Skip to main content
Back to Blogs
Provider Guide5 min read

Pymetrics Tips: What the 12 Games Measure and How to Sit Them

Pymetrics has no score. It builds a trait profile and compares it to a model of people who do the role well. What each game measures and why strategic answering backfires.

Pymetrics, now part of Harver, is used by employers including Unilever, LinkedIn, Accenture and JPMorgan, usually as an early screen. The whole assessment runs 25 to 30 minutes across around twelve short games. The single most important thing to understand before you sit it is that there is no score. There is a trait profile, and it is compared against a model built from people who do the role well.

Why that changes what preparation means

If there is no score, there is nothing to maximise. Risk tolerance is not better high or better low; it depends entirely on the role model, and you cannot see the role model. Deciding to appear bold, or generous, or cautious is therefore not strategy, it is guessing at a hidden target while giving up the one profile you can produce reliably, which is your own.

There is a second reason it backfires. Pymetrics flags inconsistent responding, and several games are built as pairs that look at the same underlying preference from different angles. Money Exchange I asks how much you send to another player; Money Exchange II asks how you respond to offers. Answering one honestly and the other strategically produces exactly the incoherence the design is looking for.

What the games measure

Risk
Balloon
Risk tolerance and learning from feedback. You pump for reward with a rising chance of losing it.
Risk
Cards
Learning from feedback and risk management, using four decks with hidden reward and penalty patterns.
Memory
Digits
Working memory span, with sequences that lengthen until you fail. Hitting the wall is the measurement.
Attention
Arrows
Attention and the ability to ignore conflicting information around the target.
Attention
Stop
Impulse control, measured by how well you withhold a response you have already begun.
Attention
Keypress
Processing speed and motor consistency over a short, sustained burst.
Perception
Lengths
Perceptual accuracy and how consistent your attention stays across many rapid trials.
Emotion
Faces
Emotion recognition from brief facial expressions.
Social
Money Exchange I and II
Trust, generosity, fairness and how you respond to unfair treatment, viewed from two sides.
Motivation
Hard or Easy
Effort allocation, through repeated choices between a harder task with a bigger reward and an easier one.
Planning
Tower
Planning and working memory, moving discs between pegs in as few moves as possible.

Adaptive games are supposed to end in failure

Digits, Stop, Arrows and Lengths all adapt. Digits keeps lengthening the sequence until you cannot hold it. The Stop task adjusts its timing so that you fail to stop roughly half the time. Failing is not a mistake in these games; failing is where the measurement happens, because your threshold is what is being located.

Knowing this matters more than it sounds. Candidates who interpret the first failure as having blown the assessment carry that into the next game, and anxiety does affect the traits being measured. Expect to hit a wall in the memory game. That is the wall being found.

What practice genuinely buys you

Practice will not and should not change your underlying tendencies. What it removes is the cost of learning mechanics during a measured task. On the balloon game, a candidate seeing it for the first time spends the first several balloons working out what the buttons do, and those balloons still count. The same is true of Tower, where the rules take a round or two to internalise, and of Money Exchange, where the structure of the exchange is not obvious on first read.

Sitting each game once beforehand means test day starts at your actual behaviour rather than at your confusion. That is the entire legitimate benefit, and it is a real one. You can play simulations of all twelve games here.

Set-up advice that actually affects results

  • Use a mouse and a keyboard, not a trackpad. Several games measure reaction time and motor consistency, and trackpad latency is noise in exactly the signal being recorded.
  • Do it in one sitting, undisturbed. The assessment is short and the attention games are sensitive to interruption.
  • Read each instruction screen fully. Several games change what they ask for partway through, and the instruction screen is where that is announced.

How Pymetrics differs from Arctic Shores

Both are game-based, and candidates confuse them constantly. Pymetrics is rooted in neuroscience research and leans towards cognitive and emotional trait measurement. Arctic Shores is a UK platform with a heavier emphasis on behavioural traits and learning agility, and sends a larger set of around fourteen games. Check which one your employer uses before you prepare, since the game mechanics have very little overlap.

Ready to Start Practicing?

Apply these strategies with our comprehensive practice platform

Start Practising Free