Technical baseline: accuracy, pace consistency, weak-foot use and first-touch-to-pass time.
62Teammate
Awareness
Awareness
67Decision
Quality
Quality
71Cooperation
Quality
Quality
58Combination
Intelligence
Intelligence
64Team
Contribution
Contribution
AI Development Hypothesis
Passing accuracy is stable; the next information gain comes from adding a teammate.
Do not infer general “social intelligence” from pass completion. Test whether timing and pass weight improve when the receiver changes position, then compare the response across repeated sessions.
FCI Where is the opportunity?
FSI How do I help the team?
FEI Can I execute under pressure?
Skill Execute the pass
Scientific guardrail: FCI, FSI and FEI are developmental constructs built from observable
football behaviors and repeated evidence. v1.0 does not claim to measure a child's general IQ,
personality, inner emotional state or fixed future potential.
v1.0 · 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
v1.0 · 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.
v1.0 · 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.
v1.0 · 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 Validated
AICS 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.
v1.0 · First Operational Global AI Football Intelligence Network
From Prototype Network to Pilot-Ready System Architecture.
Persistent stakeholder accounts, trusted agent authorization, Digital Twins, Coach OS, player-development evidence, marketplace transactions, ten-language localization and network analytics.
1 · Identity & AccountsPlayer · Guardian · Coach · Scout · Organization
2 · Trust & AuthorizationRole · scope · consent · audit · revocation
3 · Digital TwinsPlayer · Coach · Development · Org · Network
4 · Coach OS30-day journey · AICL · EC · EB
5 · Development EngineGoals · sessions · evidence · attribution
6 · Multi-Agent RuntimeAuthorized stakeholder agents
7 · Marketplace & PaymentsMatch · trial · subscription · ledger
8 · Analytics / APIsControl Tower · five regions · ten languages
Persistent State + Core Schema
Prototype UI State Becomes Operational Data.
Coach State Machine
APPLIED → ACCEPTED → IDENTITY_PENDING → VERIFIED → CURRICULUM_ACTIVE → PLAYER_ACTIVATION → EVIDENCE_BUILDING → FOUNDING_REVIEW → FOUNDING_ACTIVE → MARKETPLACE_ACTIVE
Exception states: SUSPENDED · TRUST_REVIEW · WITHDRAWN.
Every transition: actor + timestamp + reason + authorization context
+ audit event.
Development Twin
Player Twin ↔ Development Twin ↔ Coach Twin
Stores goal, baseline, plan, consent, AI recommendations, human decisions, sessions, evidence, outcomes and attribution. One player may have multiple Development Twins with different coaches.
Core persistent entities
users / stakeholder_profilesIdentity + locale + status
organizations / membershipsClub · academy · federation
players / guardians / coachesRole profiles
player_twins / coach_twinsLongitudinal state
development_twinsPlayer–Coach relationship
passportsPermissioned identity/evidence
sessions / drills / observationsActivity evidence
ai_recommendations / human_decisionsJudgment trace
contribution_records / outcomesAttribution chain
aicl / confidence / breadthCoach measurement
consents / access_grantsAuthorization
agent_authorizations / audit_eventsTrust trace
marketplace_matchesMatching rationale
subscriptions / transactionsBilling + split ledger
referrals / ambassadorsQualified network growth
Reference API Surface
Engineering Contracts for the Pilot.
POST /v1/coaches/apply
GET /v1/coaches/{id}/passport
GET /v1/coaches/{id}/aicl
POST /v1/players/{id}/invite
POST /v1/development-twins
POST /v1/sessions
POST /v1/sessions/{id}/evidence
POST /v1/coach-decisions
GET /v1/development-twins/{id}/attribution
POST /v1/guardians/consents
POST /v1/access-grants
POST /v1/agent-authorizations/check
GET /v1/audit-events
POST /v1/marketplace/matches
POST /v1/subscriptions
POST /v1/referrals
GET /v1/network/founding100
GET /v1/analytics/funnel
Authorization rule: endpoint existence never implies permission.
Protected operations evaluate identity, role, relationship, consent, scope and human-approval requirements.
Pilot Acceptance Criteria
Call v1.0 Operational Only When Real Evidence Flows.
100Accepted coaches
≥50Active Founding Coaches
≥250Active players
≥750AI-assisted sessions
5Regions
10Languages
RealGuardian consents
RealAICL · EC · EB
PaidSmall marketplace pilot
0Material safeguarding failures
Boundary: these are pilot targets, not claims achieved by this HTML
reference console.
v1.0 · 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.
v1.0 · 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.
v1.0 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.
v1.0 · 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.
v1.0 · 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.
v1.0 · 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.
v1.0 · 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.
v1.0 · 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.
BEFORE
2
Expected trajectoryMatched historical peers / within-player prior trend / benchmark
model.
COUNTERFACTUAL
3
Observed changeTechnical + FCI + FSI + FEI + physical/context evidence.
AFTER
4
Contribution recordAI recommendation → coach judgment → player action →
outcome.
TRACE
5
ConfidenceVolume, duration, reliability, comparator quality and missing-data
penalties.
TRUST
Interactive 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 L2
CAPABILITY PROJECTION
L5
before
evidence constraints
Score 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.
AI-ASSISTED REVIEW
Gabriel Ferreira · Cone Dribbling
Player video evidence30–120 sec · body + ball visible
AI PRELIMINARY RUBRIC
AI DRAFT FEEDBACK · COACH CAN EDIT
v1.0 · Global Founding 300 / First 100 Validation Cohort
300 Founding Coaches. First 100 Validate the System.
Founding 300 is Cohort 01. Phase I validates the operating model with the first 100; the next 200 expansion seats open only after activation, retention, evidence and trust gates are met.
Phase I · First 100
Validate acquisition → qualification → 30-day activation → players → AI-assisted evidence → retention → referral.
Phase II · Next 200
Scale globally only after Phase I works. Reallocate the remaining seats using observed regional performance rather than rigid quotas.
Planning distribution · Founding 300
75LATAM
70Asia
55North America
60Europe
40Africa
Illustrative AI-Augmentation distribution
L075
L155
L245
L335
L430
L520
L615
L712
L88
L95
Two independent dimensions: Human Coaching Qualification ≠ SSOT
AI-Augmentation Level. A Pro-licensed coach can begin at AI L0 because SSOT has not yet observed sufficient
AI-assisted evidence. L0 means “not yet demonstrated,” not poor coaching.
Benefits & economics
Make Cohort 01 valuable for life.
Permanent Founding designationFounding AI-Augmented Coach · Cohort 01 ·
2026 in Passport history.
Founding Coach Passport + Coach AgentIdentity, specialties, languages,
AI profile, planning and development assistance.
First-player packageProposed: first 10 players free for 3 months.
Marketplace priorityEarly visibility; later ranking becomes
evidence/reputation based.
Ambassador / Advisory pathwaysRegional, language, validation and
Founding Council roles are earned.
80% Coach / 20% Platform
Proposed for Founding 300 during first 12 months; then planned standard 70/30 marketplace split. Final commercial terms require approval.
10-module qualification
1 AI-Augmented CoachingAI assists; human remains responsible.
2 SSOT Football IntelligenceTechnical + FCI + FSI + FEI +
context.
3 Five Essential DrillsEvidence + Passing Intelligence.
4 Player Digital TwinPerformance, velocity, confidence, history.
5 AI Coach AgentPlan → observe → recommend → human validate.
6 Multi-Camera1 / 2 / 3 smartphone workflow.
7 Safeguarding & Guardian ConsentMandatory for minors.
8 Coach Passport & ReputationEvidence-based progression.
9 Marketplace & EconomicsSubscriptions, scholarships, referrals.
10 Practical QualificationReal player-development cycle, not merely a
quiz.
30-Day Activation Challenge
Identity → Learn → Activate → Coach → Prove → Founding.
Days 1–3IDENTITY: verify identity/experience, create Coach
Passport, establish AI-Augmentation L0 baseline.
Days 4–7LEARN: curriculum, safeguarding, first AI-assisted
training plan.
Days 8–14ACTIVATE: invite ≥3 players, Guardian Consent where
needed, Player Twins, session 1.
Days 15–21COACH: sessions 2–3; coach
accepts/rejects/modifies AI recommendations.
Days 22–27PROVE: baseline/current evidence + Coach
Contribution Record.
Days 28–30FOUNDING: qualification review and evidence-based
initial AI level; L0/L1 may be appropriate.
Qualified Referral Engine
A referral counts only after verification, Coach Passport, player activation and real session evidence.
1Credit
3Founding Connector
5Ambassador candidate
10Network Builder candidate
Qualified Coach K-factor = invitations × conversion × activation. ≥0.5 promising · ≥0.8
strong · ≥1.0 self-propagating.
Regional Ambassador Pathway
Active Founding CoachReal activation + ongoing engagement.
AI L3+ targetDemonstrated use, not onboarding alone.
5+ activated coach referralsQuality network contribution.
20+ players activatedDirectly or through referrals.
Trust + local-language leadershipSafeguarding, retention, terminology
and partnerships.
Founding 100 Control Tower
Scale only when Phase I works.
≥70%Application → verified
≥80%Verified → curriculum
≥70%Curriculum → 3+ players
≥70%Players → 3 sessions
≥50Active Founding Coaches
≥70%30-day retention
≥5Active players / coach
≥4AI sessions / coach / month
≥60%Longitudinal evidence within 60d
GREEN · Open next 200
≥50 active; ≥70% retained; ≥5 active players/coach; ≥60% longitudinal evidence; no material safeguarding failures.
AMBER · Improve first
30–49 active or weak activation/retention. Fix onboarding, coach value and player workflow before scaling.
RED · Do not scale
<30 active, poor repeat usage, weak evidence generation or material trust/safeguarding problems.
Priority 2 · First Operational Global Network Program
Join the Global Founding 300 AI-Augmented Coaches.
Coach better with AI. Build your verified Coach Passport. Develop real players with measurable evidence. Build your reputation and help form the seed network for the Global AI-Augmented Football Network.
Global Founding 300 · Cohort 01
We count active coaches — not registrations. Founding status requires qualification, Coach Passport onboarding, player activation and real AI-assisted development evidence.
0 / 100 Phase I Validation
PROTOTYPE COUNTER · NOT LIVE REGISTRATIONS
PROTOTYPE COUNTER · NOT LIVE REGISTRATIONS
25LATAM
25Asia
20NA
20Europe
10Africa
Active definition: verified identity/qualification + Coach Passport +
3+ active players + 3+ AI-assisted sessions + development evidence.
Apply · Interactive Prototype
Founding status is earned: an application alone does not
count toward the 100.
Automated Founding Coach Journey
Application → Active AI-Augmented Coach.
1 · Recruitment Agent
Explain program in the coach's language and capture application.
Explain program in the coach's language and capture application.
2 · Identity & Qualification Agent
Verify identity, credentials/experience and safeguarding.
Verify identity, credentials/experience and safeguarding.
3 · Coach Passport Agent
Create specialties, languages, sport profile and L0 baseline.
Create specialties, languages, sport profile and L0 baseline.
4 · AI-Coach Onboarding Agent
Teach workflow and first AI-assisted coaching plan.
Teach workflow and first AI-assisted coaching plan.
5 · Player Activation Agent
Invite 3–5 players; minors route through Guardian Consent.
Invite 3–5 players; minors route through Guardian Consent.
6 · Evidence & Development Agent
Complete 3 sessions and update Player/Coach Digital Twins.
Complete 3 sessions and update Player/Coach Digital Twins.
7 · Founding Coach Agent
Activate Cohort 01 status, marketplace profile and referral loop.
Activate Cohort 01 status, marketplace profile and referral loop.
APPLICATION
Meet your Recruitment Agent
I'll explain the Founding 100 program and help you prepare a complete application.
L0Level
0Players
0AI sessions
NoFounding
Operational KPI: 100 Active Founding AI-Augmented Coaches, not 100
applications. Human review remains at trust, safeguarding and qualification gates.
Grassroots-to-Pro Capture
Record Anywhere. Analyze Like a Pro.
Use 1–3 smartphones with one controller workflow. Start simple with a center phone, expand to Left + Right, or use Left + Center + Right for the broadest coverage.
Choose Your Camera Setup
📱 1 Phone
Center only · quick setup
Center only · quick setup
📱📱 2 Phones
Left + Right · wider coverage
Left + Right · wider coverage
📱📱📱 3 Phones
Left + Center + Right · broad/full-pitch coverage
Left + Center + Right · broad/full-pitch coverage
Workflow: mount → create session → scan QR → confirm cameras → start/stop from controller → stitch/analyze.
Controller Preview
1 CAMERA CONNECTED
CENTER
Connected
Connected
Fast setup for solo drills and smaller zones.
v1.0 · Global Coach Passport & Qualification Network
Every Qualified Human Coach Can Build a Trusted Global Coaching Identity.
SSOT Coach Passport separates verified qualification from dynamic coaching performance. Coaches can qualify through a federation-license route, documented-experience route, or SSOT qualification route. After issuance, real coaching evidence continuously updates the Passport and L0–L9 Coach Intelligence level.
Qualification lifecycle
Apply → Verify → Qualify → Issue → Coach → Level Up → Renew
The entry path can be inclusive while evidence verification remains strict. SSOT verifies credential evidence; it does not replace the issuing federation.
1
Apply
Identity, profile, language, age groups and specialties.
START
Identity, profile, language, age groups and specialties.
2
Verify
Identity, credentials, experience and safeguarding.
TRUST
Identity, credentials, experience and safeguarding.
3
Qualify
License, experience or SSOT practical-assessment route.
GATE
License, experience or SSOT practical-assessment route.
4
Issue Passport
Portable verified coaching identity becomes active.
ISSUE
Portable verified coaching identity becomes active.
5
Coach & Earn Evidence
AI reviews, sessions, player outcomes and reputation.
LIVE
AI reviews, sessions, player outcomes and reputation.
6
Level Up L0–L9
Rolling evidence updates Coach Intelligence and matching.
GROW
Rolling evidence updates Coach Intelligence and matching.
7
Renew / Maintain
Credential expiry, inactivity and safeguarding remain visible.
RENEW
Credential expiry, inactivity and safeguarding remain visible.
Clickable Coach Passport
SSOT · GLOBAL COACH PASSPORT
AI-AUGMENTED HUMAN COACHAPPLICANT
AI-AUGMENTED HUMAN COACHAPPLICANT
CO
Lucas Oliveira
🇪🇸 / 🇧🇷 · Youth Development · Passing / Midfield
Credential evidence pending verification
CRD—
EXP—
AI—
OUT—
REP—
LEVELL0
Qualification not yet issued
Formal-license routeVerified federation credential and current status.
Experience routeDocumented history + evidence + practical assessment.
SSOT qualification routeAI-Augmented Coach curriculum + practical review +
safeguarding.
Player Passport ↔ Coach Passport
The Player Twin expresses development needs. The Coach Passport expresses verified qualification, specialties and demonstrated outcomes. Together they become trusted inputs to Player × Coach × Intervention matching.
Important: SSOT verifies evidence and maintains a dynamic
coaching identity. It should not imply that SSOT replaces or issues federation licenses.
SSOT FOUNDING COACH PROGRAM · FAQ v1.0
Read This Before You Apply.
The Founding Coach Program is an active AI-augmented player-development pilot—not simply an app signup. You do not need to be an AI expert, but you should be willing to work with real players, evidence and AI recommendations while retaining your own professional coaching judgment.
1. What is the Founding Coach Program?
An initial international coaching community helping SSOT test Player × Human Coach × AI development. Coaches help evaluate player evidence, AI recommendations and real outcomes.
2. What am I committing to?
If selected, the initial commitment is an approximately 30-day structured pilot. Detailed timing, player expectations and onboarding requirements are provided before participation begins.
3. Do I need to be an AI expert?
No. We are looking for football coaches. Curiosity, coaching competence, openness to evidence and willingness to question AI matter more than technical AI expertise.
4. Is AI intended to replace the coach?
No. AI proposes; the human coach decides. Coaches can ACCEPT, MODIFY or REJECT AI suggestions and explain why.
5. What if I disagree with the AI?
Disagreement is useful evidence. SSOT aims to learn from the coach's reasoning, the intervention actually used and the player's subsequent outcome.
6. What is a Player Development Twin?
An evolving, evidence-based representation of a player's goals, observations, training history, coach decisions, AI recommendations and development over time. It is not intended as a permanent label.
7. What do I gain as a coach?
Practical AI-augmented coaching experience, a more structured player-development evidence record, and the opportunity to help shape SSOT's future coaching system. Future commercial opportunities are not guaranteed.
8. What happens after 30 days?
SSOT reviews participation, evidence quality, player outcomes, coach feedback, AI quality, safety and trust. Any later-stage participation is offered separately; applying now is not an 18–24 month commitment.
9. How are young players protected?
The platform is designed around appropriate consent, verified adult roles, permissioned access, defined coach-player relationships, provenance, auditability, revocation and safeguarding controls.
10. Who can see player videos and data?
Access should follow player or guardian permissions and legitimate roles. Private coaching evidence should not automatically become public or visible to unrelated coaches, scouts or organizations.
11. Will players or coaches be publicly ranked by AI?
The objective is development, not permanent labeling. Metrics may support evaluation and progress tracking, but should be interpreted in context with human judgment.
12. How are coaches selected?
The application considers coaching competence, commitment to player development, openness to AI/evidence, pilot-player access, communication, digital/video readiness and potential contribution. Automated assessment supports human review; it is not the final decision.
13. What happens after I apply?
Invitation → Application → Initial assessment → Human review → Follow-up/verification where needed → Pilot selection → Onboarding → Initial AI-Augmented Coaching Program.
14. What is the longer-term vision?
SSOT may progress from AI Development Twins and Triadic Coaching to Smart Academies, Coaching Intelligence Networks and eventually regional, national and global Football Intelligence League programs. These are future stages, not today's commitment.
Ready after reading the FAQ? You can start the application now.Start 10-Minute Application →
v1.2 · COMPLETE MVP PROGRAM
18–24 Month Multi-Level AI-Augmented Human Coaching Program
The next step is to turn the seven-level model into one operating progression: every level must generate real human coaching decisions, longitudinal player outcomes, provenance, trust evidence and an explicit graduation decision before scale.
HORIZON18–24 months
ARCHITECTURE7 levels
CORE UNITTriadic episode
SCALE RULEEvidence-gated
PROGRAM PHASES
Build depth before geographic scale
Complete-MVP definition: L1–L3 operate with real users and longitudinal
outcomes; L4 demonstrates cross-academy replication; L5 runs a real regional FIL. L6–L7 may begin as
governed pilots, but their data, trust and promotion contracts must work end-to-end.
Recommended immediate focus: Months 0–6
Founding 100 → active coach/player pairs → 8–12 week Development Twins → measurable ACCEPT/MODIFY/REJECT episodes → Academy Pilot. This evidence base should precede serious MARL and FIL scaling.
HUMAN COACHINGAccept · Modify · RejectCapture judgment and
rationale, not only AI output.
PLAYER DEVELOPMENTLongitudinal ΔRepeated evidence, transfer
and retention.
NETWORK LEARNINGReplicate → PromoteLocal lessons become
network knowledge only after validation.
TRUST & SAFETYHard boundaryIdentity, consent, safeguarding
and authority never enter RL action space.
v1.3 · MULTI-LEVEL COACHING CONTROL TOWER
From Founding Coach Activity to Global FIL Readiness
One operational cockpit tracks coach activation, player-development evidence, triadic human–AI decisions, academy replication, FIL readiness and the next graduation action. Network scale advances only when evidence and human-governed gates are satisfied.
SIMULATED OPERATING DATA · MVP CONTROL LOGIC
ACTIVE FOUNDING COACHES42Target for next
scale gate: 50+
ACTIVE DEVELOPMENT TWINS218Coach ↔ player
longitudinal relationships
TRIADIC COACHING EPISODES1,486AI proposal
+ human decision + outcome
LONGITUDINAL EVIDENCE COMPLETE64%Target
for green expansion: ≥70%
L1–L7 Network Readiness
Click a level to inspect the evidence contract, graduation gate and immediate action
required.
Graduation Gate Monitor
These gates are intentionally stricter than registration growth. Weak evidence pauses
scale.
Triadic Coaching Evidence Ledger
| Episode | AI proposal | Human decision | Outcome | Evidence |
|---|
Next Unlock Queue
Control Tower prioritizes what the organization should do next—not merely what the model
can technically do.
v1.4 · L1 OPERATING MANAGEMENT SYSTEM
L1 · Month 0–3: AI Development Twin
Run L1 as a 12-week evidence program, not a feature launch. The objective is to prove that SSOT can maintain a trusted player state, recommend bounded next actions, preserve human authority, measure real outcomes and improve the next recommendation.
PILOT PLAYERS20Start small enough for review
FOUNDING COACHES4≈5 players / coach
DURATION12 weeksBaseline → retention
MIN SESSIONS8 / playerTarget ≥160 episodes
PRIMARY GATEMeasurable ΔWith evidence + safety
12-Week Operating Cadence
Select a phase to inspect its deliverables and exit condition.
Player Development Twin · Canonical Record
Every recommendation must be reconstructable from the state that existed when it was
made.
STATEAge band · position · goals · baseline · workload · recent evidence
PROPOSALDrill · difficulty · repetitions · expected signal · confidence
HUMAN GATECoach/guardian accept · modify · reject · reason
OUTCOMEMeasured delta · quality · transfer · retention · safety
| Field | Example | Required? |
|---|---|---|
| Player goal | Improve first-touch control | YES |
| Baseline evidence | 3 repeated clips + score distribution | YES |
| AI recommendation | Add moving receiver at bounded difficulty | YES |
| Human decision | MODIFY · reduce speed | YES |
| Outcome | Control +6%; latency −5% | YES |
| Retention check | 7–14 days later | YES |
L1 Graduation Dashboard
Illustrative pilot controls; thresholds should be frozen before the cohort begins.
| Gate | Target | Demo | Status |
|---|
L1 → L2 DECISIONAMBER · Continue evidence
collection
Do not graduate because engagement is high. Graduate only when repeated outcomes, evidence completeness, human-review quality and safety are jointly credible.
PRODUCT OWNERPlayer Development LoopOwn onboarding, Player Twin,
session UX, recommendation display and re-test.
COACHING LEADHuman AuthorityDefine bounded interventions, reason
codes, escalation and coach-review quality.
DATA / AI LEADEvidence ContractVersion observations, proposals,
decisions, outcomes and model/policy identifiers.
SCIENCE LEADValidityPredefine constructs, repeated measures,
retention and transfer tests.
TRUST / SAFETYConsent & SafeguardingGuardian permissions, data
minimization, adult verification, revocation and incident handling.
CONTROL TOWERWeekly Gate ReviewTrack active pairs, missing
evidence, anomalies, retention, safety and next actions.
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.