Simulations
Can you do better than Kodak executives?
MIFTAN LABS AI narrative simulations place you inside historical crises with only the tools and information available at the time. You make the decisions. MIFTAN LABS system adjudicates the outcomes. Find out whether hindsight is really 20/20 — or whether the structure of the situation made failure almost inevitable.
Knowing about a bias doesn't protect you from it. You will optimize for the wrong variable. You will discount the slow-moving threat. You will protect the identity that's killing you. Not because you're unaware — because that's what the structure rewards.
Each one of MIFTAN LABS simulations compresses consequences. In the real world, structural failure takes years to surface. Here, it takes hours.
KODAK
The Innovation Trap
Board-level and C-Suite. Full day.
It's 1981. Kodak is booming. Executives are sitting on top of the most profitable company in photography. Engineers are sounding the alarm: digital is coming, film won't last forever, and they have the technology to lead the transition.
This simulation pits executives against engineers. Both sides have legitimate arguments. Neither side can win alone. Can you find the path they missed?
The conventional lesson is that Kodak should have embraced disruption. The simulation reveals why that wouldn't have saved them — and why "pivot to AI" might be the same trap in a new disguise.
MOB
Trust Under Pressure
Leadership team simulation. Half day.
Chicago, 1924. The FBI is circling. You're running the family. The booze is flowing and the neighborhood knows who's really in charge. Money, power, and loyalty are in the balance.
This simulation pits the Moretti Family against the FBI. Both sides have resources. Both sides can win. The difference is team dynamics.
The FBI needs just one person to flip. The Mob needs just one cop to take a bribe. The neighborhood has eyes, and it's impossible to know who to trust. Can you create the team dynamics needed to keep control of the neighborhood?
.jpeg)
GENGHIS KHAN
Scaling Against Incumbents
Founders and Marketers. Half day.
The year is 1200. You command a coalition outnumbered, outresourced, and underestimated. The established empires think you're a nuisance — they don't take you seriously yet. But you're playing for domination. Speed is your only advantage.
The simulation pits the Khan against his own generals. Both sides need each other. Both sides can destroy each other. The difference is how you structure power.
Micromanage and you bottleneck the empire. Delegate too much and you create rivals. Can you hold onto power and grow at the same time?
ROMAN EMPIRE
Shared Power, divided vision
Co-founders and leadership partnerships. Full day.
Constantinople, 532 AD. The empire is unstable. Riots are burning the capital. The Emperor wants to flee. The Empress refuses: "Purple makes a fine shroud."
The simulation pits Justinian against Theodora. Both hold power. Both need each other. Both see the threats differently. The difference is whether they can align before the empire burns.
Generals and courtiers are watching. When the leaders disagree, the people beneath them don't know who to follow.
Can you manage an empire if you can't come to a decision?
THE FRAMEWORK
WAR GAMES BASED ON MATH,
NOT VIBES
Most strategic simulations are storytelling exercises. A facilitator sets the scene, players make choices, outcomes get interpreted based on judgment. It feels meaningful. But there's no way to know if the result would actually happen.
MIFTAN LABS is different.
The engine is Hybrid Neuro-Symbolic AI — a dual-stack system where hard-coded mathematical logic does the calculation and AI does the storytelling.
The logic core runs on game theory — Nash equilibria, coordination dynamics, mechanism design. Every decision has defined effects. Parameters interact through specified relationships. A proprietary stress metric tracks pressure across multiple vectors — financial, operational, political, internal. Outcomes aren't interpreted. They're calculated. Auditable. Repeatable. The math runs on every move. Players see the scoreboard.
MIFTAN LABS AI layer generates the narrative — the boardroom tension, the market news, the pressure from above. No two sessions unfold the same way. But the AI doesn't decide outcomes. It tells the story the math produces.
The stochastic layer adds chaos the way reality does. Nothing is perfectly predictable. Good decisions can fail. Bad decisions can get lucky. But over time, some strategies win and others predictably lose. Just like real life.
The facilitator doesn't decide the outcome. You're not just playing against the opponent. You're playing against the math.
THE GAME THEORY
MIFTAN LABS framework was developed by a game theorist trained under Prof. Robert Aumann at Hebrew University — the Nobel laureate whose work on correlated equilibrium and common knowledge explains why rational actors, with perfect information, still produce irrational collective outcomes.
MIFTAN LABS simulations are built on mechanism design, correlated equilibrium, and repeated game dynamics — the formal structures that explain why rational actors, making optimal individual choices, produce collectively catastrophic outcomes.
Nash equilibrium traps. Cascade matrices with directional coefficients. Phase transitions when parameters cross thresholds. O-SSR tracking organizational stress across entropy, capacity, flow, and distribution.
Monte Carlo validation. Strategy comparison showing why aggressive approaches fail — concentrated pressure doesn't trigger cross-parameter cascades as effectively as adaptive pressure that pushes multiple thresholds simultaneously.
The identity trap formalized: when survival-optimal choices and identity-consistent choices stop overlapping, the organization cannot adapt without destroying itself. The math has been tested. The strategies have been ranked. The cascades have been mapped.
That is what players are up against.
BEHAVIORAL ECONOMICS
Game theory tells you what rational actors should do. Behavioral economics tells you what they actually do — and why the gap between those two things can be catastrophic.
Loss aversion. Commitment escalation. Sunk cost entrapment. The planning fallacy. Confirmation bias under stress. These aren't quirks. They're predictable patterns that appear in every organization, in every crisis, across every industry. Kahneman and Tversky. Thaler. Ariely. Four decades of research proving that smart people make the same mistakes — not because they're stupid, but because they're human.
MIFTAN LABS simulations are tuned to exploit these patterns. You'll feel loss aversion pulling you toward protecting what you have instead of risking what you need. You'll escalate commitment to failing strategies because you've already invested too much to quit. You'll filter information to confirm what you already believe — especially under pressure. Not because you don't know better, but because that's what pressure does.
These aren't flaws in the game; they're human defaults. Real crises trigger people to make decisions that can blow up in their faces. The simulations are designed to do the same.
The math models the structure. The behavioral economics models you. And knowing about a bias doesn't protect you from it. That's the point. You have to feel the pattern to see the pattern.
What Makes This Different
Most strategic simulations are complex theater. A facilitator sets the scene. Players make choices. The facilitator interprets what happens based on judgment, experience, and whatever outcome feels logical to them. For participants, it feels meaningful. And it might be. There's no way to know if it's true.
MIFTAN doesn't work that way.
The engine runs whether you like the result or not. Every move has defined effects. Parameters interact through specified relationships. When you make a decision, the math calculates what happens — not the facilitator. You can argue with a consultant. You can't argue with a model.
The AI tells the story. It plays your opponents. It narrates the boardroom tension, the analyst calls, the talent defections. But it doesn't decide outcomes. It translates the math into consequences you can feel.
That's the point. You don't experience a spreadsheet. You experience a company falling apart — or pulling together — and only afterward do you see why. The simulation doesn't care about your title, your confidence, or your narrative. It measures what you actually did. It shows you what it cost.
The mechanics are proprietary. What you take away is the insight.
About Us
Sarah Nadav
Behavioral economist. Serial entrepreneur. Game theorist.
Sarah studied game theory at the Hebrew University of Jerusalem under Professor Robert Aumann in the early 2000s — the Nobel laureate whose work on repeated games and common knowledge explains why rational actors, pursuing rational strategies, produce collectively irrational outcomes. Arms races. Market failures. Organizational collapse. The mathematics are the same.
Instead of building a career in academia or NGOs, she decided to become a serial fintech founder (VC-backed), with experience across every company stage — from early team at Vayyar (now a unicorn) to Senior Behavioral Designer at Wix Payments (NASDAQ: WIX). The pattern: applying behavioral economics to systems where decisions compound — finance, healthcare, organizational design.
Member of the World Economic Forum Expert Network since 2017, contributing to global initiatives on economic policy and systemic risk. Featured in Forbes, Bloomberg, TIME. Author of a bestselling book on financial decision-making during crisis, quoted in numerous books and textbooks.
Sarah contracted COVID in 2022. What followed was a cascade of life-threatening complications that were repeatedly misdiagnosed. No single decision was irrational. Each institution followed the procedure. Taken together, those procedures produced a system that could not coordinate, could not adapt, and could not see the patient in front of it. Every attempt to survive triggered failures elsewhere in the system, with catastrophic personal consequences. She lived the trap.
That's when the theory stopped being academic. The same dynamics that killed Kodak — loss aversion, information asymmetry, identity-consistent decisions overriding survival-optimal ones — were killing her. She watched rational actors, following rational procedures, produce outcomes that made no sense. The math was the same. She'd just never been the one inside it.
MIFTAN came out of that. Not despite the collapse — because of it.
The simulations are built on all of it — the game theory, the behavioral research, the experience of watching organizations make predictable mistakes under pressure. The math is rigorous. The scenarios are real. The traps are the ones she's seen kill companies. Players get to beat the odds, outperform history, and take away the lessons without paying the price.
Working With MIFTAN
Sarah designed the frameworks and analytical models. Delivery is handled by a trained facilitation team operating to the same standards.
Each engagement: pre-brief, live simulation, structured debrief, executive summary. Custom simulations for boards and executive teams are scoped individually.
Client discretion and analytical integrity are foundational.
Advanced applications for defense, government, and classified contexts are handled separately and are not publicly described.
The Engine
MIFTAN is powered by Hybrid Neuro-Symbolic AI—a dual-stack system where mathematical logic does the calculation and AI does the storytelling.
The Logic Core is a Python-based simulation engine built on a proprietary metric: the Organizational System Stress Ratio (O-SSR). It tracks how coordination capacity, resource flow, identity constraints, and narrative coherence interact under pressure. Parameters are interdependent. Cascades propagate through specified relationships. When parameters cross thresholds, phase transitions trigger—the system doesn't degrade linearly, it breaks.
The stochastic layer adds fog of war. Good decisions can fail. Bad decisions can get lucky. But over time, the math converges. Monte Carlo validation across thousands of runs shows which strategies win and which predictably lose.
The Interface Layer is a constrained LLM that functions as referee and narrator. It translates validated mathematical outputs into scenario briefings and game narrative. It does not decide outcomes. It does not modify the core model. It tells the story the math produces.
The system is designed to identify what conventional metrics miss: the pre-collapse phase, where organizations remain operationally functional but are structurally unstable. The period before failure becomes irreversible.
The mechanics are proprietary. What you take away is the insight.