PICM — Predictive Influence & Concentration Module
Architecture Ecosystem: Structured Reality Standards™
Architecture Family: PGA™ — Prediction Governance Architecture
Parent Standard:
Collective Prediction Governance Standard (CPGS)
Operational Layer: Collective Prediction Governance Layer
Category:
Governance & Enforcement
Subcategory:
Prediction Governance
Type: Parent Standard Module
Governed Space:
Predictive Influence & Concentration
Version: 1.0
Status: Canonical · Open Module
Effective Date: 9 September 2026
Compatibility: OOF® Methodology OS™ · GOA™ · ORA™ · VALIDOS™ · INTEGROS® · ADIT® · AGA™ · AIG® · ASGA™ · CLIA® · SIMULOS®
AI-Readable: Yes
Authority: OOF®
Protection:
MIP® — Methodological Intellectual Property
Canonical Language:
English (UCL™)
Governed Space
Predictive Influence & Concentration
Canonical Definition
Predictive Influence & Concentration Module (PICM) defines the governance conditions under which influence over an aggregated predictive representation is identified, attributed, measured or qualified, structurally interpreted, and assessed for material concentration before the aggregate may proceed toward final Collective Prediction Representation.
PICM determines whether apparent collective participation corresponds to materially distributed predictive influence or whether the resulting aggregate is disproportionately shaped by particular:
predictive sources; source families; dependence clusters; model families; institutions; information origins; weighting structures; capital positions; market participants; Predictor–Reality Proximity positions; decision-proximate or outcome-informed sources.
It transforms:
Governed Prediction Aggregation + Governed Predictive Source Diversity Profile + Governed Prediction Dependence & Variation Profile + Contribution Weights + Relevant Influence Relationships
into:
Governed Predictive Influence & Concentration Profile
Its foundational principle is:
Broad participation does not establish distributed predictive influence.
Operational Role
PICM is the fourth internal governance space of the Collective Prediction Governance Standard (CPGS) .
PSDM establishes:
what predictive source diversity exists.
PDVM establishes:
how dependent or independent the predictive contributions are and how they vary.
PAGM establishes:
how eligible predictive contributions are aggregated.
PICM establishes:
who or what materially determines the aggregated result.
The focused operational question is:
How is predictive influence distributed across the collective structure, where is it concentrated, and does the apparent collective result materially depend upon a limited number of sources, clusters, information origins, models, capital positions, or other influence structures?
The primary output is:
Governed Predictive Influence & Concentration Profile
Internal Module Flow
Predictive Source Diversity → Prediction Dependence & Variation → Prediction Aggregation → Predictive Influence & Concentration → Collective Prediction Representation
Current Position: Predictive Influence & Concentration
Predictive Influence
Predictive Influence represents the material effect that a predictive contribution, source, cluster, methodology, information origin, weighting structure, or other predictive component has upon an aggregated predictive representation.
Participation ≠ Influence
A source may participate while having little effect on the final aggregate.
Another source may represent only one contribution while materially determining the result.
Nominal Participation
Nominal Participation represents presence within the collective prediction mechanism.
Effective Influence
Effective Influence represents the actual material contribution to the resulting aggregate.
Nominal Participation ≠ Effective Influence
Influence Distribution
Influence Distribution describes how predictive influence is allocated across the contributing structure.
Distributed Influence
Influence is relatively distributed where no small source, cluster, or structural origin disproportionately determines the aggregate.
Distributed Participation ≠ Distributed Influence
Influence Concentration
Influence Concentration exists where a limited number of predictive sources, source clusters, models, institutions, capital positions, information origins, or other structural components disproportionately shape the aggregate.
Large Collective ≠ Low Influence Concentration
Source-Level Influence
Influence may be attributed to individual predictive sources.
Source Count ≠ Source Influence
Contribution-Level Influence
A specific predictive contribution may exert material influence because of:
assigned weight; extreme predictive position; aggregation method; timing; market capital; methodological priority.
Contribution Weight ≠ Complete Influence
A nominal weight does not always capture actual effect on the aggregate.
Marginal Influence
Marginal Influence represents how materially the aggregated result changes when a contribution is modified, removed, or reduced.
High Weight ≠ High Marginal Influence by Default
Removal Sensitivity
Removal Sensitivity measures or qualifies how the aggregate changes when a predictive source or contribution is removed.
One Contribution Removed → Material Aggregate Change
may indicate concentrated influence.
Removal Sensitivity ≠ Prediction Error
Source-Family Influence
Multiple nominal sources may belong to one effective source family.
Their combined influence should remain governable.
Many Interfaces + One Model Family ≠ Distributed Model Influence
Model-Family Concentration
Model-Family Concentration exists where a collective prediction is materially shaped by one model family despite multiple model instances or services.
Multiple Models ≠ Distributed Model Influence
Agent Concentration
Multiple agents may share a common predictive origin.
Agent Count ≠ Distributed Agent Influence
Institutional Concentration
Predictive influence may be concentrated within one institution or organizational network.
Multiple Analysts ≠ Distributed Institutional Influence
Ownership Concentration
Different sources may share common ownership or control.
Different Legal Entities ≠ Distributed Ownership Influence
Methodological Concentration
A collective may contain many sources while most effective influence comes from one methodology.
Source Diversity ≠ Methodological Influence Diversity
Information-Origin Concentration
Many predictive contributions may ultimately depend upon one information origin.
Many Predictions ≠ Many Information Origins
Upstream Influence
An upstream information source may indirectly influence many downstream predictions.
Upstream Source Not Directly Included ≠ No Upstream Influence
Hidden Influence
Influence may exist even where the influential source does not directly participate in aggregation.
Direct Contribution ≠ Only Form of Predictive Influence
Dependence-Cluster Influence
Predictions grouped into a dependence cluster may collectively exert disproportionate influence.
Cluster Size ≠ Independent Influence
Cluster Weight
The combined effective influence of a dependence cluster should remain governable.
Nominal Contribution Weight ≠ Independent Cluster Weight
Influence Duplication
Influence Duplication occurs where one predictive origin enters the aggregate through multiple downstream contributions.
Repeated Pathways ≠ Independent Influence
Influence Amplification
Influence Amplification occurs where aggregation structure causes one predictive origin to receive greater effective influence than its underlying informational contribution would suggest.
Aggregation Weight ≠ Informational Contribution
Influence Attenuation
Influence may also be reduced through:
dependence adjustment; weight caps; segmentation; normalization; methodological rules.
Reduced Influence ≠ Invalid Prediction
Influence Transfer
Influence may move from one source to another through intermediate predictive systems.
Example:
Expert Forecast → AI Retrieval → Multiple AI Predictions → Aggregate
The expert may exert substantial indirect influence even though only AI predictions appear as direct contributors.
Intermediate Transformation ≠ Elimination of Upstream Influence
Influence Chain
An Influence Chain describes how predictive influence propagates across sources or systems.
Long Influence Chain ≠ Independent Influence
Influence Network
A collective predictive system may contain a network of direct and indirect influence relationships.
Influence Network ≠ Contribution List
Influence Centrality
Some sources may occupy structurally central positions within an influence network.
Centrality ≠ Predictive Accuracy
Dominant Predictive Source
A Dominant Predictive Source is a source whose effective influence materially shapes the collective result.
Dominant Source ≠ Correct Source
Dominant Source Cluster
A Dominant Source Cluster is a group of structurally related predictive contributions that collectively exert material influence over the aggregate.
Dominant Cluster ≠ Majority of Participants
Dominant Information Origin
A Dominant Information Origin exists where substantial collective influence traces back to one information source or environment.
Information-Origin Dominance ≠ Source-Count Dominance
Influence Concentration Across Dimensions
Concentration may exist across different dimensions:
source; institution; model; methodology; information origin; geography; capital; Predictor–Reality Proximity; decision authority.
Low Concentration in One Dimension ≠ Low Overall Concentration
Multi-Dimensional Concentration
A collective may appear distributed across one dimension while remaining concentrated across another.
Example:
100 traders may represent many individuals but most capital may belong to five accounts.
Participant Diversity ≠ Capital Diversity
Capital Influence
In market-based prediction systems, capital may determine contribution influence.
Capital ≠ Predictive Merit
Capital-Weighted Influence
A participant with greater capital may move a market-derived predictive signal more strongly than many smaller participants.
One Trader ≠ One Unit of Influence
Market Participation ≠ Equal Predictive Influence
Market Concentration
Market Concentration exists where a limited number of positions materially shape the market-derived predictive signal.
Number of Traders ≠ Distribution of Market Influence
Liquidity and Influence
Low liquidity may allow relatively small positions to materially change a market signal.
Small Trade ≠ Small Influence in Low-Liquidity Market
High Liquidity ≠ Distributed Influence by Default
Position Size
Position size may contribute to market influence.
Position Size ≠ Predictive Reliability
Trading Frequency
Repeated trading may increase apparent activity without increasing independent predictive participation.
Trade Count ≠ Predictor Count
Account Count
Multiple accounts may belong to one actor or coordinated group.
Account Count ≠ Independent Participant Count
Coordinated Influence
Multiple participants may coordinate their predictive actions.
Multiple Accounts ≠ Distributed Influence
Strategic Influence
Participants may intentionally attempt to shift the collective signal.
Signal Influence ≠ Predictive Accuracy
Market Manipulation Determination
PICM may identify influence patterns requiring further governance review.
It does not determine legal market manipulation.
Influence Pattern ≠ Legal Misconduct Determination
Predictor–Reality Proximity
Predictor–Reality Proximity may materially affect interpretation of influence.
Relevant positions may include:
Public-Information Predictor
Information-Advantaged Predictor
Decision-Proximate Predictor
Outcome-Informed Predictor
Outcome-Influencing Predictor
Outcome-Controlling Predictor
High Predictor–Reality Proximity ≠ Misconduct
Predictor–Reality Proximity ≠ Predictive Influence
A predictor may be close to the predicted reality while having little influence over the collective aggregate.
These dimensions must remain separate.
Outcome Influence
A predictive source may influence the real-world outcome itself.
Predictive Influence ≠ Outcome Influence
Predictive Influence concerns influence over the collective prediction.
Outcome Influence concerns influence over the predicted reality.
Outcome Influence ≠ Predictive Accuracy
Outcome Control
A predictor may possess authority capable of materially determining the outcome.
Outcome Control ≠ Predictive Skill
Decision Authority
Decision authority should remain distinguishable from predictive influence.
Decision Authority ≠ Collective Predictive Legitimacy
Outcome-Informed Influence
A source with privileged outcome information may strongly influence the collective prediction.
Outcome Knowledge ≠ Predictive Capability
Influence from Informational Advantage
Information advantage may legitimately cause a source to contribute valuable information.
However, the collective representation should preserve where a supposedly broad consensus is materially driven by one informationally advantaged source.
Information Advantage ≠ Distributed Collective Intelligence
Predictive Reputation Influence
Sources with strong reputations may influence other contributors.
Reputation ≠ Predictive Accuracy by Default
Authority Signaling
Participants may follow predictions because of perceived authority.
Authority Signaling ≠ Independent Confirmation
Consensus Influence
Existing consensus may influence later predictions.
Consensus as Input ≠ Independent Consensus
Feedback Influence
The aggregate itself may become an input into future predictive contributions.
Collective Prediction → Contributor Update → Collective Prediction
This can create feedback loops.
Feedback Loop ≠ New Independent Information
Reflexive Collective Influence
A collective prediction may influence both:
future predictions; the predicted reality.
Predictive Reflexivity
Predictive reflexivity occurs where the collective prediction alters subsequent predictive behavior.
Operational Reflexivity
Operational reflexivity occurs where the collective prediction contributes to changes in Operational Reality.
Predictive Reflexivity ≠ Operational Reflexivity
Self-Reinforcing Prediction Structure
A collective signal may become stronger because contributors repeatedly respond to the signal itself.
Growing Consensus ≠ Growing Independent Evidence
Self-Weakening Prediction Structure
A collective prediction may cause actors to take preventive action, reducing the probability of the predicted outcome.
Prediction Influence ≠ Prediction Failure
Influence Threshold
An Influence Threshold may define when a source, cluster, or structural origin becomes materially influential.
Threshold ≠ Universal Constant
The appropriate threshold depends upon the aggregation mechanism and governance purpose.
Concentration Threshold
A Concentration Threshold may define when influence concentration requires qualification or further governance treatment.
Concentration Threshold ≠ Automatic Illegitimacy
Influence Share
Influence Share represents the proportion of effective predictive influence attributable to a source or structure where measurable.
Influence Share ≠ Contribution Count Share
Effective Influence Share
Effective Influence Share may account for:
nominal weight; dependence; indirect influence; source-family relationships; aggregation sensitivity.
Nominal Weight Share ≠ Effective Influence Share
Effective Predictor Count
An Effective Predictor Count may estimate how many materially independent influence-bearing predictive sources exist after relevant dependence and concentration are considered.
Raw Predictor Count ≠ Effective Predictor Count
Effective Predictor Count ≠ Exact Number of Independent Minds
It is a governance representation, not necessarily a literal count.
Concentration Ratio
Where useful, concentration may be represented through defined ratios.
Concentration Ratio ≠ Complete Influence Structure
A single metric should not replace structural interpretation.
Influence Distribution Profile
An Influence Distribution Profile represents the distribution of effective predictive influence across relevant dimensions.
Influence Distribution Profile ≠ Predictive Performance Profile
Concentration State
PICM may classify concentration as:
Broadly Distributed
Moderately Concentrated
Materially Concentrated
Highly Concentrated
Dominantly Concentrated
Concentration Uncertain
Unable to Determine
Broadly Distributed
No limited source or cluster materially dominates the aggregate under the applicable methodology.
Moderately Concentrated
Some concentration exists but the aggregate retains substantial distributed influence.
Materially Concentrated
A limited set of sources or structures has substantial influence requiring explicit interpretation.
Highly Concentrated
A small number of sources or structures strongly determine the aggregate.
Dominantly Concentrated
The aggregate is effectively controlled or determined by one source, cluster, information origin, or equivalent influence structure.
Dominantly Concentrated ≠ Invalid Prediction by Default
The state requires transparent representation.
Concentration Uncertain
Material concentration may exist but cannot be sufficiently resolved.
Unable to Determine
Available information does not support a legitimate concentration determination.
Unable to Determine ≠ Distributed Influence
Influence Qualification
Where concentration is material, the aggregate should preserve an Influence Qualification.
Example:
Aggregate materially influenced by one model family representing approximately 64% of effective predictive influence.
Qualification ≠ Disqualification
Concentration Adjustment
Where methodology permits, aggregation may be reconsidered when excessive concentration is identified.
PICM identifies the concentration.
It does not silently rewrite PAGM methodology.
Concentration Identified ≠ Automatic Reweighting
Aggregation Reassessment Trigger
Material concentration may trigger reconsideration of aggregation methodology.
Reassessment Trigger ≠ Automatic Aggregation Change
Influence Conflict
Different influence-analysis methods may produce materially different concentration conclusions.
Influence Conflict ≠ Average the Conclusions
Influence Uncertainty
Influence may be difficult to determine where:
ownership is opaque; source relationships are incomplete; market positions are hidden; model provenance is unknown; indirect influence chains are unresolved.
Unknown Influence ≠ No Influence
Hidden Concentration
Hidden Concentration exists where apparent distributed participation conceals common ownership, common origin, coordinated activity, or indirect structural dominance.
Visible Diversity ≠ Distributed Influence
Influence History
Influence structures may change over time.
Current Influence ≠ Historical Influence
Historical Concentration
A collective prediction's historical interpretation should preserve the concentration structure existing when the aggregate was formed.
Later Distribution ≠ Historical Distribution
Influence Version
Material changes in influence structure should create a new governable version.
Influence Change ≠ Historical Replacement
Source Entry
A new source may materially change influence distribution.
New Source ≠ Material Influence Change by Default
Source Exit
Removing a dominant source may materially alter collective structure.
Source Removal ≠ Proportional Influence Reduction by Default
Weight Change
Changes in aggregation weights may alter effective influence.
Weight Change ≠ Predictive Source Change
Dependence Change
New information about predictive dependence may change the interpretation of effective influence.
Same Aggregate Number ≠ Same Influence Structure
Collective Stability
An aggregate may remain numerically stable while the sources determining it change substantially.
Numerical Stability ≠ Structural Stability
Relationship to PSDM
PSDM governs Predictive Source Diversity.
PICM uses the source structure to determine how influence is distributed.
Source Diversity ≠ Influence Diversity
Relationship to PDVM
PDVM governs Prediction Dependence & Variation.
PICM uses dependence relationships to identify duplicated, clustered, indirect, or concentrated influence.
Prediction Dependence ≠ Predictive Influence
Relationship to PAGM
PAGM establishes aggregation method, contribution weights, and dependence adjustments.
PICM assesses the resulting effective influence structure.
Aggregation Weight ≠ Complete Influence Structure
Relationship to CPRM
PICM provides the influence and concentration structure required before final Collective Prediction Representation.
CPRM determines how the collective result may legitimately be represented.
Influence Assessment ≠ Final Collective Representation
Relationship to Predictor–Reality Proximity
Predictor–Reality Proximity may be preserved where material to understanding concentrated influence.
High Predictor–Reality Proximity ≠ Misconduct
Relationship to Prediction Markets
PICM governs predictive influence and concentration within market-derived collective signals.
It does not determine securities law, gambling law, market-abuse law, or other jurisdiction-specific legal conclusions.
Influence Pattern ≠ Legal Conclusion
Relationship to AI Ensembles
PICM identifies whether an apparent multi-model collective is materially dominated by:
one model family; one upstream model; one retrieval system; one orchestration layer; one information source.
Model Count ≠ Model Influence Distribution
Relationship to ORA™
ORA™ governs Operational Reality.
PICM does not establish whether a predictor actually controls or changes Operational Reality.
It governs the relevance of such relationships to collective predictive influence.
Governance Requirements
PICM requires that:
1. Predictive Influence remain distinguishable from participation; 2. Nominal Participation remain distinguishable from Effective Influence; 3. distributed participation not automatically establish distributed influence; 4. Influence Concentration remain identifiable; 5. large collective size not automatically imply low concentration; 6. source count remain distinguishable from source influence; 7. contribution weight remain distinguishable from complete influence; 8. Marginal Influence remain assessable where material; 9. Removal Sensitivity remain assessable where appropriate; 10. removal sensitivity remain distinguishable from prediction error; 11. source-family influence remain governable; 12. model-family concentration remain identifiable; 13. multiple models not automatically establish distributed model influence; 14. agent count remain distinguishable from distributed agent influence; 15. institutional concentration remain identifiable; 16. ownership concentration remain identifiable where material; 17. different legal entities not automatically establish distributed ownership influence; 18. methodological concentration remain governable; 19. information-origin concentration remain identifiable; 20. upstream influence remain governable; 21. indirect influence remain preservable; 22. direct contribution not be treated as the only form of influence; 23. dependence-cluster influence remain identifiable; 24. cluster size remain distinguishable from independent influence; 25. nominal contribution weight remain distinguishable from independent cluster weight; 26. Influence Duplication remain identifiable; 27. repeated pathways not automatically create independent influence; 28. Influence Amplification remain governable; 29. aggregation weight remain distinguishable from informational contribution; 30. Influence Attenuation remain traceable where material; 31. reduced influence not imply invalid prediction; 32. Influence Transfer remain governable; 33. intermediate transformation not automatically eliminate upstream influence; 34. Influence Chains remain traceable where material; 35. long influence chains not establish independence;
36. Influence Networks remain representable where useful; 37. Influence Centrality remain distinguishable from predictive accuracy; 38. dominant predictive sources remain identifiable; 39. dominant source clusters remain identifiable; 40. dominant information origins remain identifiable; 41. concentration across multiple dimensions remain assessable; 42. low concentration in one dimension not establish low overall concentration; 43. Participant Diversity remain distinguishable from Capital Diversity; 44. capital influence remain distinguishable from predictive merit; 45. market participation remain distinguishable from equal predictive influence; 46. market concentration remain identifiable; 47. trader count remain distinguishable from influence distribution; 48. liquidity effects on influence remain governable; 49. high liquidity not automatically establish distributed influence; 50. position size remain distinguishable from predictive reliability; 51. trade count remain distinguishable from predictor count; 52. account count remain distinguishable from independent participant count; 53. coordinated influence remain identifiable where material; 54. strategic influence remain distinguishable from predictive accuracy; 55. influence patterns remain distinguishable from legal misconduct determinations; 56. Predictor–Reality Proximity remain distinguishable from Predictive Influence; 57. high Predictor–Reality Proximity not imply misconduct; 58. Predictive Influence remain distinguishable from Outcome Influence; 59. Outcome Influence remain distinguishable from predictive accuracy; 60. Outcome Control remain distinguishable from predictive skill; 61. Decision Authority remain distinguishable from collective predictive legitimacy; 62. outcome knowledge remain distinguishable from predictive capability; 63. Information Advantage remain distinguishable from distributed collective intelligence; 64. reputation influence remain governable; 65. authority signaling remain distinguishable from independent confirmation; 66. consensus influence remain identifiable; 67. feedback influence remain governable; 68. feedback loops remain distinguishable from new independent information; 69. Predictive Reflexivity remain distinguishable from Operational Reflexivity; 70. growing consensus not automatically establish growing independent evidence; 71. Influence Thresholds remain explicit where used; 72. Concentration Thresholds remain explicit where used; 73. threshold exceedance not automatically establish illegitimacy; 74. Influence Share remain distinguishable from Contribution Count Share; 75. nominal weight share remain distinguishable from Effective Influence Share; 76. raw predictor count remain distinguishable from Effective Predictor Count; 77. Effective Predictor Count not be represented as an exact number of independent minds by default; 78. concentration metrics remain distinguishable from complete influence structure; 79. concentration states remain explicit; 80. Dominantly Concentrated remain distinguishable from invalidity; 81. Unable to Determine remain available as a legitimate state; 82. Unable to Determine not be interpreted as distributed influence; 83. material concentration remain qualified; 84. qualification remain distinguishable from disqualification; 85. concentration identification not automatically rewrite aggregation weights; 86. aggregation reassessment triggers remain distinguishable from automatic aggregation change; 87. influence conflicts remain preservable; 88. unknown influence not be treated as no influence; 89. Hidden Concentration remain identifiable; 90. visible diversity remain distinguishable from distributed influence; 91. historical influence structures remain reconstructable; 92. later influence distribution not overwrite historical distribution; 93. influence changes remain versionable; 94. source entry not automatically imply material influence change; 95. source removal not imply proportional influence reduction; 96. weight changes remain distinguishable from source changes; 97. dependence changes remain capable of changing influence interpretation; 98. identical aggregate numbers remain distinguishable where influence structures differ; 99. numerical stability remain distinguishable from structural stability; 100. Predictive Influence & Concentration remain distinct from final Collective Prediction Representation.
Governance Logic
PICM applies the following governance logic:
Governed Prediction Aggregation
Governed Predictive Source Diversity Profile
Governed Prediction Dependence & Variation Profile
Contribution Weights / Dependence Adjustments
Relevant Source / Market / Model / Influence Relationships
↓
Nominal Participation Structure Established
↓
Effective Contribution Structure Established
↓
Direct Predictive Influence Assessed
↓
Indirect / Upstream Influence Assessed
↓
Source-Family Influence Assessed
↓
Dependence-Cluster Influence Assessed
↓
Model / Agent / Institutional Concentration Assessed where applicable
↓
Information-Origin Concentration Assessed
↓
Ownership / Methodological Concentration Assessed where material
↓
Capital / Market Influence Assessed where applicable
↓
Predictor–Reality Proximity Context Preserved where material
↓
Predictive Influence Distinguished from Outcome Influence
↓
Influence Chains / Networks Assessed where necessary
↓
Influence Duplication / Amplification Assessed
↓
Marginal Influence / Removal Sensitivity Assessed where material
↓
Effective Influence Shares Established where determinable
↓
Effective Predictor Structure Established
↓
Concentration Across Relevant Dimensions Assessed
↓
Hidden Concentration Assessed
↓
Reflexive / Feedback Influence Assessed where material
↓
Influence Qualification Established
↓
Concentration State Established
↓
Historical Influence Structure Preserved
↓
Governed Predictive Influence & Concentration Profile
↓
Collective Prediction Representation
Where sufficient determination cannot be established:
Predictive Influence & Concentration Governance Gap → Influence Relationship Clarification / Source-Origin Analysis / Market Concentration Analysis / Model-Family Analysis / Concentration Qualification / Concentration Uncertain / Unable to Determine
Minimum Implementation Framework
A minimum implementation of PICM must establish:
A. Collective Prediction Reference
The candidate collective prediction or aggregated predictive representation.
B. Influence-Bearing Source / Structure
The source, contribution, cluster, model family, institution, information origin, market position, or other influence-bearing structure.
C. Influence Type
Direct, indirect, upstream, clustered, capital-based, model-based, informational, reflexive, or other relevant influence.
D. Nominal Contribution Weight
The weight assigned through aggregation where applicable.
E. Effective Influence
The materially interpreted influence after relevant dependence and structural relationships are considered.
F. Influence Chain / Relationship
Where indirect influence is material.
G. Influence Share
Where determinable.
H. Concentration Dimension
Source, model, institution, information origin, methodology, ownership, capital, or other relevant dimension.
I. Dominant Source / Cluster
Where identifiable.
J. Removal / Marginal Sensitivity
Where materially useful.
K. Predictor–Reality Proximity Context
Where relevant to interpretation.
L. Influence Qualification
Any material limitation, concentration, hidden dependence, or uncertainty.
M. Concentration State
At minimum:
Broadly Distributed / Moderately Concentrated / Materially Concentrated / Highly Concentrated / Dominantly Concentrated / Concentration Uncertain / Unable to Determine
N. Influence & Concentration Rationale
A traceable explanation supporting the state.
O. Historical Influence Version
The reconstructable influence and concentration structure applicable when the aggregate existed.
Governance Outputs
PICM may produce:
Governed Predictive Influence & Concentration Profile; Influence Distribution Profile; Nominal Participation Structure; Effective Influence Structure; Source-Level Influence Record; Contribution-Level Influence Record; Marginal Influence Record; Removal Sensitivity Record; Source-Family Influence Record; Model-Family Concentration Record; Agent Concentration Record; Institutional Concentration Record; Ownership Concentration Record; Methodological Concentration Record; Information-Origin Concentration Record; Upstream Influence Record; Hidden Influence Record; Dependence-Cluster Influence Record; Cluster Weight Record; Influence Duplication Record; Influence Amplification Record; Influence Attenuation Record; Influence Transfer Record; Influence Chain; Influence Network; Influence Centrality Record; Dominant Predictive Source; Dominant Source Cluster; Dominant Information Origin; Multi-Dimensional Concentration Profile; Capital Influence Record; Market Concentration Record; Liquidity Influence Record; Coordinated Influence Record; Strategic Influence Record; Predictor–Reality Proximity Context; Outcome Influence Distinction; Consensus Influence Record; Feedback Influence Record; Predictive Reflexivity Record; Operational Reflexivity Reference; Influence Threshold; Concentration Threshold; Influence Share; Effective Influence Share; Effective Predictor Count; Concentration Ratio; Concentration State; Influence Qualification; Influence Conflict Record; Influence Uncertainty Record; Hidden Concentration Record; Predictive Influence & Concentration Governance Gap; historical influence state; Influence Version.
The primary governed output is:
Governed Predictive Influence & Concentration Profile
which becomes an input to:
Collective Prediction Representation.
Critical Distinctions
Broad Participation Does Not Establish Distributed Predictive Influence
Participation ≠ Influence
Nominal Participation ≠ Effective Influence
Distributed Participation ≠ Distributed Influence
Large Collective ≠ Low Influence Concentration
Source Count ≠ Source Influence
Contribution Weight ≠ Complete Influence
High Weight ≠ High Marginal Influence by Default
Removal Sensitivity ≠ Prediction Error
Many Interfaces + One Model Family ≠ Distributed Model Influence
Multiple Models ≠ Distributed Model Influence
Agent Count ≠ Distributed Agent Influence
Multiple Analysts ≠ Distributed Institutional Influence
Different Legal Entities ≠ Distributed Ownership Influence
Source Diversity ≠ Methodological Influence Diversity
Many Predictions ≠ Many Information Origins
Upstream Source Not Directly Included ≠ No Upstream Influence
Direct Contribution ≠ Only Form of Predictive Influence
Cluster Size ≠ Independent Influence
Nominal Contribution Weight ≠ Independent Cluster Weight
Repeated Pathways ≠ Independent Influence
Aggregation Weight ≠ Informational Contribution
Reduced Influence ≠ Invalid Prediction
Intermediate Transformation ≠ Elimination of Upstream Influence
Long Influence Chain ≠ Independent Influence
Influence Network ≠ Contribution List
Centrality ≠ Predictive Accuracy
Dominant Source ≠ Correct Source
Dominant Cluster ≠ Majority of Participants
Information-Origin Dominance ≠ Source-Count Dominance
Low Concentration in One Dimension ≠ Low Overall Concentration
Participant Diversity ≠ Capital Diversity
Capital ≠ Predictive Merit
One Trader ≠ One Unit of Influence
Market Participation ≠ Equal Predictive Influence
Number of Traders ≠ Distribution of Market Influence
Small Trade ≠ Small Influence in Low-Liquidity Market
High Liquidity ≠ Distributed Influence by Default
Position Size ≠ Predictive Reliability
Trade Count ≠ Predictor Count
Account Count ≠ Independent Participant Count
Multiple Accounts ≠ Distributed Influence
Signal Influence ≠ Predictive Accuracy
Influence Pattern ≠ Legal Misconduct Determination
High Predictor–Reality Proximity ≠ Misconduct
Predictor–Reality Proximity ≠ Predictive Influence
Predictive Influence ≠ Outcome Influence
Outcome Influence ≠ Predictive Accuracy
Outcome Control ≠ Predictive Skill
Decision Authority ≠ Collective Predictive Legitimacy
Outcome Knowledge ≠ Predictive Capability
Information Advantage ≠ Distributed Collective Intelligence
Reputation ≠ Predictive Accuracy by Default
Authority Signaling ≠ Independent Confirmation
Consensus as Input ≠ Independent Consensus
Feedback Loop ≠ New Independent Information
Predictive Reflexivity ≠ Operational Reflexivity
Growing Consensus ≠ Growing Independent Evidence
Concentration Threshold ≠ Automatic Illegitimacy
Influence Share ≠ Contribution Count Share
Nominal Weight Share ≠ Effective Influence Share
Raw Predictor Count ≠ Effective Predictor Count
Effective Predictor Count ≠ Exact Number of Independent Minds
Concentration Ratio ≠ Complete Influence Structure
Influence Distribution Profile ≠ Predictive Performance Profile
Dominantly Concentrated ≠ Invalid Prediction by Default
Unable to Determine ≠ Distributed Influence
Qualification ≠ Disqualification
Concentration Identified ≠ Automatic Reweighting
Reassessment Trigger ≠ Automatic Aggregation Change
Influence Conflict ≠ Average the Conclusions
Unknown Influence ≠ No Influence
Visible Diversity ≠ Distributed Influence
Current Influence ≠ Historical Influence
Later Distribution ≠ Historical Distribution
Influence Change ≠ Historical Replacement
New Source ≠ Material Influence Change by Default
Source Removal ≠ Proportional Influence Reduction by Default
Weight Change ≠ Predictive Source Change
Same Aggregate Number ≠ Same Influence Structure
Numerical Stability ≠ Structural Stability
Source Diversity ≠ Influence Diversity
Prediction Dependence ≠ Predictive Influence
Aggregation Weight ≠ Complete Influence Structure
Influence Assessment ≠ Final Collective Representation
Influence Pattern ≠ Legal Conclusion
Model Count ≠ Model Influence Distribution
Enterprise Application
PICM enables organizations to see the real power structure behind a collective prediction.
A forecasting platform may display:
10,000 participants
while a small number of sources materially determine the result.
An AI ensemble may contain:
50 agents
while most effective influence originates from one base model.
A prediction market may show broad trading activity while a few capital positions dominate price formation.
An expert consensus may contain dozens of contributors while most forecasts inherit one upstream institutional forecast.
PICM makes these structures governable before the collective result is represented as broad predictive consensus.
Use Case 1 — 25,000 Traders, but Four Accounts Drive the Market Signal
A prediction market displays:
82%
for the probability of a regulatory approval occurring before the end of the year.
The market contains:
25,000 participant accounts.
At first sight, the 82% signal appears to represent broad collective belief.
PICM examines the influence structure.
Four large accounts represent a substantial share of active capital, frequently move the market price, and account for most of the marginal changes in the displayed probability.
Thousands of smaller participants contribute activity but have limited effect on the resulting market signal.
The critical governance question is:
Does participation by 25,000 accounts justify representing the 82% market signal as broadly distributed collective judgment when a small number of positions materially determine the price?
PICM establishes a:
Highly Concentrated
influence state and requires that the concentration remain visible to downstream collective representation.
Number of Traders ≠ Distribution of Market Influence
Broad Participation Does Not Establish Distributed Predictive Influence
Use Case 2 — Fifty AI Agents, One Upstream Predictive Origin
An organization deploys:
50 autonomous forecasting agents
to predict whether a major industrial project will reach commercial operation within two years.
The agents use different roles, prompts, and task descriptions.
Their outputs are aggregated and presented as:
consensus across 50 AI agents.
PICM traces the influence structure.
Most agents rely upon:
the same base model; the same retrieval infrastructure; the same project dataset; one upstream risk forecast repeatedly referenced across the system.
The aggregate therefore contains fifty visible contributions but is materially shaped by one narrow predictive infrastructure.
The critical governance question is:
Does a fifty-agent architecture create distributed predictive influence when the majority of those agents inherit their predictive structure from the same upstream system?
PICM identifies:
Model-Family Concentration
Information-Origin Concentration
and:
Dominant Upstream Influence
The collective representation must preserve that structure.
Agent Count ≠ Distributed Agent Influence
Many Predictions ≠ Many Information Origins
Canonical Boundary
PICM governs the identification, attribution, qualification, structural interpretation, concentration assessment, and historical preservation of predictive influence within an aggregated predictive structure. It governs direct and indirect influence, source-family and dependence-cluster influence, model, agent, institutional, ownership, methodological and information-origin concentration, capital and market influence, Influence Chains, Influence Networks, Effective Influence, marginal influence, removal sensitivity, Predictor–Reality Proximity context, feedback and reflexive influence, concentration states, Hidden Concentration, and influence history. It does not determine Predictive Source Diversity, establish Prediction Dependence, select the aggregation methodology, silently reweight predictive contributions, determine legal misconduct, establish Operational Reality, determine final Collective Prediction legitimacy, govern Prediction Communication & Reliance, or assess predictive performance.
PICM governs:
who or what materially influences the aggregated predictive result and how concentrated that influence actually is.
It does not determine:
whether the resulting collective structure may ultimately be represented as a legitimate Governed Collective Prediction.
Canonical Closing Statement
A collective prediction may have thousands of visible contributors while only a few sources carry the weight of the result.
PICM exposes the sources, models, information origins, capital positions, clusters, and indirect pathways through which predictive influence actually flows.
Collective participation becomes meaningful only when governance can distinguish who contributed from who effectively determined the signal.