Things are breaking down

A late-stage game-dev sprint is shifting from “make it work” polish toward concrete balance and player decision design, with the team aiming to have a fully playable build by Monday. The most consequential change is a simplification to how tower upgrades and enhancement cards scale: enhancements are being treated as equivalent to tower upgrades of the same rarity. That means a “common” speed boost (e.g., +0.25 attack speed) won’t create runaway balance problems across different tower contexts—like a catapult designed for slow fire—because the system becomes easier to reason about: a legendary enhancement effectively mirrors gaining one level with a legendary tower upgrade.

Alongside that systems rethink, the team is actively debating how much numerical feedback to show players. One camp prefers displaying numbers because the audience likes seeing “number go up,” while the other argues for a cleaner interface without raw stats to better communicate the game’s feel. The practical concern is that removing numbers can make it harder to judge how much certain changes actually improve performance, so the team plans to iterate on UI and balancing while keeping some quantitative signals in the loop.

The focus then narrows to level pacing and early-game clarity—especially how many waves the first level should contain, how hard it should be, and what rewards should look like after the player’s deck is partially depleted. With a limited starting hand (roughly six to eight cards, including golden harvest, enhancements, and towers), the group lands on the idea that the first level should be easy to pass, allowing some enemies to get through while costing the player only a bit of health. A specific target emerges: three to four waves for the first level, with tower placement rules designed to encourage meaningful early choices—such as making archer towers the default “one” and other towers “two,” so players can’t just drop a single option and call it done.

A second major design argument centers on when players make decisions. Some want enhancements in the opening hand, while others worry that too many choices at the start overwhelm new players. The compromise leans toward reducing decision load on wave one: introduce a clean, understandable setup first (possibly with a pre-game choice like “more health” or a riskier rare card), then let the first real strategic decisions arrive after the initial wave. The team also discusses adding a “chest” or random reward before play to create a single early fork without turning the opening into a spreadsheet of options.

Finally, the sprint includes operational realities: UI and animation systems are being ripped out and rebuilt, memory leaks are being fixed (event listeners weren’t being freed, causing enemy references to linger and memory to climb during levels), and the team is trying to stop constant flavor changes so it can ship content and mechanics. Even internal “day crew vs night crew” friction is framed as a productivity problem—resolved through alignment meetings—so the group can lock the end-state, build levels, balance waves, and keep the game fun rather than endlessly re-engineering it.