Lessons for my own (Godot) games

What's worth stealing, why, and what to watch out for.

Steal: the dual-life loop framing

A single character living two distinct mini-games that consume each other's outputs is a rare design shape and almost always interesting. Action-roguelite + shopkeeper is the Moonlighter twist — but the pattern generalizes:

• Combat by day + farming by night (Stardew).
• Dungeon by day + cooking sim at night (untouched indie space).
• Battle royale + base-builder hybrid.

What matters: the same items / resources flow through both, and each half evaluates them by a different rubric. The tension between "what's good in mini-game A" and "what's good in mini-game B" is where the design lives.

Steal: loot value as a placement problem

The backpack puzzle is the most original mechanic in the game. Almost no other indie has tried "where you put loot determines what it sells for" with this much commitment.

Generalizes naturally:

• Spell loadouts where adjacency triggers combos.
• Crew assignments on a ship (Sea of Thieves-ish).
• Garden plot layouts where adjacent crops boost each other.
• Inventory tetris where items have synergy patterns, not just shape.

The lesson: don't make loot just have a value — make it have a value equation that the player solves.

Steal: the Hades-style map even in non-roguelikes

Moonlighter 2's map is one of the cleanest "graft a Hades-style DAG onto a non-Hades game" examples around. The pattern works because:

• Players see the run shape before committing.
• Path choices feel meaningful even when individual encounters are simple.
• One UI screen unlocks days of design discussions about node-type balance.

This is portable to anything with sequential encounters. Not just roguelites. Tower defense waves, narrative branching games, RPG dungeons — all candidates.

Steal: layered multipliers as the price equation

The pricing equation base × quality × popularity × showcase + tip is a clear template for player-facing depth via multipliers. Each multiplier has a different sourcing path (loot → backpack → calendar → furniture → live shop), so each gives the player a different lever to pull.

If you're designing economy systems, layered multipliers from independent sources is more interesting than one big number plus modifiers — because each multiplier is a different mini-decision.

Steal: cite your inspirations

Digital Sun being open about Hades inspiration helped reception. Players appreciate knowing what the dev was reading; reviewers respond better to "Hades-inspired with an original twist" than to "we totally invented this." Honesty is cheaper than mystery here.

Cautionary lesson: re-platforming a sequel costs charm

Moonlighter 1 → 2 went from 2D pixel art to 3D isometric. The mechanical depth went up; the charm went down. Reviewers consistently flag this. Some original-game fans aren't coming back.

For my own future sequels: if you change visuals alongside mechanics, audiences perceive both as a single shift. Plan for the cost.

Cautionary lesson: don't replace charm-mechanics without a great replacement

The original's price-discovery minigame had charm. The sequel's multi-multiplier math has depth. They are not the same thing. If you replace a charm-mechanic with a math-mechanic, you'd better make the math itself charming — Spire does this with "risk vs reward fractally"; Moonlighter 2's math is currently not labeled with a similar identity, so it reads as opaque to new players.

Cautionary lesson: in EA, the exchange rates are everything

Multiple reviewers note that the ratios between systems (gold per relic, perk per encounter, blacksmith cost) feel off in early access. Even when each individual system is fine, the exchange rates between them is what makes the loop feel rewarding or grindy. Ship those tuned, or expect EA reviews to reflect it.

Cautionary lesson: variety is good, but milestones matter

The original Moonlighter had clean milestone progression — each new dungeon was a real escalation. The sequel has more options, less clear milestones. Players need to feel "I just unlocked the next thing." If your meta-progression is mostly variety, you still need some mile-markers, or it feels like grinding for nothing.