Sky: COTL and Social Play

Social Play: "any activity performed near other players, which encourages simulating the thoughts and feelings of others" - John Hughes, thatgamecompany @GDC 2020
My Role
Project Lead & UX Designer
Tools/Skills Used:
Figma, User Flow, Affinity Diagram, Prototyping, User Survey
Duration
4 weeks (Feb 2021 - March 2021)
Project overview:
Open world MMO game Sky: Children of the Light emphasizes social interaction between players. The game’s Friends List feature, or "Friend Constellation,” is the main method for users to intentionally interact with another known user. As a veteran player myself, I wanted to conduct research to see how the Friend Constellation’s mechanics affected player experience.

Read final paper here:

See Final Prototype
Problem:
Lack of user flexibility makes it difficult to locate friends. This has lead to inefficient user workarounds.
Research aim:
1. Observe user behavior and specific pain points
2. Learn how the Friend Constellation affects social play
3. Identify user journey and uncover UX blockers
How are friends displayed?

Currently, the friend constellation:

  • Shows up to 72 friends across 5 screens
  • Has no feature to remove/search for friends
  • Holds 25 favorited friends on the first screen
  • Changes friend “star” positions periodically
  • Allows interaction with friends — can favorite, block, gift, and join friends
How do you interact with friends that aren’t shown?

Players who exceed 72 friends will still only see 72 friends at once. Those invisible friends still exist, but are unreachable in the moment.

The only alternative to meeting through luck is a player-found workaround: temporarily blocking friends to remove names and make space until the specific friend appears

Recruited 4 social, veteran users:

Search Criteria:

  • Behavior: plays weekly
  • Demographics: played for +1 month, has +72 added friends
  • Technology: has mobile device with Sky and Discord apps
Research participant demographics:

Survey results showed all participants log on +3 times weekly, add 1-5 new friends weekly, and have played the game for anywhere between 9-14 months.

Methods:

Summative Usability Testing

  • Qualitative insight
  • 25 - 45 min individual sessions
  • Self reported quantitative data

Each session started with 2 tasks:

  • Try to find a friend you don't see in your friend constellation at this moment
  • Please bring any friends removed in task 1 back to the friend constellation (if applicable)

Followed by a semi-structured interview:

Focused on previous behavior with making and finding friends as well as how users would want to find friends

And finished with a System Usability Scale Survey.

Analysis:

1. Affinity Diagram

Transcribed, grouped and coded user inputs & behavior

2. User Flow

Used user behaviors to detail every possible step and route of the current journey

3. System Usability Scale Survey

Exported to Google sheets and calculated manually

“The constellation is very cute, but it’s not super useful in terms of finding people” - User G

Critical finding: there is no clear method to locate a friend.

In fact, the constellation, though aesthetically pleasing, is disorganized and unintentionally causes users—especially those with larger friend lists—to lose contact against their wishes. User-found workarounds like favoriting, blocking, and adding emojis to friend names are not scalable and create unnecessary steps in the user journey

Evidence from usability testing:
  • User P blocked +80 friends in task 1 unaware that the favorited and emoji-labelled friend in question was on the main constellation from the start.
  • User G was notified in-game a friend was online (this only occurs for favorited friends), but could not find their name on the constellation to play with them.
  • User H was with a friend in-game, left to equip an item, but could not find and rejoin their friend through the constellation until they blocked numerous players.

"Invisible" friends tend to be those a user has not contacted in a while. However, if a user decides to interact with one, there's no guaranteed way to do so—the user flow shows how users can easily fall into a cumbersome cycle of searching and/or blocking others. Success isn't guaranteed for those experienced with the blocking tactic, either. As a result, the  average SUS score regarding users' experiences with the friend constellation is below what's desired. Each user's SUS score correlated with how much they struggled with Task 1.

Average SUS Score:

63.11%

Allow users to organize/customize the Friend Constellation.

Users proposed the constellation could show all their friends or have an all encompassing list paired with filter functions. Some users wanted to view friends by most recently online to either see who's online or to block inactive players to make space--I interpreted this to mean players want organization. However, if the goal is to address this study's issues, a search feature would have prevented any of the user anecdotes above where users could not find a specific friend they wanted to contact. See my prototype below.

Further research should be conducted on a "remove friend" feature and organizational tools for the friend constellation.

Prototype:

I created a new magnifying glass icon for the friend constellation in the top right corner. Clicking it opens a search bar that appears over the same background as the friend constellation.

As the user types, the screen displays friend names that start with what's typed. If there is a large amount of friends that begin with the same letters, you can tap the space outside the search bar and browse among the 5 screens. Beyond this prototype, there should be a loading indicator, and clicking on a friend should open up the interaction menu like the original constellation.

Now imagine this scenario: my old friend Mona messaged me on Discord asking to catch up through Sky, but because I have a large friends list and haven't played with her in months, I don't  see her name. However, with my prototype, we can easily find each other. My prototype acts as a simple solution to a specific, but frustrating problem while staying consistent with Sky's current design and style.

Future questions to explore:
  • What percentage of active users have over 72 friends?
  • What percentage users want to be able to remove a friend?
  • Why are users blocking players and does that align with the original purpose behind the blocking feature?