Pymetrics Games Risk and Decision-Making 4 — Questions and Answers
Question 1: In the Pymetrics 'Money Exchange' game, if you are the proposer, offering a split that feels fair to the other party (even if not equal) is strategically important because:
- The algorithm always rewards 50/50 splits only
- The responder can reject the offer, leaving both parties with nothing (Correct answer)
- Proposers get a bonus for generosity
- Responders must always accept any positive offer
Correct answer: The responder can reject the offer, leaving both parties with nothing
In ultimatum-style exchanges, responders who feel treated unfairly often reject offers, resulting in zero for both parties.
Question 2: Which of the following best describes 'framing effects' in the context of Pymetrics risk tasks?
- Choosing differently based on whether outcomes are described as gains or losses, even when mathematically identical (Correct answer)
- The visual layout of the game interface influencing choices
- The order in which answer options are presented affecting selection
- Background noise in the testing environment biasing decisions
Correct answer: Choosing differently based on whether outcomes are described as gains or losses, even when mathematically identical
Framing effects occur when logically equivalent options are chosen differently because one is framed as a gain and another as a loss.
Question 3: A Pymetrics candidate is risk-averse in the gain domain but risk-seeking in the loss domain. This pattern is most consistent with:
- Expected utility theory
- Prospect theory as described by Kahneman and Tversky (Correct answer)
- The efficient market hypothesis
- The Kelly criterion for bet sizing
Correct answer: Prospect theory as described by Kahneman and Tversky
Prospect theory specifically predicts risk aversion for gains and risk-seeking for losses, reflecting how people evaluate outcomes relative to a reference point.
Question 4: In a Pymetrics task where you must decide how much to invest in a risky project, which strategy would likely score best for roles requiring financial risk management?
- Invest the maximum every round to show boldness
- Invest nothing to demonstrate capital preservation
- Calibrate investment size to the risk-reward profile of each round (Correct answer)
- Invest randomly to appear unpredictable
Correct answer: Calibrate investment size to the risk-reward profile of each round
Calibrating investment to each round's risk-reward ratio demonstrates sophisticated financial judgment and disciplined risk management.
Question 5: When Pymetrics games require you to learn which of several options pays out most often, what learning style does the algorithm typically reward?
- Sticking to the first option chosen without exploring
- Systematic exploration followed by exploitation of the best option (Correct answer)
- Randomly switching options to collect all payouts
- Only choosing the option with the highest single observed payout
Correct answer: Systematic exploration followed by exploitation of the best option
Systematic exploration to identify the best option, then exploiting it, reflects adaptive reinforcement learning that Pymetrics scores favorably.
Question 6: In Pymetrics games involving interpersonal trust (e.g., deciding how much money to send to a partner who may or may not return it), high senders who receive nothing back and then send high amounts again are demonstrating:
- Rational updating of trust based on evidence
- Persistent high baseline trust regardless of feedback (Correct answer)
- Strategic deception to confuse the algorithm
- Optimal tit-for-tat strategy
Correct answer: Persistent high baseline trust regardless of feedback
Continuing to send high amounts after being betrayed suggests an unconditionally high trust baseline rather than evidence-based trust calibration.
Question 7: In a Pymetrics risk task that tracks your performance across 20 rounds, which learning curve would an algorithm most favorably interpret?
- Flat performance across all 20 rounds
- Improving performance after early exploration as you identify optimal strategies (Correct answer)
- Declining performance indicating increasing fatigue
- Highly variable performance with no discernible pattern
Correct answer: Improving performance after early exploration as you identify optimal strategies
An upward learning curve, especially after early exploration, signals adaptive intelligence and the ability to optimize strategy over time.
Pymetrics Games Risk and Decision-Making 4
In the Pymetrics 'Money Exchange' game, if you are the proposer, offering a split that feels fair to the other party (even if not equal) is strategically important because: