How can we use one’s perceptions of self to push for behavioral change?
Using cognitive behavioral theories to implement value-driven changes in League of Legends that will encourage a desire to be a kinder player.
Role(s)
Lead UX Researcher Product Designer
Duration
Aug 2023 - Present
Team
Geoff Kaufman (advisor)
Farhan Ahmed
Junghyun Choy
Laura Delince
Kyle Kawakami
Hnin Lei Lei Tun
Gaurav Nemade
Methods
Digital Prototyping
Usability Testing
Contextual Interviews
Surveys & Questionnaires
Focus Groups
Expert Consults
System Usability Scale
Why Tackle Gaming Toxicity?
Within multiplayer games, 74% of players have reported witnessing or experiencing toxic behavior. Many rationalize toxicity as an inherent, natural part of gaming, but research has shown that this type of behavior has long-term negative effects on players causing personal and interpersonal harm way beyond the screen.
Our team has been working with Prof. Geoff Kauffman to explore the root causes of online toxic behavior and experiment with cognitive behavioral theories to find persuasive interventions to gaming toxicity. Our focus is not to use extrinsic repercussions to punish toxic players, but rather to push players so that they are motivated to be kinder regardless of incentives or punishments.
How might we motivate players to have an intrinsic desire to exhibit kind behavior within online gaming environments?
How did we approach this?
Pick 3 values and describe why they resonate with you
When a player first makes an account, they will be prompted to pick and reflect on 3 values that they want to grow in. They will later be able to swap these as they see fit.
-
Players will become more self-aware by having to make a deliberate and effortful choice to reflect on their top values. This sets the stage for self-awareness theory which states that when people are self aware, they will evaluate their current behavior with their higher internal standards. If these differ, they are likely to change their behavior to escape this self-discrepancy.
See you and your teammates’ values before the match
Pre-game loading screen
Close-up of character card
As the match starts up, the loading screen will include every player’s values at the top of their avatar.
-
This uses priming to subconsciously activate the mental schemas associated with those values. Having this in mind right before the match will lead to value-driven behavior during the match.
This screen also leads users to perceive a collective adherence to values, creating a social norm. In order to conform to this norm, players will be more motivated to also adhere to their own listed values.
See you and your teammates’ values during the match
Overall in-game view
Close-up of teammate values
During the match, each players will see their own values next to their avatar. They will also be able to see their teammates’ values displayed on screen at all times.
-
By intentionally positioning the values right next to each player’s avatar, a direct association is created between identity and value (similar to how a username becomes an identity). By having values as identifiers, players are less likely to behave against them since doing so would cause cognitive dissonance. This is also a way of priming since the values are constantly in view.
Endorse your teammates at the end of the match
Once the match ends, players will be asked to endorse their teammates on their unique values. If they feel like their teammates demonstrated any of the values shown, they can click on them to endorse.
-
This feature will cause players to retroactively reflect on their own values and behavior. We have found that when having to evaluate others under a positive framework, players will tend to find at least one positive trait. This activates social proofing since they now have the knowledge that others are behaving well, thus putting more pressure on themselves to follow that norm and adhere to their values as well.
Progress through the value tiers and receive rewards
As players receive endorsements from their teammates, they will progress through that value’s tiers. There will be checkpoints within each tier that will reward players with in-game prizes.
-
Earlier tiers will have more checkpoints, giving more frequent rewards. As players reach the final tiers, rewards will become less frequent and more sporadic.
At the beginning, players will adhere to their values because of the extrinsic motivation of the rewards. However, as they approach the final tier, the motivation will shift to intrinsic. Due to the laborious journey to get to the top tier, players will feel like they have been behaving well because they truly embody the values. As the rewards turn less frequent and more sporadic, players become more convinced that their behavior is a core part of who they are regardless of rewards.
How did did we get here?
To explore the problem space and familiarize ourselves with the topic, we took the first couple of weeks to conduct background research. The goal was to understand the existing literature, the trends in the field, and the current situation for League players.
Researched existing literature in virtual gaming environments, online toxicity, impulsive behavior, and cyberbullying.
Interviewed experts in fields of gaming, HCI, and game design to understand their perspective.
- Ravi Kawade (Professional Dota 2 Player)
- Elaine Fath ( Lead Game Designer @ Center for Transformational Play at Carnegie Mellon)
- Erik Harpstead (System Scientist at Carnegie Mellon’s HCII)
- Julian Frommel (HCI & Games Researcher at Utrercht University)
Recruited a group of 6 League of Legends players at varying levels of experience and expertise to discuss their perceptions of the game environment regarding teammate interactions, toxicity, and their overall behavior in the game.
After gathering our notes and insights from our background research methods, we prioritized the following 4 main takeaways to take into account as we designed our solution:
Anonymity online allows people to feel like a detached sense of identity regarding their virtual personas. This can lead them to engage in toxic behavior that is “out of character”.
There is dissonance in behavior and values since players are aware that toxic behavior is “wrong”, but will still engage in it.
Players tend to blame others for their losses, suggesting a lack of self-accountability.
Most people report being victims of toxicity but few admit being perpetuators of it. This suggests a lack of self-awareness since most players have likely engaged in both.
Players feel a sense of helplessness when it comes to reporting toxic behavior since they don’t see direct consequences when they do it.
Toxicity has been normalized within League of Legends, making people less likely to report toxic behavior when they witness or experience it.
Behavioral changes should come from within. By keeping things playful, people are more likely to be receptive to change.
Beware of reactance which is when people feel negatively to interventions because they sense that their autonomy is being restricted.
The challenge: Change people’s behavior in a way that it shifts from extrinsically motivated to intrinsically motivated.
In short… make people want to be good.
A big challenge when navigating interventions for human behavior is ensuring that the change is sustainable. Thus, we wanted to be mindful about pushing players to engage in “slow” or “system 2” thinking, by giving space for reflection. By engaging in laborious thinking, players can begin to deconstruct their impulse behaviors, and rebuild their online identity starting from their core values.
We want to ensure players are behaving well not just because they can get rewarded for it, but because they believe that they are good people at the core.
Extrinsic Motivation - doing something because of external motivation
Intrinsic Motivation - doing something because we find it intrinsically valuable, interesting, or enjoyable
How did we test our solution?
For the following weeks, we went through 3 rounds of iteration and testing to refine our design idea. For each round, we used a similar testing protocol: 1) walk through an overview of the new feature idea 2) pick their values 3) play a full match 4) endorse teammates 5) complete a follow up interview/survey.
-
Main Features:
Paper prototype in which participants place paper strips with their values at the bottom of their laptop screen
Players can choose to upvote or downvote teammates on each value
Key Takeaways:
Players did not think about values during the game
Endorsement screen is overwhelming
Players noted that some values are easier to get upvoted on than others and some values are too similar to each other
-
Main Features:
Improved list of values based on the Rokeach Value Theory
Short descriptions for each value to prevent definition ambiguity
Can only upvote teammates (or do nothing)
Key Takeaways:
Players still did not think about their values during the match
Participants associated the values to both their gameplay styles and their lifestye outside of the game
Players found themselves thinking about how they could better embody their values for the next game
-
Main Features:
Players are asked to write a short reflection for each value choice with a cool-down timer
Loading screens before each match display every player’s values
Values are displayed during the match right next to each player’s avatar
Key Takeaways:
Participants found themselves looking at their own and their teammates’ values at multiple points thorughout the match
Participants are skeptical about adoption within the League community and fear players will not take the reflections seriously
Participants are concerned that beginner players will have a hard time moving up their values because of poor game performance
Players can see their values at the bottom of the screen
Players pick 3 values that resonate with them most
Endorsement “screen” with both upvote and downvotes
Communication between players through the chat
What did people think?
To get a quick and general idea of the usability of the feature, we used a System Usability Scale (SUS) to get a baseline how people perceived their interactions with our final prototype. Further usability testing would be required to get a more comprehensive understanding. However, this provides a good starting point to figure out possible problem areas.
1. I think that I would like to use this system frequently.
4/5
2. I found the system unnecessary complex.
2.2/5
3. I think the system was easy to use.
4/5
4. I think that I would need the support of a technical person to be able to use this system.
1/5
5. I found the various functions in this system were well integrated.
4.2/5
6. I thought there was too much inconsistency in this system.
1.2/5
7. I would imagine that the most people would learn to use this system very quickly.
4.6/5
8. I found the system very cumbersome to use.
1.6/5
9. I felt very confident using the system.
4.2/5
10. I needed to learn a lot of things before I could get going with this system.
2/5
In conclusion: The system is generally understandable at face value, but there are certain complexities that might not be clearly communicated causing some confusion and insecurity for users. This is something we are planning on further exploring.
Where do we go from here?
Huge thanks to Prof. Kaufman and our wonderful team! <3
Coming out of this project, I have grown immensely as a researcher and designer. From leading a multidisciplinary team through conversations around shaping a project, to designing user testing protocols and rapidly prototyping, I now feel more equipped to tackle complex problem spaces.
As we continue our research with Prof. Kaufman, we plan on refining our solution, thinking about more moments in which self-reflection could be prompted. We will also be working with the Collaborative Against Hate to continue testing and developing our ideas. The goal is to conduct more longitudinal testing to explore the long term effects of our solution. Something we are curious about is the extent to which online environments can transform one’s sense of self, and whether improvements in online behavior can translate to improvements in offline behavior. These would be valuable insights not just within League of Legends, but hopefully for other online environments as well.
Currently, we have received IRB approval and funding from Carnegie Mellon’s College of Fine Arts Creative Inquiry Grant to continue this journey!