Social Cognition
Die-Roll Task
Also called: die-rolling task, die-under-the-cup task, mind game, cheating game, Fischbacher and Föllmi-Heusi paradigm
Measure honest reporting at the group level: participants privately think of a number, roll an on-screen die and say whether it matched, scored by the share of Yes answers against the 1 in 6 chance predicts.
4-6 minUntimed
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Preview any test free after you sign up.
- Papers, 2021 to 2025
- 35
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures honest reporting at the group level. Before each roll of an on-screen die, the participant privately thinks of a number from 1 to 6, then reports whether the die showed it; each match earns points. Chance predicts a match on 1 roll in 6, and the test never learns the number.
Constructs
- Honest reporting
- Dishonesty measured at the group level
- Response to incentives
- Moral decision making
Research uses
- Behavioral economics research on honesty
- Moral psychology and integrity research
- Studies of incentives and rule following
- Group comparison within a study
How the paradigm works
The Die-Roll Task measures honest reporting at the level of a group. In the classic version, people roll a die in private and report the number, which sets their payment. No single report can be checked, but a group's reports can be compared with the spread of numbers an honest group would give (Fischbacher and Föllmi-Heusi, 2013). This version is the "mind game" (Jiang, 2013; Kajackaite and Gneezy, 2017): before each roll the participant thinks of a number from 1 to 6 and keeps it private, the die on the screen rolls, and the participant answers Yes if it showed that number and No if not. Each Yes earns points. The test knows every roll but never the number thought of, so it cannot tell a true Yes from a false one.
More about the paradigm
The standard session has 20 rolls, as in Jiang's online version (1 to 40 as a setting; 1 gives the one-shot game). Each roll lands on a whole number from 1 to 6 with equal chance, drawn from the session's seed and saved with the answer. Each Yes earns 10 points by default (1 to 1,000); the study decides what points are worth, and the payment line says whether they are exchanged for money. One check question with feedback comes first, and there is no time limit. As in Kajackaite and Gneezy (2017), the instructions name neither honesty nor lying, and studies usually explain the purpose of the task afterwards.
The main score is the share of rolls answered Yes. Anyone reporting truthfully answers Yes on 1 roll in 6 (0.167) on average, whatever numbers they tend to think of. Over a group, the mean share minus 1/6 estimates how often reports went beyond chance: in Kajackaite and Gneezy's one-shot mind game, 32% to 49% of participants answered Yes, depending on the stake, against the 16.7% that chance predicts. For one person a high share can arise by chance; with 20 rolls, about 1 truthful person in 10 answers Yes 6 or more times. The first roll's answer, the share in each half of the rolls and the time to answer are also reported, and answers under 100 ms count as anticipations. The scores are research measures for group comparisons: they never show that one person reported falsely, and they do not diagnose.
Scores
- Reported Match Rate%
- Share of rolls answered Yes: the die showed the number thought of. Chance predicts 1 in 6 (0.167).
- Compare with 0.167 over a group
- Rate Above Chance%
- The reported match rate minus 1/6. Its mean over a group estimates how often reports went beyond chance.
- A group-level measure; for one person it can be above 0 by chance
- Matches Reportedcount
- Rolls answered Yes.
- Informational
- First Roll Answered Yes0/1
- 1 when the first roll was answered Yes, 0 when No; comparable with one-shot designs.
- Compare with 1 in 6 over a group
- Match Rate, First Half%
- Share of Yes answers over the first half of the planned rolls.
- Informational
- Match Rate, Second Half%
- Share of Yes answers over the second half of the planned rolls.
- Informational
- Median Answer Timems
- Median time from the landed die to the answer, over answers at or after 100 ms.
- Informational
- Median Yes Timems
- Median time to answer Yes, over answers at or after 100 ms.
- Informational
- Median No Timems
- Median time to answer No, over answers at or after 100 ms.
- Informational
- Points Wonpoints
- Points earned over all rolls.
- Informational
- Check Correctflag
- 1 when the check question on when to answer Yes was answered correctly, 0 when not.
- 1 supports interpretation
- Answers Under 100 mscount
- Answers given less than 100 ms after the die landed.
- 0 expected
- Rollscount
- Rolls answered.
- Informational
Use and evidence
35 papers from 2021 to 2025 name this task in their title or abstract, up 30% compared with 2016 to 2020. That ranks it 279 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 10 |
| 2024 | 4 |
| 2023 | 3 |
| 2022 | 10 |
| 2021 | 8 |
| 2020 | 6 |
| 2019 | 5 |
| 2018 | 7 |
| 2017 | 6 |
| 2016 | 3 |
| 2015 | 1 |
| 2014 | 0 |
| 2013 | 0 |
| 2012 | 1 |
| 2011 | 1 |
| 2010 | 0 |
| 2009 | 0 |
| 2008 | 0 |
| 2007 | 0 |
| 2006 | 0 |
| 2005 | 0 |
| 2004 | 0 |
| 2003 | 0 |
| 2002 | 0 |
| 2001 | 0 |
| 2000 | 0 |
Use grade D
30 to 99 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
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 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Each roll
- The participant thinks of a number from 1 to 6 and keeps it private, presses Roll, and answers Yes if the die shows that number or No if it does not. The test never asks for the number.
- Rolls
- 20 (standard; 1 to 40). Each lands on a whole number from 1 to 6 with equal chance, drawn from the session's seed and saved with the answer.
- Points
- 10 per Yes (standard; 1 to 1,000). Points are hypothetical by default; a setting changes the instruction line to say the study pays.
- Check
- One question with feedback before the first roll: the participant thought of 4 and the die shows 2.
- Wording
- The instructions name neither honesty nor lying. There is no time limit.
- Scores
- The share of rolls answered Yes, against the 1 in 6 that chance predicts, and that share minus 1/6. Also the first roll's answer, each half of the rolls and the time to answer.
Fischbacher and Föllmi-Heusi (2013) had people roll a real die in private and report the number. Online, the test rolls the die, so this version uses the mind game of Jiang (2013) and Kajackaite and Gneezy (2017): the private part is the number the participant thinks of, which the test never learns. A share of Yes answers above 1 in 6 is meaningful over a group, but one person can exceed it by chance, so the scores never show that a participant reported falsely. Studies usually explain the purpose of the task afterwards. The scores support research 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 answers, one Yes or No per roll. The number the participant thought of is never recorded, so no answer is right or wrong, accuracy is not judged against chance, and one answer throughout is accepted. A run short of its planned rolls and a wrong check answer are flagged for review. Answers under 200 ms on at least half the rolls invalidate.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. The test cannot tell a true Yes from a false one, so it judges only how the answers were given, never what they say about the participant. 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 |
|---|---|---|
| Rolls | 20 | 1 to 40 |
| Points per match | 10 | 1 to 1,000 |
| 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 Die-Roll Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Fischbacher U, Föllmi-Heusi F, 2013; Kajackaite A, Gneezy U, 2017), with the standard settings (rolls: 20; points per match: 10; payment line: Hypothetical points). The main score was reported match rate (share of rolls answered Yes: the die showed the number thought of).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Fischbacher, U., & Föllmi-Heusi, F. (2013). Lies in disguise: An experimental study on cheating. Journal of the European Economic Association, 11(3), 525-547. https://doi.org/10.1111/jeea.12014The private die roll, and comparing a group's reports with chance.
- Jiang, T. (2013). Cheating in mind games: The subtlety of rules matters. Journal of Economic Behavior & Organization, 93, 328-336. https://doi.org/10.1016/j.jebo.2013.04.003Choosing in the mind before an online die is thrown, over 20 rounds.
- Kajackaite, A., & Gneezy, U. (2017). Incentives and cheating. Games and Economic Behavior, 102, 433-444. https://doi.org/10.1016/j.geb.2017.01.015Thinking of a number from 1 to 6 and answering whether the die matched it.
Source papers
- Kajackaite and Gneezy (2017): in the one-shot mind game, 32%, 47%, 41% and 49% of participants answered Yes at stakes of $1, $5, $20 and $50, against 16.7% by chance (75 participants per stake). Jiang (2013, working paper version): over 20 rounds of an online mind game with two choices, the choice that paid more was reported on 64% of rounds against 50% by chance.Paradigm source, Adults
Questions
How long does the Die-Roll Task take?
About 4 to 6 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?
Reported Match Rate, Rate Above Chance, Matches Reported, First Roll Answered Yes, Match Rate, First Half, Match Rate, Second Half, Median Answer Time, Median Yes Time, Median No Time, Points Won, Check Correct, Answers Under 100 ms and Rolls. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Die-Roll Task used in research?
35 papers from 2021 to 2025 name it in their title or abstract, up 30% compared with 2016 to 2020. That ranks it 279 of 395 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Die-Roll Task 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?
Rolls (1 to 40), points per match (1 to 1,000) 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.
Related tests