Section 01 — The Foundation
Ideas don't fail.
Reality does.
Every year, millions of ideas enter the world with full conviction and collapse on contact. Not because they were wrong. Because they met a society they didn't understand.
The wrong framing. The wrong audience. The wrong moment. A phrase that felt precise to the creator but felt threatening to the receiver. A message that resonated in one group and backfired in another.
Section 02 — The Problem
The gap between what you intend and what people hear is where everything dies.
You have an idea. You believe in it. But the moment it leaves your hands, it enters a system you cannot see: a network of human minds shaped by competing beliefs, fears, identities, and social pressures.
Traditional tools show you reach metrics after the fact. Surveys give you thin opinion slices before. Neither shows you what actually happens inside a social system when your idea hits it at scale.
◈ The Framing Problem
The same idea with different wording can produce completely different reactions in the same audience.
◈ The Sequence Problem
Who hears an idea first determines everything. Starting in the wrong group means it never escapes.
◈ The Identity Problem
Ideas that challenge core identity do not just fail. They make the opposition stronger and harder to reach.
◈ The Virality Problem
An idea can be deeply believed by everyone who hears it and still go nowhere if the wrong people are carrying it.
Section 03 — How It Works
Four steps from idea to insight.
You submit text. The engine does everything else. Here is what happens under the hood.
01
You write your idea
Anything between 3 and 500 characters. A product claim, a campaign headline, a policy statement, a tweet.
02
A synthetic population is built
20 to 300 AI agents are generated from 12 real-world archetypes. Each agent carries seven behavioral vectors: beliefs, identity, emotional triggers, social behavior, epistemology, and worldview alignment — optionally conditioned on a country profile.
03
The simulation runs
Agents react, interact, influence each other, and shift stances across multiple rounds or network ticks. Thirty-one behavioral mechanics — drawn from behavioral economics, social influence research, network science, and structural sociology — govern how beliefs move, and every agent hears its own neighbourhood rather than the whole crowd.
04
A verdict and patterns surface
You get a decision-first verdict, virality, polarization, and trust scores, plus opinion clusters, a full agent event feed, and network spread metrics.
Section 04 — Mission and Vision
Turn failure into a simulation, not a scar.
Every idea you care about carries risk. AnalyticalG steps in front of that risk before the campaign runs, before the message ships, before you bet on an outcome you cannot take back. Run the scenario. Read the room. Walk away knowing.
A future where no idea ships without a mirror.
We are building toward a world where creators no longer guess and strategists no longer hope. Where the distance between your intention and your audience collapses into something you can measure, study, and act on before it matters.
Before you publish, you know. Before you commit, you have already simulated.
Section 05 — The Two-Phase System
One simulation. Two dimensions of truth.
Society processes an idea in two distinct ways: internally (belief) and externally (spread). These are not the same thing. They do not predict each other. You need both.
Phase 01
Intelligence
What happens inside minds?
When your idea arrives, does it resonate or provoke? Does it align with existing beliefs or threaten identity? Does it change minds or harden opposition?
Virality
Polarization
Trust
Phase 02
Architecture
What happens between minds?
Does your idea have the structural conditions to travel? Who carries it, who blocks it, where does it tip from contained to widespread?
Spread Efficiency
Tipping Point
Resistance
The Four Outcomes — Why You Need Both Modules
High Intelligence, Low Architecture
Deep resonance. No reach.
Your idea converts every mind it touches but never escapes the first cluster. Seeded among believers and carried by low-influence agents, it dies inside the room it convinced.
A nuanced argument that converts specialists but never reaches the mainstream.
High Architecture, Low Intelligence
Everywhere. Believed by no one.
Your idea reaches 80% of the network but polarization is high and trust is low. It spread as noise, not signal. It touched everyone and changed nothing.
A viral headline that generates outrage but leaves zero lasting conviction behind.
High Intelligence, High Architecture
Spreads fast. Changes minds.
The most powerful outcome. Your idea moves through the full network and shifts belief at scale. This is the zone AnalyticalG helps you identify, understand, and use responsibly.
A message that reaches the mainstream and shifts how people think.
Low Intelligence, Low Architecture
The most honest result.
No traction. No spread. The simulation tells you before you invest that this idea has no structural fit with the society you are targeting. That is not failure. That is information.
Most ideas, as it turns out. Now you find out before you publish.
Phase 01
Intelligence
The Belief Layer — What happens inside minds?
Phase 1 deploys up to 300 synthetic agents, each carrying seven behavioral vectors: beliefs, identity, emotional triggers, social behavior, how they decide what counts as true, and which story about the world they already live inside. Your idea enters their world and every agent reacts in isolation first. No social influence yet. Just raw, unfiltered first contact across 300 different minds.
Then the social layer activates. High-influence voices reshape the undecided. Contrarians harden. Tribal clusters form around shared identity. The idea you launched stops belonging to you. It is now being processed, distorted, and amplified simultaneously across a society you can observe in motion.
The question Intelligence answers is not whether they like it. It is what happens to belief, trust, and polarization across the population after your idea has passed through it. Four simulation rounds. A complete map of how minds moved, and why.
Simulation Flow
Round 0 — First Contact
Every agent reacts to the raw idea in isolation. No social influence yet. Pure individual reception.
Round 1 — First Influence
Agents hear the loudest voices from Round 0. Belief starts to shift through social proof and contrarian response.
Rounds 2 to 3 — Emergence
Clusters form. Minority groups harden. Fatigue sets in for repeat speakers. Momentum builds or collapses.
Final Round — Equilibrium
The population arrives at a stable state: consensus, fracture, or polarised deadlock.
Round-by-Round Flow
The Emergence Mechanics — 31 Live, 8 Founding
These are not filters applied from outside. They are the natural patterns of how human beliefs move when exposed to other human beliefs. The eight founding mechanics below launched the engine; today thirty-one are live, spanning four research layers: behavioral economics (anchoring, identity entrenchment, hot-cold emotional states), social influence (norms, fluency, authority, reciprocity, mimicry), structural context (platform speed, institutional trust, vulnerability), and network science (complex contagion, neighbourhood-local exposure). Each one is grounded in published social science and covered by regression tests so it cannot silently stop working.
01
Opinion Inertia
People resist change based on how open they are. High-openness agents shift easily. Identity-anchored agents barely budge.
02
Social Proof
When enough people believe something, conformist agents follow. The bandwagon accelerates as adoption grows.
03
Contrarian Trigger
When majority agreement hits 65%, Rebel archetypes automatically flip. Consensus itself generates opposition.
04
Visibility Bias
High-influence, recent speakers generate more reaction. Older voices fade. Recency shapes what gets amplified.
05
Emotional Cascade
Fear and anger spread faster than measured responses. Emotional intensity is a direct multiplier on reach.
06
Speaker Fatigue
Agents who speak too often tire out and speak less. Attention is finite. Repetition loses its effect.
07
Minority Resilience
Small resistant groups facing a large majority dig in harder. Social pressure does not always break resistance.
08
Social Momentum
Fast-growing ideas get a further boost. Stalling ideas lose credibility with undecided agents over time.
What Intelligence gives you
Decision Verdict
A decision-first read of the run: what the population did with your idea and what to change before it ships.
Virality Score
How fast and how far did the idea spread through the population?
Polarization Score
0.0 means consensus. 1.0 means the population is split exactly in half.
Trust Score
How much credibility did the idea receive from the synthetic population?
Emotion Breakdown
Distribution of curiosity, skepticism, fear, anger, agreement, and indifference.
Agent Event Feed
Every individual reaction visible, showing the full conversation as it happened.
Ensemble Mode — Confidence Band
Instead of one run, ensemble mode executes five seeded simulations of the same idea and reports the spread across them as a confidence band. The band shows how much the outcome varies within the engine itself — population and trajectory variance, not external validation. A narrow band means the result is stable across plausible societies. A wide band is itself a finding: your idea's fate depends heavily on who hears it first.
Phase 02
Architecture
The Spread Layer — What happens between minds?
Phase 2 asks a different question entirely: can your idea travel? Where Intelligence models what people think, Architecture models what they do: who they tell, who stops listening, and whether the idea reaches enough nodes to sustain itself without the original push.
AnalyticalG builds a trust-weighted network of 100 to 300 agents. The idea seeds into the top 8% by influence score, then releases. Every subsequent agent independently decides: adopt, share, resist, or ignore. Resistors form structural firewalls. A single highly-connected resistor can stop propagation dead in its tracks.
The tipping point (the exact moment when spread becomes self-sustaining) is the critical metric. If it never arrives, you know before you invest. If it does, you know at which tick, through which pathway, and at what adoption rate it crossed the threshold.
Propagation Flow
Setup — Network Construction
A network of 100 to 300 agents is built. Connections form based on the affinity matrix: who trusts whom and how strongly.
Tick 0 — Seed Injection
The idea is seeded into the top 8% by influence score. The most connected, most trusted agents receive it first.
Ticks 1 to N — Propagation
Every agent with the idea decides: adopt, share forward, resist, or ignore. Resisters create network firewalls.
Tipping Point — Phase Transition
At a critical threshold, spread becomes self-sustaining. The simulation identifies the exact tick this occurs, or confirms it never will.
Network Propagation — Visual Model
Not an illustration — this graph is grown by the same Holme-Kim procedure the engine uses (preferential attachment with triad closure), and the colouring is the result of running an actual complex-contagion cascade across it, where an agent adopts only after three distinct neighbours already have. Seed influencers (bright blue) inject the idea. It radiates outward through adopters. Resistors (red) form blocking clusters. Passive agents never reach the threshold.
Network Theory Under the Hood
Ideas do not spread like viruses. Information can pass through a single contact, but belief is a complex contagion: an agent moves only when enough of its trusted contacts have already moved. This is not an analogy the engine gestures at — it is implemented. Contested claims require reinforcement from several distinct sources before belief shifts at full strength, and hearing the same loud voice twice does not count. Resistance is never subject to that requirement: refusing to adopt needs no corroboration.
Both simulation layers run on a graph with realistic clustering, so agents hear their own neighbourhood rather than the whole crowd, and the triangles that carry reinforcement actually exist. The lineage is Granovetter's threshold models and weak ties (1973, 1978), Centola & Macy on complex contagion (2007), and Holme & Kim on clustered scale-free networks (2002). It is why network shape, not message quality alone, decides whether your idea travels.
Diagram 01 — Simple vs. Complex Contagion
Raw information hops along single ties. Belief needs multiple confirming voices before an agent crosses its adoption threshold.
Diagram 02 — Bridge Width Between Communities
Weak ties are enough to carry news between clusters, but a belief must arrive from several directions at once. Narrow bridges leak information and block conviction — which is why the same idea can saturate one community and never escape it.
Diagram 03 — The Threshold Cascade
Two fates of the same idea. Cross critical mass and adoption becomes self-sustaining; stall below it and the network's own inertia rolls it back. Architecture reports the exact tick the crossing happens — or that it never does.
The 12 Archetypes
The synthetic population is built from 12 archetypes drawn from behavioral psychology. Each carries a unique belief vector, emotional profile, and social role. Together, they represent the full ideological range of a real society.
Aspirational Striver
Status-driven. Amplifies aspirational ideas.
Skeptical Realist
Evidence-first. Slows emotional contagion.
Trend Follower
Conformist. Primary viral amplification layer.
Authority Believer
Institutional trust. Anchor against disruption.
Rebel Thinker
Contrarian. Auto-inverts on majority consensus.
Nihilist Observer
Detached. Resists emotional manipulation.
Corporate Professional
Risk-managed. Bridges clusters pragmatically.
Creator Hustler
High influence. Fastest transmission node.
Risk-Averse Planner
Fear-anchored. Blocks aggressive ideas.
Social Justice Advocate
Moral focus. Amplifies systemic narratives.
Tech Optimist
Future-positive. Receptive to disruption.
Traditionalist
Past-oriented. Maximum resistance to novelty.
What Architecture gives you
Adoption Curve
Tick-by-tick graph of what fraction of the network adopted the idea over time.
Spread Efficiency
How effectively did the idea use the network's available connections? Scored 0 to 1.
Tipping Point
The exact tick where adoption became self-sustaining, or confirmation that it never did.
Resistance Clusters
Which archetypes formed active firewalls and how strong was their blocking signal?
Key Transmission Nodes
The top agents responsible for carrying the idea across the network.
Section 08 — Grounding
The engine reads the room before it simulates.
A simulation of an imaginary society is a thought experiment. A simulation seeded with what the real world is saying right now is an instrument. Before a run, the ingestion pipeline checks what is actually circulating about your topic and feeds that read into the engine's topic analysis.
◈ LIVE
Global News Coverage
Live worldwide news monitoring surfaces the headlines, themes, entities, and tone circulating around your topic — deduplicated down from raw syndication noise to distinct stories. This is the "what’s circulating" panel, and it flows into the topic read the agents receive.
◈ WIRED
Prediction Markets
Where a matched market exists on Polymarket or Kalshi, the pipeline pulls the market-implied probability — real people pricing real money on the outcome — as a grounding input. Market price histories also serve as an external calibration reference the engine’s trajectories are backtested against.
◈ BUILT
Social Discourse
Connectors for Reddit (including real reply graphs), Bluesky, and YouTube are built into the same pipeline to add opinionated, conversational text alongside news — the layer where people take positions rather than report them. Activation is credential-gated and rolling out.
Ungrounded runs say so
Grounding degrades gracefully and honestly. If no real-world signal is available for your topic — no coverage, no matched market, a source outage — the run is explicitly labeled as ungrounded rather than quietly falling back. You always know whether you are looking at a simulation informed by the world or a simulation of the engine's priors alone.
Calibration, not validation
Backtests against market price trajectories and 12 historical information events are calibration references — external signals we did not hand-pick, used to tune the engine. They are not independent validation, and we do not present them as such. Until an independent validation number exists, the honest frame for this platform is stress-testing and failure-mode discovery, not quantitative prediction. No personal data is stored at any point: sources are processed in aggregate only.
Section 09 — The Loop
The two phases do not just coexist. They create a feedback loop.
Intelligence tells you if the idea lands. Architecture tells you if it travels. Each answer raises a new question that only the other module can answer.
The Iteration Loop
Intelligence
Run your idea through the belief layer.
What you learn
Rebel Thinkers and Traditionalists are hardening. Trust score: 0.41. Polarization: 0.78.
New question
Is it the framing or the idea itself causing resistance?
Refinement
Adjust the framing based on what Intelligence revealed.
What you learn
Remove identity-threatening language. Soften the challenge to authority.
New question
Does the refined version now have the conditions to spread?
Architecture
Run the refined idea through the network spread layer.
What you learn
Spread efficiency: 0.71. Tipping point at tick 6. But Authority Believers are forming a resistance cluster.
New question
What in the idea triggers institutional rejection?
Intelligence
Return to Intelligence, focused on the Authority Believer archetype.
What you learn
The framing implies institutional failure. Authority Believers protect institutions reflexively.
New question
Can we reframe without losing the core claim?
The Loop Continues
Each run teaches you something specific about your idea's relationship with society.
What you learn
The loop is not a flaw to be resolved. It is how learning works.
New question
You stop when you understand, not when the scores are perfect.
Why the loop works
Each module answers what the other cannot
Intelligence tells you what minds do with your idea. Architecture tells you whether that idea ever reaches those minds at all.
Every re-run is a hypothesis test
Change one variable and run again. The delta in scores tells you exactly which lever moved the needle and why.
Randomness keeps the results honest
The same input never produces identical output. Each run is a plausible variation of reality, not a fixed prediction — and ensemble mode turns that spread into a confidence band you can see.
Section 10 — Experimental Platform Notice
Experimental Project
AnalyticalG is an experimental simulation platform built by Truble Marketeer. It has not been validated against real-world behavioural data at scale. All outputs are synthetic, probabilistic, and exploratory.
No liability for the misuse of data generated by this platform rests on AnalyticalG or Truble Marketeer. Users assume full responsibility for how they interpret and act on simulation results.
Scores are not predictions, guarantees, or professional advice of any kind. Do not use simulation outputs as the sole basis for consequential decisions.
Section 11 — The Promise
Test ideas before
they meet reality.
AnalyticalG does not make ideas good. It makes the relationship between an idea and its audience visible, before the cost of being wrong has been paid.
Phase 01
Intelligence
31 behavioral mechanics · Grounded topic reads · Network-local belief propagation
Phase 02
Architecture
Network graph · Tick propagation · Tipping point detection
The Loop
Iterate
Refine. Re-run. Learn. Until you understand your idea's relationship with society.