What this measures
Each pump adds points. Bank them and they are yours. Pump once too often and the balloon bursts, taking everything on it. The burst point is hidden and different every balloon.
This is the Balloon Analogue Risk Task (BART), published by Lejuez and colleagues in 2002 and one of the most used behavioural measures of risk-taking. Its appeal is that it captures what you actually do under uncertainty rather than what you say about yourself on a questionnaire.
The maths, since the game will not tell you
Burst points here are drawn from a small range, so the expected value of continuing falls with every pump: early pumps are nearly free, and past a certain point each one risks more than it adds.
The optimal strategy is therefore a fixed, moderate number of pumps, banked every time — not "pump until it feels risky". That intuition is exactly what the task is designed to expose, because feelings do not track expected value well.
- Too few pumps — you leave a lot on the table across eight balloons.
- Too many — bursts wipe out the accumulated gains faster than the extra pumps add.
- Inconsistent — usually the worst outcome, and the most common.
The pattern worth noticing about yourself
The interesting data is not your score, it is how your behaviour changes after events.
- After a burst, do you get cautious? Most people do, sharply — even though the next balloon is statistically independent. That is the gambler's fallacy running in the other direction.
- After a big bank, do you get bold? Also common, also unjustified.
- Do you chase? Pumping harder to recover a loss is the single most costly pattern here, and it is the same pattern that shows up in real financial decisions.
A stable, unchanging strategy across all eight balloons is both the optimal play and the rarest one.
Why hiring assessments use it
BART-style tasks appear in assessment batteries because they are hard to fake. You can tell a questionnaire you are "measured but decisive"; you cannot easily perform that across eight balloons under time pressure without it showing.
What is being read is usually not "risk-taking is good or bad" — it is consistency. Wild swings between extreme caution and extreme risk read poorly almost everywhere, while a steady policy reads well regardless of where it sits.
Frequently asked
Q. Is there a best number of pumps?
A. Given the range used here, a mid-range number banked consistently beats both extremes
over eight balloons. Working out roughly where that sits is the puzzle.
Q. Does BART predict real-world risk-taking?
A. Modestly. It correlates with self-reported risk behaviours in several studies, but the
relationships are small and it is a research instrument, not a diagnostic.
Q. Is the burst point really random?
A. Yes, drawn fresh for every balloon. There is no pattern to learn — which is precisely
what makes the after-a-burst caution irrational.
For a different decision-under-uncertainty task, try the card decks, which measures learning from feedback rather than risk appetite.