A safer space for kind gamers
Overview
Kind Words is a video game where players send supportive letters to strangers.
There has a been recent influx of letters on suicide that is affecting the regular playerbase, reflected in the game reviews.
The goal is to protect players from these alarming messages and guide the ones in crisis to find better help resources
Role
UX Designer
User Research, Wireframing, Prototype and Testing
2024
“There were letters with so much pain I felt incapable of helping”
01. Research
Compared to other communities, the game’s security filters were low
Popular online forums like Reddit deal with inappropriate messages with word filters and volunteer moderators, stopping them from being shown to others. Reporting comes as a tertiary solution.
Kind Words, instead, relies first on user reports to moderate the community.
Word filters fall on the weaker side, and moderators are few, but most in the same timezone.
02. Analysis Insights
Reporting letters doesn’t fix gamers being exposed to them
- People posting alarming messages online is not good for anyone.
- Regular players need to be protected. Community moderation and dealing with SOS is not on them.
- All the game can do is direct people in crisis to proper resources, but it can be done better.
03. Solutions
Improving security and helplines
1. Stronger Global Filter to keep Suicidal Messages under control
2. Real message from active moderator sent to the person in distress, directing them to the helpline resources.
3. Custom word filters for each player to avoid triggers, for example, eating disorder trigger words.
4. Improving helplines by country (IP Location Sensitive) and sort them by nature of the problem
5. Adding an emergency button in the helplines and interface
6. Adding a sad face / happy face voting system
Sad faces = message is less shown around the game
04. Prototype Test
Gamers loved the filters, but didn’t understand the voting system
The good: all players were very happy about the filter and customised filter options! they used adjectives like ‘‘helpful’’ and ‘‘great’’. Additionally they appreciated the improved helpline resources.
The bad: they found the happy/sad voting system hard to understand: ‘‘happy or sad about what?’’ ‘‘will the other person know I don’t like the message?’’ the non-solution was ambiguous and frankly, an overkill, so It had to go. Also, they didn’t have much faith in the correct-use message for new players, which is understandable.
Final Prototype
05. Conclusion
I improved my research approach and talked to the developers
Conclusion: Many assumptions at the beginning. I was struggling to separate myself from the research (I’m sensitive, I’m a gamer).
How did I fix it? I found an assumptions checklist by Erika Hall that now will become a staple in my design process.
Next Steps: I’m currently talking to the developers. For now, they want to leave the game as it is. They are very open, and this project’s solutions might be integrated in the future, specially when a chat version of the game is coming out soon.