What a Simulation Shows About a Decision I Made Without One

In 2022 I exited a data privacy startup I had co-founded. I had a material stake in it after two years. There was friction at the end. I left.

That decision was made on gut, financial pain, and the sense that I was the wrong person in the wrong seat at the wrong time. Not on a model. I did not have these tools then.

Four years later, I ran the decision through a structured thinking protocol and a probability model.

The model validated the exit. But that is not the finding.

The finding is this: the gut had already computed the answer. The model gave me language for why.

And the language was not what I expected. I thought I left because of execution problems — friction, misalignment, wrong-seat energy. The model says otherwise. What it says is below, and it changed what lesson I take from the exit.

The structural insight

Before the numbers: what kind of decision was this?

A stay-or-exit founder call is not one decision. It is a chain of conditions that must all fire for the stay scenario to produce a positive outcome. In this case:

1. The regulatory environment (India’s DPDPA) must actually enforce — not just pass, but become meaningful
2. The company must survive long enough to see that demand arrive
3. The company must then execute well enough to capture a share of it
4. The payout must be large enough to justify the years of compounding cost

Each condition multiplies the others. The early gates dominate. If enforcement does not arrive, execution quality is irrelevant. If the company does not survive, execution quality is irrelevant. By the time execution quality matters, it is already the last gate in a chain.

The friction felt like an execution problem. It was actually a signal about a timing and survival bet — a bet I was not willing to hold for an unknown number of years against conditions I did not control.

Before the model: The Blade

Before running any numbers, I used an AI-assisted critical thinking workflow built around The Blade protocol. Six questions designed to surface what you’re not seeing before you start computing.

The most important output was a framing constraint: this simulation is post-hoc. The decision was made on gut and friction, not analysis. Publishing it as validation of a calculated choice would be dishonest — and that dishonesty would undermine everything the field note claims to show.

The Blade’s cut:

• One thing to do: Frame this explicitly as retrospective. These tools did not exist when I decided.
• One thing to stop: No characterization of the people who continued building. The analysis is about the structure of the decision, not the people inside it.

The model

Two scenarios. 10,000 runs each. Every input is a probability distribution, not a fixed value.

Scenario A — Exited 2022: Income growth rate, year-over-year growth consistency, capital that stopped burning after exit. These inputs are partly informed by what actually happened — which means the comparison is directional, not a controlled experiment.

Scenario B — Stayed: Annual cash burn, probability DPDPA enforcement becomes meaningful, probability the company survives long enough to see demand, probability of executing well enough to capture it, size of a blockbuster outcome if all three conditions fire simultaneously.

Scenario B’s model did not settle to a stable mean. That non-convergence is a finding, not a flaw: it is the statistical signature of a two-mode outcome. Almost always bad, rarely transformative, nothing in between. The 1.6% blockbuster tail keeps pulling the average — which is why the median tells you more than the mean here.

What drives each outcome

The model attributes how much of the variance in each scenario is explained by each input. This is where the last-gate structure becomes visible.

In Scenario A, growth trajectory and capital saved together explain nearly all the variance. Starting income explains almost nothing — it gets compounded into irrelevance within two years.

In Scenario B:

Those who stayed are not running an execution problem. They are running a timing and survival problem.

Improving execution from poor to strong barely moves the distribution — because the other conditions gate it first. The company’s fate in Scenario B is more clock-dependent than skill-dependent.

This is what the gut was computing in 2022, without language for it.

What this is and what it is not

The model validated the exit. The gut had already decided that. These are not independent findings.

What the model added: a structural frame for why the gut was probably right. Not because execution was weak or the team was wrong, but because the bet required patience through conditions I did not control, in a regulatory environment with uncertain timing, with capital burning throughout.

The 1.6% blockbuster scenario is real. Those who stayed are running that bet. The model shows exactly what lever matters most if you are running it: survival probability, not execution quality.

AI did not tell me whether leaving was morally right, relationally clean, or personally justified. It did not resolve the human texture of the exit. What it did was make it possible to interrogate the structure of the decision with more rigour than I had access to at the time.

AI can make retrospective analysis cheaper and the structure of a bet more visible. It cannot frame the question correctly, choose sensible inputs, or tell you what the model cannot know. Those remain human judgments — and the first one is always: what kind of decision is this, actually?

Tools: AI-assisted /critical-thinking — The Blade (State / Failure / Unseen / Actors / Constraint / Cut). AI-assisted /monte-carlo — 10,000 runs per scenario, Latin Hypercube Sampling, Sobol first-order sensitivity indices.

All figures are multiples of the original stake. No absolute financial figures.