It's Just an Experiment  ·  tiny experiments & growth loops
An interactive note on tiny experiments

It's Just an Experiment

Three games that show why a drifting world rewards loops over ladders, why a small, repeatable promise is almost impossible to fail, and why many tiny reversible bets quietly beat one big plan — after Anne-Laure Le Cunff's Tiny Experiments.

↓   play the first game
Game One · Ladders & loops

The goal you set is for a world that won't stay still

Pick a five-year goal and march straight at it — that's a ladder. But what you'll want, and where the reward actually sits, keeps drifting while you climb. A loop takes one small step, looks at where things moved, and re-aims. Same effort, two very different places to land.

Below, the faint line is where the reward actually is — it wanders, because the world and you both change. The ladder commits at the start. The loop re-aims every step. Press play, then shift the world and watch which one copes.

Ladder · gap at the end
Loop · gap at the end
Loop · re-aims
0
where the reward actually is the ladder (fixed at the start) the loop (re-aims each step)
Time runs left to right. The ladder is a straight bet on a prediction; the loop is a short conversation with reality, repeated.

The ladder isn't lazy or foolish — it's committed. That's the problem. Commitment to a fixed target is only as good as the prediction underneath it, and we are famously bad at predicting what we'll want. The loop never predicts very far. It just keeps a tight, honest contact with the present and lets the destination emerge.

“We don't go in circles; we grow in circles.”

Anne-Laure Le Cunff · Tiny Experiments

The loop has a shape: Pact → Act → React, then you decide — persist, pause, or pivot — and go round again. Each turn is cheap, so you can afford a lot of them. Which raises the obvious worry: doesn't running this many little trials mean failing constantly? That's Game Two.

Game Two · The PACT

How to make a promise that's almost impossible to fail

A SMART goal grades you on the outcome: lose 10 kg, hit 10k subscribers, ship the launch. But the outcome leans on a hundred things you don't control. A PACT grades you on the output — the action you actually take — which is entirely yours. Le Cunff's four words: Purposeful, Actionable, Continuous, Trackable. Written as one sentence: “I will [action] for [duration].”

I will write 100 words every morning for two weeks.
Chance you can call it a success
Of the result, the share you control
Success vs. how much rides on the world. Grade on the outcome (red) and luck decides; grade on the output (green) and you do.

So the trick isn't grit. It's moving the finish line onto ground you own. A small, repeatable, binary action — did I show up, yes or no — is something you can almost always honour, whatever the world does. Failure all but disappears, and what's left is data.

Which is why the resistance you feel before starting isn't a verdict — it's a signal. Le Cunff borrows a three-part check: Head (is this the right thing?), Heart (does it pull me?), Hand (can I actually do it?). When all three light up, she calls it aligned aliveness — and the procrastination melts. Tap the ones that are true:

Head
Is this the right thing to do?
Heart
Does it genuinely pull me?
Hand
Can I actually do it now?
Aligned aliveness — go
Procrastination isn't a flaw to crush; it's a reading on which of the three is missing.

“It's almost impossible to fail when you see everything as an experiment.”

Anne-Laure Le Cunff · Tiny Experiments
Game Three · Optionality

Why many small bets beat one big one

Here's the punchline the first two games were built for — and it's the same one the sibling essay makes about markets. A tiny experiment has a capped downside (it's small and reversible — you can stop) and an open upside (it might just work, or open a door you couldn't have planned). Run a lot of them and the math turns convex: your worst case stays flat while your best cases pile up.

Worst case (the most you can lose)
Typical year (median outcome)
A lucky year (top 5%)
1,000 simulated years median worst case
Each dot is one possible year. The big plan is a single coin flip with your whole stack; the many-experiments year floors the loss and keeps the tail.

Notice the slider's left edge. Run zero experiments and the downside is perfectly safe — and nothing ever happens. Le Cunff's warning about hoarding options: it's “like spending your whole life filling up the gas tank without ever driving.” Optionality only pays if you spend it. The convexity isn't in having the experiments. It's in running them.

“When we fixate on finding one singular purpose, we rule out the side quests that help us grow the most.”

Anne-Laure Le Cunff · Tiny Experiments
In one line

Stop predicting the path. Run it.

The world drifts, so loops beat ladders. The finish line you own can't make you fail. And a pocketful of small, reversible bets compounds into the one big thing you could never have planned for. Three games, one conclusion — and, as the neuroscientist says, it's almost impossible to fail when you see everything as an experiment.

“Replace the old linear model of success with a circular model of growth, in which goals are discovered, pursued, and adapted — not in a vacuum, but in conversation with the larger world.”

Anne-Laure Le Cunff · Tiny Experiments

The ideas here are Anne-Laure Le Cunff's, from Tiny Experiments: How to Live Freely in a Goal-Obsessed World (2025) and her writing at Ness Labs — the PACT (“make a PACT instead”), the growth loop, the Triple Check (after Hugo Kehr), and the bridge to optionality she draws in “The optionality fallacy,” where she meets Taleb's antifragility head-on. The interactive games are my own toy models — deliberately simple, directionally honest. Its sibling, It's Just Math, makes the same convexity argument for money.

Get the next simulator

One email when a new concept goes playable — plus the model behind it. No noise.

No spam, unsubscribe any time.