The document describes CYCLES, a serious game designed to train people to recognize and mitigate cognitive biases. It discusses two games that target specific biases like confirmation bias and representativeness. Gameplay involves puzzle-like challenge rooms that expose players to biases. Studies found the game significantly improved bias knowledge from pre-to-post-testing and eight weeks later. The document outlines challenges in generalizing the game to new audiences like law enforcement, teenagers, and businesses. It provides examples of adapting content and mechanics to make the concepts more age-appropriate and relevant to different groups.
2. What is CYCLES?
• Part of IARPA program, designed to train
Intelligence Analysts
• Serious game that teaches bias recognition,
discrimination, mitigation
• Two games
– Phase 1: Fundamental Attribution Error,
Confirmation Bias, and Bias Blind Spot
– Phase 2: Anchoring, Projection, and
Representativeness
•Proven to train effectively
3. 3
Game Video Links
• Trailer Game 1
https://www.youtube.com/watch?v=5QqY4KmX1S4&feature=youtu.be
• Trailer Game 2
• https://www.youtube.com/watch?v=-St0x9a5uq4&feature=youtu.be
4. 4
CYCLES: Basic Lesson
We (our brains) are and always will be biased
BUT
There are simple thinking tools that can help
Always remember your thinking tools
CYCLES games let players experience this lesson
5. What are Cognitive Biases?
• Confirmation Bias
– People search for or interpret information in a way that confirms their
preconceptions. Often preceded by priming.
• Fundamental Attribution Error
– People over-emphasize personality-based explanations for behaviors
observed in others while under-emphasizing the role and power of
situational influences on the same behavior.
• Bias Blind Spot
– Individuals are unaware of their own cognitive biases, even when they
can recognize cognitive biases in others.
6. CYCLES Game - Phase 1
• Design: 30 minute, casual
3rd person puzzle-style
game
• Series of “challenge
rooms”, “review rooms”
Confirmation Bias
Fundamental Attribution
Error
Bias Blind Spot
8. 8
Implementation: Challenge Types
Real world text-based quizzes at end of levels
Challenge Review Hub
• Review knowledge
• Real world
examples
• Learn and observe
bias
• Core bias training
• Examples of bias
• Walk-through
complex concepts
• Reiterate lessons
9. Repetition vs Length
Bias Reduction
Immediately after and 8-weeks after playing
Exp 3 Repetition Exp 4 Length
Immediate
Immediate
Immediate
Immediate
8-weeks
8-weeks
8-weeks
8-weeks
Video avg.
post 11%
All conditions have sig. pre to post change; sig. different than video
10. Bias Knowledge More than Doubled
Best game: Repeated Play Exp 3a
Video avg.
post 99%
Video avg.
8-week 33%
Pre to Post N = 335
8-week N = 198 All conditions statistically significant (p<.05) difference with pre-test
Immediate
Immediate
8-weeks
8-weeks
Knowledge Improvement
11. 11
CYCLES Game - Phase 2
• ~60 minute, 1st person puzzle
adventure game
• Premise: Mr. Q will train you to
survive decision-making
challenges
• A survival experience set in a
remote carnival
Anchoring Bias
Projection Bias
Representativeness
Bias
12. 12
Phase 2 Variables
Rewards Interactivity SATs
Choices in content and
instruction vs. straight-
through play
Practice with
Structured
Self-Critique-
like system vs.
without
Sounds, visuals, prizes,
bronze-silver-gold
ratings accompanying
feedback vs. feedback
alone
13. 13
Implementation: Challenge Types
Real world text-based quizzes at end of levels
Tents Booths Shops
• Practice decision-
making
• Based on carnival
games
• Point-and-click
puzzles
• Learn and observe
bias
• Core bias training
• Examples of bias
• Walk-through
complex concepts
• Apply decision-
making
• Familiar settings
(shopping)
• Point-and-click
puzzles
14. 14
CYCLES Design Learnings
• Narrative Alignment
– Align player role with concepts
– Trustworthiness of feedback, player
coaching paradigm
• Bias Elicitation
– Players may believe they are not
biased
• Simplify Interactions
– Phase 1 is a more guided experience
than Phase 1
– Simplified language, mitigations
• Infographics for Complex Material
15. 15
Design Challenges
• Emphasize the learner is not at fault
• Challenge difficulty “sweet spot”
• Text density balance
• Dynamic sensitivity to learning progress
• Value of failure and replay
• “Slow down” mechanics
• Humor is important!
• Practice needs to be built in
• Literacy supports
16. 16
Challenges for General Adaptation
• Content easily updated to be applicable for target audience
– Quiz questions
• Target platform
– Flexible code base
• People could feel like there was something wrong with them
when introduced to the topic.
– Solution: Optical illusion
● Cognitive biases are difficult to demonstrate on an individual
scale
18. 18
Design challenges for LEO
● Targeting
○ Law Enforcement Officers
○ First Responders
● Feedback
○ Narrative Context - Superhero training
Target audience think of themselves as superheroes
○ Order -
•Phase 2 game (superhero training classes)
•Phase 1 game (superhero mission)
bonus, this training ends on Bias Blind Spot
● Tradeoff between muscle memory training and the “slow
down” message
19. 19
Design challenges for empathy
Cognitive Bias and thought processes affecting empathy are
similar
Biases we proposed teaching:
• Projection Bias
• Fundamental Attribution Error
• Confirmation Bias
• Bias Blind Spot
Some existing CYCLES content was less applicable for CYCLES of
Empathy
Explaining to students why empathy was important to learn
20. 20
Challenges for Teen audience
Complex language and concepts
(“Representativeness”), amount of
reading
Younger audience have more limited
attention span than a college or
professional audience
Player motivation - original audience was
more intrinsically motivated
21. 21
Key Adaptations
• Went from a 9th grade reading level to 5th / 6th
on Flesch Kincaid.
• Real-world quiz questions were made more
audience centric.
• Used the more direct mechanics from CYCLES
Phase 2
• Narrative: “School for not being a jerk”
23. 23
Bias -> Empathy
Representativeness Use to get buy in on bias (with simpler
problems like coin flips and probability)
Anchoring Bias (Feelings sub-bias) to talk about how our
strong emotional association with an event can
prevent us from seeing the whole event clearly
Projection Bias (renamed Similarity) to describe how others
like different things than they do
Bias Blind Spot Used learnings from our other game to inform
the design for adding Bias Blind Spot to Ph 2
(New) Used existing teaching structure and
functionality for new content Feelings Bias
24. 24
Design challenges for Business
● Forecasting
● Hiring
● Artifical Intelligence
● Risk Assessment
● Communication
25. 25
Change Management
Q x A = E
Even though some design
variables were established as
not affecting teaching
effectiveness, they do affect
marketing/perception
Example: Minimalistic art in
Phase 1 game