Pymetrics Balloon Game

The pymetrics balloon game asks you to pump a balloon for $0.05 a pump and click Collect before it pops; a popped balloon loses its money. Prep sites that have rebuilt it describe 39 balloons in three colours, each colour with its own popping pattern, and no time limit (JobTestPrep, Game Assessment Prep). It is one of the trait games in the Harver (formerly pymetrics) battery: it reads how much risk you take and whether you learn from the balloons that pop. pymetrics' own support centre says the games have "no right or wrong answers" (pymetrics Support Center), so the aim is not a high score but an honest, steady run. You can try the same pump-and-collect mechanic free on our balloon game.

What is the pymetrics balloon game?

It is a version of a lab task psychologists have used for over twenty years, the Balloon Analogue Risk Task or BART (Lejuez et al., 2002). In the lab and in the pymetrics game the structure is the same:

• Each pump inflates the balloon and adds $0.05 to it.
• Collect banks the balloon's money and brings the next balloon.
• If the balloon pops, the money on that balloon is gone. Money you already banked is safe.

Harver does not publish the rules. Prep providers agree on the main details: JobTestPrep describes "39 rounds", balloons "in one of 3 colors", and "no time limit" (JobTestPrep); Game Assessment Prep's reconstruction uses blue, orange and yellow balloons, with orange the most fragile and blue the sturdiest (Game Assessment Prep, updated 28 September 2026). Treat the colour details as reconstruction, not specification. Your invitation's instruction screen is the source of truth.

The game sits inside the wider pymetrics assessment, which Harver describes as "12+ interactive, gamified experiences" taking "approximately 25 minutes" on average (Harver). pymetrics' help centre adds that each game "lasts only 1-3 minutes" and that you cannot pause a game once it starts (pymetrics Support Center).

What does the balloon game measure?

Harver lists risk tolerance, decision making and learning among the traits its games assess (Harver). The balloon game is where risk tolerance is easiest to see, and prep sites describe two layers to it:

1. Risk-taking. How far you push each balloon before banking. JobTestPrep frames it as how cautiously or adventurously you bank (JobTestPrep).
2. Learning. Whether you notice that the colours behave differently and adjust. JobTestPrep frames it as spotting that "each balloon color has a different explosion pattern". Game Assessment Prep adds loss sensitivity: whether you pump less after a balloon of that colour pops.

Neither layer has a correct setting. A cautious profile and a bold profile are both normal human profiles, and each fits some jobs better than others. What reads poorly is play that looks random: bold on one balloon, timid on the next of the same colour, with no visible reason.

How is the balloon game scored?

There is no pass mark and no single balloon score that an employer sees. pymetrics compares your whole trait profile with a profile built for the role. When you reuse results for a second application, the support centre says "your existing gameplay data will be re-assessed against the new role profile you have applied to" (pymetrics Support Center). Prep providers report that the recruiter's report shows "Matching traits" and "Traits to probe" rather than raw numbers (JobTestPrep).

So the money you bank is not your result. A candidate who banks less by playing carefully can match a role better than one who banks more, if the role's profile favours caution. This is why "maximise the money" advice from some prep sites misses the point. JobTestPrep itself offers two strategies, a "Max Money" one and a "Max Balloons" one, depending on which profile you want to show; the risky part is performing a profile that is not yours, because it is hard to keep up consistently across a dozen games, and you do not know which profile the role is looking for.

How should you play the balloon game?

There is no winning pump count, but there is a calm, consistent way to play that shows your real profile clearly:

• Pick a starting rule for each colour and stick to it for a few balloons. For example, "about ten pumps on a colour I have not seen pop yet". A rule makes your choices look deliberate, because they are.
• Learn per colour, not per balloon. One pop is weak evidence. Two or three pops of the same colour at low pump counts is a pattern; adjust that colour, not all of them.
• Do not chase losses. Pumping harder on the next balloon to win back a popped one turns a steady profile into a noisy one. Game Assessment Prep's guide warns against it too.
• Collect on purpose. Decide your stopping point before you start pumping each balloon, then stop there. Reflexive clicking is what makes a run look erratic.
• Ignore the clock. Prep sites report no time limit, so there is nothing to gain by rushing.

None of this changes who you are. It removes the noise that makes your real profile hard to read.

How to practise the balloon game

pymetrics says "you cannot preview the games before playing" and "there is no prep time required" (pymetrics Support Center). That is true of the trait itself. What a practice run does is make the format familiar, so the first balloon on the day is not the first balloon you have ever seen.

Our balloon game uses the same pump-and-collect mechanic: $0.05 a pump, Inflate with Space and Exchange with Enter, and a popped balloon loses its money. It is not a copy of Harver's game. It has 45 balloons that differ by shape (round, wide, tall) rather than colour, because it is modelled on the Arctic Shores version of the task (see our Arctic Shores balloon game guide). The learning layer is the same idea: each shape bursts within its own range, reshuffled every session, so players who notice which shape is fragile bank more.

What first attempts look like on GamePrep. Among 53 players' first attempts on GamePrep (October to November 2026), the median banked $8.90 across 45 balloons. The middle half banked between $8.10 and $10.50; the top tenth banked $12.12 or more and the bottom tenth $6.00 or less. That is a small sample, so read it as a rough picture. For scale, the norm table on our game was set from the expected value of simple strategies on its burst ranges: pumping every balloon five times averages about $8.90, ten times about $10.80, and adjusting by shape about $13.30. A typical first attempt, in other words, banks about what a cautious fixed rule would, and the gap above it comes from noticing the shapes. Our score benchmarks page shows the latest figures.

Remember the limit: our score rewards learning the shapes, because that is the part we can measure fairly. pymetrics does not grade you on money. Use the practice to get comfortable with the mechanic, then on the day play the way you naturally would.

Where the balloon game sits in the pymetrics battery

The balloon is one of the "play it straight" games. The others in that group (cards, money exchange, hard-or-easy, keypresses) also read traits. A few games do reward practice because they measure capacity: the digits game (memory span), the arrows game (attention) and, to a degree, the stop game and the tower game. Our complete pymetrics games guide covers all of them, and Harver vs pymetrics explains why your invitation may carry either name.

FAQ

How many balloons are in the pymetrics balloon game? Prep providers report 39 balloons in three colours. Harver does not publish the number, so check the instruction screen in your invitation.

How much is each pump worth? Prep providers consistently report $0.05 per pump. It is play money; nothing is paid out.

Is there a right number of pumps? No. pymetrics' support centre says the games have no right or wrong answers. The game reads your risk tolerance and how you adjust after pops, and different roles favour different profiles.

Do the balloon colours matter? Prep sites report that each colour has its own popping pattern, so noticing which colours pop early is part of what the game reads as learning. Their exact odds are reconstructions, not Harver figures.

Is there a time limit? Prep providers report none. pymetrics says each game lasts 1 to 3 minutes and cannot be paused once started.

Can I retake the pymetrics balloon game? Not freely. pymetrics allows candidates "to play the games once every 330 days", and within that window your existing results are re-assessed for each new role you apply to.

Is the balloon game colour-blind friendly? pymetrics offers gameplay accommodations for ADHD, dyslexia and colour blindness, chosen in a "Game Configuration" pop-up before you start (pymetrics Support Center).

Is the pymetrics balloon game the same as the Arctic Shores one? The mechanic is the same lab task, but the details differ: Arctic Shores' version shows 45 balloons of different shapes, while prep sites describe pymetrics' as 39 balloons of different colours.

GamePrep is an independent practice platform and is not affiliated with pymetrics or Harver. Game sets vary by employer and change between cycles; treat your invitation and its instructions as the source of truth. We provide related practice only, never leaked questions or answers.