RELEASE Devlog: From Prototype to Playable

Hey everyone! :)

DungeonMart is finally releasing on Itch.io. It's been wild 2 weeks since my last devlog, and honestly? The game I'm shipping today is barely recognizable from what I showed you back then. :)

Let me walk you through what changed, what I learned, and why sometimes you need to reconsider your initial direction entirely.

The Big Pivot: Three Changes That Might Have Redefined Everything 

Remember when I said I was wrestling with "volume vs. margin" gameplay? Looking back, I think I might have been asking the wrong question.

After playtesting the prototype, three things seemed to become clear:

1. Categories > Sizes 

My original system had items sorted by physical size (Small/Medium/Large). It seemed clean and logical, but I started to wonder if it was really the right fit for this game.

The issue, I think, was that players struggled to strategize meaningfully. "Do I stock more medium items?" doesn't feel like an interesting decision to make. But "Do I focus on Weapons or Potions this week?" seemed to engage players more.

The fix: I replaced the size system with 6 item categories (Weapons, Shields, Potions, Armor, Traps, Magic). Each category now has its own unlock progression through objectives. I believe this gives players more meaningful choices about which market segments to pursue.

Implementation: This was roughly a 2-day refactor that touched basically everything—shelf systems, restock UI, customer demand logic. Claude Code helped scaffold the UI changes while I rewired the core systems. I think the hardest part wasn't the code itself; it was updating 35 items in the database and creating matching shelf prefabs.

2. Mouse Over WASD 

The keyboard movement felt... off to me. Like, "why am I pretending this is an action game?" kind of off :/. You're running a shop. You're clicking things. I kept wondering why I was making players walk around with WASD when they might just want to point at stuff and interact.

The fix: I tried a full transition to point-and-click instead. Click shelves to restock. Click delivery boxes. Click HUD buttons. I added hover for visual feedback (pink outline pulse—I think it looks pretty good :))

Implementation: I rewrote the player controller to use raycasts + NavMesh pathfinding. Added hover detection with emission-based materials for feedback. The whole thing took maybe a day, partly because the game logic was decoupled from input—having a component-based architecture really paid off here.

3. Objective-Based Progression 

My tier system was just "earn money → unlock stuff." It worked, but I wasn't sure it felt very engaging.

The fix: I moved toward parallel objective tracking across 5 types (Revenue, Customers Served, Items Sold, Days Survived, Hybrid goals). Each objective unlocks specific rewards—new item categories, shop segments, upgrades. I think this makes progression feel more earned rather than just grinded.

Example objectives:

"Sell 50 Weapons to unlock Shields category" "Serve 100 customers to unlock shop expansion" "Survive 10 days with profit to unlock Tier 3 items"

Implementation: This was probably the biggest lift—ObjectiveManager.cs, ObjectiveDataSO, filtering UI, progress bars, completion badges. I used ScriptableObjects for data-driven design (all objectives live in the inspector, which makes balancing easier). Claude Code helped with the UI scaffolding while I built the tracking logic.

What Seems to Be Working 

After shipping ~40 features, here's what I'm fairly happy with so far:

The Systems Architecture: Everything runs on events and ScriptableObjects. I believe this approach makes it relatively easy to add new content—want to add a new objective? Create an SO, link it, done. Want to rebalance 35 items? Edit a CSV, hit import, done. This flexibility definitely helped me during crunch.

CSV Item Importer: Building this tool in Week 2 might have been the smartest decision I made :). I've rebalanced the economy 11 times. Without this tool, I think I'd still be manually editing item #8.

Phase-Based Pacing: Morning (restock) → Business (customers) → End of Day (orders/upgrades) seems to create a natural rhythm. I compressed the business phase to 90-120 seconds with fast customer spawns—it feels better to me than the slower 5-minute phases I started with.

Customer Demand System: I show trending categories with demand bubbles. I think this gives players information to make more strategic restocking choices. It's a simple visual, but I believe it has a noticeable impact.

What Might Need Work :/

Let's be honest about the rough edges:

Restocking UX Feels Clunky: You click a shelf, get a menu, pick items, confirm. But you can't go back if you misclick. You can't see "where" you're restocking until you're in the menu. I think this needs a redesign—maybe drag-and-drop? Maybe preview mode? I'm still working through what might work better.

Balancing is Uncertain: I've tried tuning for 15-20 minute completion, but I'm one person playtesting my own game. I'm not confident the difficulty curve is right. That's partly what release feedback is for—I need outside perspective.

Some Upgrades Might Feel Underwhelming: Express Checkout (25% faster transactions) and Bulk Ordering (5x items, 10% discount) work mechanically, but I'm not entirely sure they feel impactful yet. Feedback will help here.

How I Approached Building This 

Iteratively, rather than through heavy planning. I didn't design the category system in a document—I built the size system, played it, felt unsure about it, and replaced it. Same with mouse controls. Same with objectives.

Data-driven everything. ScriptableObjects for items, objectives, upgrades, customer types. I think this approach let me iterate on balance without touching code as much. CSV import for the item database. I tried to make everything that could be designer-facing into something designer-facing (even though I'm the only designer).

Cut features fast. I removed a whole "License Upgrade" system because I felt objectives served that purpose better. Cut an "Extended Hours" concept. If something didn't seem to serve the core loop, I tried to axe it.

What's Next 

The game is releasing today on Itch.io as a free browser game (WebGL). 20-30 minute playtime. Six item categories. Objective-driven progression. Trending demand system. And yes, you're arming the dungeon monsters against the heroes.

I need feedback—I'd really appreciate honest thoughts from outside my own perspective :). Does the economy feel right? Is restocking engaging or tedious? Do the objectives guide you or confuse you?

This is v1.0 of a prototype that might become something bigger, or it might teach me what not to do next time. Either way, I've learned a ton building it.

Play it. Break it. Tell me what you think.