Project 03 · Game
Open-world farming and life sim, validated with 50,000 beta players.
PlaceholderFarm RPG — key art16:10
Project Resume
Progression is layered on different clocks. Something finishes within a session, something finishes overnight, and something takes a week. A player who has ten minutes gets a complete loop; a player who has an hour gets a project. The interface exists to answer one question fast: what is ready right now.
PlaceholderProblem 116:10
A large share of new players left during the first session.
PlaceholderProblem 216:10
Players stalled at a predictable point in the mid game.
PlaceholderResults & Learnings16:10
Media
PlaceholderFarm RPG — key art16:10
Farm RPG — key art
Open-world farming and life sim.
PlaceholderFarm RPG — key art16:10
PlaceholderProgression UI16:10
Roles
What I did
Full Case Study
Design Brief
A farming sim has no fail state and no finish line, which means nothing forces a player to come back. The design problem was to give an open-ended game a reason to be opened tomorrow, without turning it into a chore list.
Constraints
Overview
Progression is layered on different clocks. Something finishes within a session, something finishes overnight, and something takes a week. A player who has ten minutes gets a complete loop; a player who has an hour gets a project. The interface exists to answer one question fast: what is ready right now.
Key Design Choices
PlaceholderProgression UI4:3
Problem
Problem
A large share of new players left during the first session.
Cause
Onboarding introduced every system up front, so nothing was completed before the player had to absorb the next thing.
Solution
Cut onboarding down to a single loop taken all the way to a reward, and deferred everything else to the moment it became relevant. The player leaves the first session having finished something.
Result
Problem
Problem
Players stalled at a predictable point in the mid game.
Cause
Content was paced against an assumed session length that the data did not support.
Solution
Re-paced the progression around the session lengths players actually had, and made long projects resumable rather than restartable.
Result
Learnings
Fifty thousand players is a different job from a playtest. At that scale you stop debating whether a decision is good and start reading whether it worked, and the two answers are not always the same.
Retention in a game with no fail state is entirely a pacing problem. Nothing in this project was fixed by adding content — the fixes were all about when content became available and whether a session could be finished.
“Games are a way to turn ideas into feelings.”— Unir Reis
Next project
Summoned Guardians
