FoMind
Meet Lin: a computer science graduate who loves new tech and spends weekends coding at home. Which is more likely?
- A: Lin works at an internet company.
- B: Lin works at an internet company and is building his own app on the side.
Many people pick B because it fits Lin better. But B is just a slice of A: everyone who fits B also fits A, so B can never be more likely than A. Mistaking “more detailed and more believable” for “more probable” is the conjunction fallacy.
Every added condition lowers the odds (or leaves them unchanged)
When an outcome needs several things to go right at once, the overall probability is the product of each step. Say you’re planning a hike with friends next Saturday:
- Good weather: 80%
- Everyone is free: 70%
- You get train tickets: 90%
If these are roughly independent, the chance that all three work out is 0.8 × 0.7 × 0.9 ≈ 50%. Each one looks promising on its own, yet together it’s a coin flip.
A starker example: a project has 5 stages, and you’re 90% confident in each. The chance that all go smoothly is 0.9⁵ ≈ 59%. The more stages there are, the wider the gap between “every step looks fine” and “the whole thing works.”
If the steps affect each other, replace each one with “the chance this step succeeds, given the previous ones did” and multiply those — the logic is the same.
Try the multiplier below: add or remove steps, adjust each probability, and watch how fast the total drops.
This is where the interactive demo “Multi-step probability multiplier” lives: in the FoMind app you can drag the controls and see results update live.
Using it in forecasts
In the “Key uncertainties” field, list each condition that has to hold, estimate each one, multiply, and compare the result with your first gut feeling. If your gut was much higher, you were probably won over by a vivid, complete story.
The more specific a story, the more convincing it feels — but every extra condition makes it less likely, not more.