Run Your Life as an Experiment
- Difficulty
- Easy
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- —
Sparks' antidote to over-planning and existential second-guessing, and the organising idea behind his free workbook Experiment Without Limits. Instead of deciding whether something is right for you, you frame it as a hypothesis — if I do X, Y will happen — find the fastest way to test it, and commit to a fixed window during which the question is closed. Signs of progress are defined in advance so you know what a working experiment looks like, and you keep pulling levers: more of what seems to work, less of what does not. The commitment is what creates the power, because it stops you revisiting the decision every day and frees the mental bandwidth that comparison against unlimited optionality consumes. At the end you review honestly and choose the next experiment.
Origin
Extracted from Deep Dive with Ali Abdaal
How to run it
- 1
Write the hypothesis
State a guess in the form 'if I do X, Y will happen' — if I start putting out YouTube videos I will find my tribe; if I live with a flatmate I will be happier than living alone. Naming the guess makes it testable rather than agonising.
- 2
Find the fastest possible test
Deconstruct the hypothesis to the smallest action that generates real feedback. Sparks' bias is towards action over intellectualising; instead of five-year plans, rent the van for a week and see what living in it is actually like.
Pro tip Raise the stakes to tighten the loop: Sparks got funnier by performing stand-up rather than watching comedy specials.
- 3
Commit to a fixed duration
Give the experiment enough time to work — thirty days, three months, a year, or six months minimum for something as slow as poker. During that window the question of whether to do it is off the table.
Watch out Cutting the window short guarantees noise. Sparks warns that in poker you should expect six to twelve months before even a semblance of profit.
- 4
Define the signs of progress in advance
Decide what you will see along the way if the experiment is going well — which inputs are leading to good outputs — and how you will double down or accelerate if you see them.
- 5
Collect data and pull levers while it runs
Treat the whole period as continuous data collection. Notice which days feel fulfilling, which ones don't, what happens when you go outside or work out harder, and adjust the dial rather than the direction: a bit more of this, a bit less of that.
Pro tip The framing removes pressure. It stops being 'is this my life' and becomes 'let's see what this button does'.
- 6
Review only at the end, then choose the next experiment
At the close of the window take the designated step back: what did I love, what did I not, what went well, what didn't. Then decide to continue, adjust, or do something else entirely — and start the next experiment.
Watch out Asking 'is this what I want to do with my life' every single day is the failure mode this framework exists to remove.
In the wild
Ali framed moving into an expensive London studio and building an in-person team as a one-year experiment rather than a permanent decision. Going all in for the fixed window meant not relitigating the choice, and the year produced a large amount of learning. At the end of it the next experiment presented itself — given that he is staying in London, what does a studio inside his own house look like, and what if he lived with his brother and sister-in-law.
→ A year of committed learning followed by a clean, low-drama decision about the next configuration.
Ali had long liked the idea of living in a van, but had filed it under a future life — five years out, in the US, with a visa. Re-listening to a Noah Kagan interview where Kagan simply rented a van for three weeks to try it, Ali realised he could rent a van in London and live in it for a week purely as an experiment, and added it to his list of experiments to run.
→ A five-year fantasy converted into a one-week test that can be run immediately.
Common mistakes
Reopening the decision every day
Constantly asking whether this is really what you want burns bandwidth and prevents commitment. Sparks confines that question to a designated time before and after the experiment.
Trying to optimise before you have any feedback
You cannot optimise up front. Sparks' point about YouTube applies generally — the reps are their own reward, and only publishing tells you what people respond to.
Chasing unlimited optionality
Bouncing between options means never giving any of them a fair run. Committing for a defined period is what lets the experiment actually produce a signal.
From the transcript
“and my idea of framing as an experiment is well what's an hypothesis i have what's a guess that hey if i do x y…”
“if i could kind of train one mindset into people is to approach things experimentally to try it see what happens if you like the…”
“there's a real power in this commitment because a challenge or opportunity that we have is that life has unlimited optionality”
From the episode
World Poker Champion On The Science Of Decision Making - Chris Sparks
Chris Sparks