A long development cycle teaches you things. Some are about the specific game. Others are about how to make games in general. This piece collects the lessons that stuck.
Scope discipline matters more than ambition
Our early design documents were ambitious. We had five gods planned (we cut to four), seven progression layers (we cut to five), elaborate co-op systems at launch (we deferred). Every cut hurt. Every cut was right.
Ambition is cheap. Scope discipline is expensive. The discipline is what ships games.
Players surprise you
We designed the parchment system assuming most players would equip their best parchments before each run. Beta data showed players were enjoying the experimentation more than the optimization. People were running builds we thought of as suboptimal because they were fun. We adjusted accordingly.
The lesson: do not impose your idea of optimal play on players. Watch what they actually do.
Balance is iterative, not solvable
We thought we could balance the four gods system to perfect pick rates. We could not. Pick rates have political and cultural dynamics beyond pure mechanical balance. We learned to balance for "all four are viable" rather than "all four are equally chosen."
See our devlog on the four gods system for the pick rate story.
Community shapes everything
The Discord community has shaped Lost Tower more than any single design decision. Testers caught balance issues we missed. Players surfaced edge cases we never imagined. The community is not just a marketing channel; it is a development resource.
What we will carry forward
Scope discipline. Iterative balance. Watching players rather than guessing. Building community early. Saying no to features that don't fit.
These are the lessons that will carry into Klyppy and beyond.