Social Cognition
Traveler's Dilemma Game
Also called: traveller's dilemma, travelers dilemma game
Measure how far people undercut a simulated partner: each round both name a number from 100 to 200, the lower one counts for both with a bonus of 5 or 100 points, scored by the mean number and the bonus effect.
4-7 minUntimed
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Try it here first, with no account.
- Papers, 2021 to 2025
- 13
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how far below the highest number people name in the traveler's dilemma, and how the size of the bonus moves them. In each round the participant and a simulated partner name a number from 100 to 200; the lower number counts for both, with a bonus of 5 or 100 points for the one who named it.
Constructs
- Strategic reasoning
- Steps of reasoning about a partner
- Expectations about others' choices
- Weighing a bonus against a shared payoff
Research uses
- Behavioral economics research on bounded rationality
- Game theory experiments on equilibrium and learning
- Studies of strategic reasoning
- Group comparison within a study
How the paradigm works
The Traveler's Dilemma Game measures how far below the highest number people name when undercutting a partner pays a little, and how the size of that payment moves them. In the game, which Basu (1994) named, two players name a whole number in a range at the same time. The lower number counts for both, and the player who named it gets a bonus taken from the other. When the bonus is above 1, the best reply to a known number is one below it, so only the lowest number survives step-by-step reasoning about what a rational partner would name: it is the game's only equilibrium, whatever the bonus (Goeree and Holt, 2001). People rarely play it when the bonus is small. Goeree and Holt (2001) had 50 paid students play once with numbers from 180 to 300: with a bonus of 180, close to 80% named the lowest number, and with a bonus of 5, roughly the same share named the highest, an average of 280. Played over repeated rounds with new partners and numbers from 80 to 200, average numbers stayed above 190 with a bonus of 5 and fell to near the lowest number with bonuses of 50 and 80 (Capra, Goeree, Gomez and Holt, 1999).
More about the paradigm
Here the participant plays 20 rounds, each with a new partner, and types a whole number from 100 to 200. The bonus is 5 points in 10 rounds and 100 points in the other 10, each bonus once in every two rounds, in an order drawn from the session's seed. The partners are simulated. With a chance of 0.8, each names the number Goeree and Holt's players named most often at that kind of bonus, the highest with the small bonus and the lowest with the large one, and otherwise any number from 100 to 200, from numbers drawn before the session, so nothing the participant does changes them. Against these partners, naming 199 with the small bonus and 100 with the large one earns the most points on average. The standard instructions say each partner is a computer program that follows set rules; a neutral wording leaves that line out, and the last screen then says so. Studies usually explain the purpose afterwards. One check question with feedback comes first, there is no time limit, and after each round the participant sees the partner's number and the points. The study decides what the points are worth.
The main score is the mean number named. It is also reported for each bonus, and the bonus effect is the mean with the small bonus minus the mean with the large one, the effect Goeree and Holt found and the equilibrium does not predict. The share of small-bonus rounds at the highest number, the share of large-bonus rounds at the lowest number, the first number with each bonus and the points won are reported too. A high number can reflect a hope that the partner names a high number too, a guess about a program or a wish to keep both players' points high, so the scores describe choices in a game played for points, never a person's character, and they do not diagnose. People who know the partners are programs may name different numbers from people playing other people, so compare people who saw the same instructions.
Scores
- Mean Numberscore
- Mean of the whole numbers from 100 to 200 the participant named over the rounds.
- Read across a group; no better direction
- Mean Number, Small Bonusscore
- Mean number named in the rounds with the bonus of 5 points.
- Informational
- Mean Number, Large Bonusscore
- Mean number named in the rounds with the bonus of 100 points.
- Informational
- Bonus Effectscore
- Mean number with the small bonus minus the mean number with the large bonus.
- Above 0 means higher numbers when undercutting pays little
- Highest Number Rate, Small Bonus%
- Share of the small-bonus rounds in which the participant named 200, the highest number.
- Informational
- Lowest Number Rate, Large Bonus%
- Share of the large-bonus rounds in which the participant named 100, the lowest number and the game's equilibrium.
- Informational
- First Number, Small Bonusscore
- The number named in the first round with the small bonus.
- Informational
- First Number, Large Bonusscore
- The number named in the first round with the large bonus.
- Informational
- Points Wonpoints
- Points over all rounds.
- Informational
- Check Correctflag
- 1 when the check question on the points for two different numbers was answered correctly, 0 when not.
- 1 supports interpretation
- Median Choice Timems
- Median time from a round's screen to the number sent, over numbers sent at or after 100 ms.
- Informational
- Numbers Sent Under 100 mscount
- Numbers sent less than 100 ms after the round's screen appeared.
- 0 expected
- Roundscount
- Rounds played.
- Informational
Use and evidence
13 papers from 2021 to 2025 name this task in their title or abstract, down 48% compared with 2016 to 2020. That ranks it 437 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 3 |
| 2024 | 1 |
| 2023 | 5 |
| 2022 | 1 |
| 2021 | 3 |
| 2020 | 3 |
| 2019 | 3 |
| 2018 | 6 |
| 2017 | 5 |
| 2016 | 8 |
| 2015 | 6 |
| 2014 | 12 |
| 2013 | 8 |
| 2012 | 6 |
| 2011 | 12 |
| 2010 | 4 |
| 2009 | 9 |
| 2008 | 4 |
| 2007 | 3 |
| 2006 | 1 |
| 2005 | 2 |
| 2004 | 1 |
| 2003 | 2 |
| 2002 | 0 |
| 2001 | 0 |
| 2000 | 1 |
Use grade E
Fewer than 30 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 0
- Registered clinical studies using it as an outcome
- 0
Note on the counts: includes game-theory papers that run no task.
How the grades work
Use: A is 1,000 or more papers from 2021 to 2025, B 300 to 999, C 100 to 299, D 30 to 99 and E fewer than 30. Evidence: A means at least 25 papers report its test-retest reliability and at least 25 registered clinical studies use it as an outcome; B, at least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers; C, at least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies; D, little published measurement evidence found. The counts measure use. They include every paper that names the task in its title or abstract. Grades describe the published research on this task, not ConductCognition's version of it.
Source: ConductCognition test census 0.22.0, counted 3 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Each round
- The participant and a new partner each type a whole number from 100 to 200 at the same time. Equal numbers give each player that many points. Otherwise the lower number counts for both: the player who named it gets the round's bonus added, and the other loses the bonus.
- Bonus
- 5 points in half the rounds and 100 points in the other half, shown at the top of each round. Each bonus comes once in every 2 rounds, in an order drawn from the session's seed and saved.
- Rounds
- 10 for each bonus (standard; 4 to 20), 20 in all. Points are hypothetical by default; a setting changes the instruction line to say the study pays.
- Partners
- Computer programs, a new one in each round. With a chance of 0.8, a partner names the number Goeree and Holt's players named most often at that kind of bonus: 200 with the small bonus and 100 with the large one. Otherwise it names any whole number from 100 to 200, each equally likely. Their numbers are drawn from the session's seed before the first round and saved, and nothing the participant does changes them. The standard instructions say each partner is a computer program that follows set rules; a neutral wording leaves that line out, and the last screen then says so.
- Results
- After each number, for 2.5 seconds: the partner's number and the participant's points for the round.
- Check
- One question with feedback before the first round: the points for naming 150 when the partner names 120, with the small bonus.
- Scores
- The mean number, overall and for each bonus; the bonus effect, the small-bonus mean minus the large-bonus mean; the share of small-bonus rounds at 200 and of large-bonus rounds at 100; the first number with each bonus.
100 is the game's only equilibrium at either bonus, yet Goeree and Holt (2001) found that a small bonus moves most people to the highest number and a large one to the lowest. Against these partners, naming 199 with the small bonus and 100 with the large one earns the most points on average. Here the partners are computer programs, the numbers and points are ConductScience's own, and the screens say number and bonus. People who know the partners are programs may name different numbers from people playing other people, so compare people who saw the same instructions. Studies usually explain the purpose afterwards. The scores describe choices in a game played for points, never a person's character, and do not diagnose.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: the scored run's numbers, one per round, from 100 to 200. No number is right or wrong, so accuracy is not judged against chance, and the same number in every round is accepted. A run short of its planned rounds and a wrong check answer are flagged for review. Numbers sent under 200 ms on at least half the rounds invalidate.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. The lowest number is the only equilibrium, and the highest number is what most people name when the bonus is small (Goeree and Holt, 2001), so naming one number throughout is a real way to play. An answer under 200 ms is faster than a simple visual reaction, so it cannot reflect the item. Following response-time effort screening for untimed tests (Wise and Kong 2005), the session is invalid only when at least half the answers are that fast.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Rounds per bonus | 10 | 4 to 20 |
| Partner wording | Computer programs (standard) | Computer programs (standard), Neutral: your partner |
| Payment line | Hypothetical points (standard) | Hypothetical points (standard), The study pays for points |
Yearly plans can change settings within these limits. Free and pay-as-you-go studies use the standard settings.
Need a setting this test does not have?
Devices and languages
- Untimed test: runs on desktops, laptops, tablets and phones. Participant devices
- Languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese.
- Available on: Every plan, including the free plan.
- No percentile. Scores and quality flags only. How norms work
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Traveler's Dilemma Game in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Basu K, 1994; Capra CM, Goeree JK, Gomez R, Holt CA, 1999), with the standard settings (rounds per bonus: 10; partner wording: Computer programs; payment line: Hypothetical points). The main score was mean number (mean of the whole numbers from 100 to 200 the participant named over the rounds).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Basu, K. (1994). The traveler's dilemma: Paradoxes of rationality in game theory. American Economic Review, 84(2), 391-395.
- Capra, C. M., Goeree, J. K., Gomez, R., & Holt, C. A. (1999). Anomalous behavior in a traveler's dilemma? American Economic Review, 89(3), 678-690. https://doi.org/10.1257/aer.89.3.678Repeated rounds with new partners and bonuses from 5 to 80: higher numbers with smaller bonuses.
- Goeree, J. K., & Holt, C. A. (2001). Ten little treasures of game theory and ten intuitive contradictions. American Economic Review, 91(5), 1402-1422. https://doi.org/10.1257/aer.91.5.1402The game played once with a bonus of 5 and of 180, and the most common numbers the partners use here.
Source papers
- Goeree and Holt (2001): played once for money with numbers from 180 to 300, close to 80% of 50 students named the lowest number with a bonus of 180, and roughly the same share named the highest with a bonus of 5. Capra, Goeree, Gomez and Holt (1999): over repeated rounds with numbers from 80 to 200, average numbers stayed above 190 with a bonus of 5. Other ranges, bonuses and partners will give other numbers.Paradigm source, Adults
Questions
How long does the Traveler's Dilemma Game take?
About 4 to 7 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops, tablets and phones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Mean Number, Mean Number, Small Bonus, Mean Number, Large Bonus, Bonus Effect, Highest Number Rate, Small Bonus, Lowest Number Rate, Large Bonus, First Number, Small Bonus, First Number, Large Bonus, Points Won, Check Correct, Median Choice Time, Numbers Sent Under 100 ms and Rounds. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Traveler's Dilemma Game used in research?
13 papers from 2021 to 2025 name it in their title or abstract, down 48% compared with 2016 to 2020. That ranks it 437 of 495 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.22.0, counted 3 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Traveler's Dilemma Game have?
In the published research, 0 papers report its test-retest reliability and 0 registered clinical studies use it as an outcome (evidence grade D in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Rounds per bonus (4 to 20), partner wording (Computer programs (standard), Neutral: your partner) and payment line (Hypothetical points (standard), The study pays for points). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
Participants can take this test in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. English is on every plan, and the other languages come with yearly plans. Have a native speaker read the machine-translated text before launch.
Is this a clinical test?
No. ConductCognition is for research use only. It runs, scores and exports the test; clinical interpretation is outside the platform.
Every test: the full test list.
Run this test in your next study
Start free with 10 free participants on any test, then use participant credits or a yearly plan.