Crafting is one of the most repeated and most often misapplied mechanics in indie game design. A well-designed crafting system can add hours of value; a poorly designed one turns a game into an inventory simulator. The difference is not graphical fidelity but design discipline: which recipe, which material, at what tempo, for what reward.
At Althera Games, while designing the crafting system for Potion Rise Simulator from scratch, we first took apart the best examples in the genre, including Recettear, Potion Craft, Stardew Valley, and Moonlighter. In this article we share a practical design map that runs from the anatomy of the crafting loop through recipe-discovery philosophies, the material supply chain, and the specific decisions we made for Potion Rise Simulator. Thematically, our simulation economy design article is the economic cousin of this piece.
The Anatomy of a Crafting Loop
A crafting system is not a single mechanic; it is a loop. At its simplest, the loop has four phases: planning, sourcing, production, reward. The player chooses what to make, gathers materials, applies the recipe, then receives the result (money, progression, story). If any of those four phases is weak, the loop breaks; the player gets bored sourcing, fatigued producing, or disappointed at reward.
In a healthy loop each phase has its own tempo. Planning should be fast and information-dense; if the player can't decide in five seconds, your UI is wrong. Sourcing should be moderately paced and physical; the player should move and observe. Production should be rhythmic and emotional; the player should enjoy the process. Reward should be fast and clear; the player should know they won. Tuning those four tempos is the actual work of crafting design.
A practical check: a single full pass through your loop should run between 3 and 7 minutes. Loops shorter than 3 minutes feel like clickers, not mechanics; longer than 7 fatigue the player on rotation. You can observe this band across games from Stardew Valley's harvest cycle to Recettear's single-day commerce. Loop length sets the rhythm of your game.
Recipe Discovery: Open vs Hidden
The biggest philosophical decision in crafting design is how the player learns a recipe. There are two main approaches: open discovery and hidden discovery.
In open discovery, the player sees the formula the moment a recipe is unlocked. Most of Stardew Valley follows this model: when you learn a recipe, the ingredients and result are presented openly. This produces planning pleasure; the player calculates which materials to gather, how much to produce, which recipe is profitable. For strategically inclined players, this is a satisfying feel.
In hidden discovery, the player experiments to find the recipe. Old Minecraft crafting tables, Don't Starve's combination logic, Astroneer's material experiments fall in this category. This produces discovery pleasure; the player gets surprise and "aha!" moments. The risk is that experimentation cost can easily turn punishing; if the player loses hours of materials on a wrong combination, discovery turns to frustration.
A practical middle path is hybrid discovery. Announce that a recipe exists (for example, "Health Potion recipe unlocked"), but let the player figure out the exact ratios or optional components experimentally. This is the core wit of Potion Craft: navigating the liquid map ingredients trace is both discovered and experimental. In Potion Rise Simulator we use the hybrid model: recipes are unlocked through in-game booklets, but every recipe's optional bonus components (a rare herb that produces a "premium" variant) are left to the player's experimentation.
The Material Supply Chain
The unsung hero of crafting is the supply chain the player doesn't quite notice. The recipe itself depends on where its required materials come from. If every material comes from a single source, the loop stays shallow; if materials come from too many sources, the loop becomes tiring. The ideal supply chain has 3 to 5 tiers, and each tier presents a distinct gameplay activity.
For Potion Rise Simulator the chain looks like this: seed (purchased from NPCs or found while exploring) → plant (grown in the player's garden, requires water and sun) → extract (drawn from plants in the workshop through methods like distillation or fermentation) → potion (combined extracts using a recipe, the heart of the loop). Four tiers give the player four distinct mental activities: bargaining, patience, experimentation, payoff. That variety enriches the loop on its own.
The most common mistake in supply-chain design is an uneven time gap between tiers. If seed-to-plant takes 10 seconds and plant-to-extract takes 10 minutes, the player gets stuck at the chain's middle. Stardew Valley solves this through evenly paced tiers and parallelizable tasks. In Potion Rise Simulator we set a target average of 30 to 90 seconds per tier; the player never waits at any tier and never blasts through one instantly either.
A practical technique: backward design is the safest approach. Start with the final product (say "Healing Potion"), then work backward: what do you need to make this? What do you need to make that? What do you need to make that? Stop after 3 to 5 tiers. This tends to prune unnecessary branches that forward design would otherwise grow. Our UE5 indie development article covers how to implement this with data tables in the engine.
The Risk-Reward Balance
At its core, a crafting game is a risk-reward system. The player invests time and materials, and gets value back. If that equation is unbalanced the game becomes either too easy (no risk) or punishing (no reward worth the risk). In a well-designed crafting system the player makes a small bet on every choice.
Recettear's bargaining is a clean example. The customer asks for a price; if you say a higher number, the customer may walk or initiate a haggle. Quote too low and you lose money; quote too high and you lose the sale. Every interaction presents a micro-choice of risk. That simple balance keeps the game interesting for hours.
For Potion Rise Simulator the risk-reward design runs at three levels. Seed level: which plant will you grow, in which season, for which NPC tier? Recipe level: would you risk burning that hard-earned rare ingredient on an experimental recipe? Customer level: a basic potion sells for 50 silver to a peasant, a premium potion for 500 silver to an aristocrat, but if the aristocrat refuses you, that premium potion sits in your inventory. The three levels combine, and players run different strategic experiments per session.
The depth of crafting comes not from the count of materials but from the player remembering why they made a certain choice. The recipes that stick aren't the hardest to compute, they're the ones with stories.
Comparing Recettear, Potion Craft, Moonlighter
Three games in the genre solve the same problem with different design emphases. Studying them side by side clarifies what your own crafting system should pick.
Recettear (2007): as a shop simulator, it pushes crafting to a secondary plane; the main mechanic is customer interaction and price negotiation. Crafting here is not the foreground but a way of stocking the shop. The player descends into the dungeon for three days, collects materials, then returns to town and runs sales in the shop. Recettear's genius is that crafting and trade aren't separate systems; they are the two halves of the same money loop.
Potion Craft (2022): re-imagines crafting as a map-based journey. The liquid moves across a map as ingredients are added, and reaching a target zone determines the desired potion. The mini-game is so satisfying that it carries the brand of the game. Lesson from Potion Craft: think of crafting as a space rather than a list, so the player feels something physical through the process.
Stardew Valley (2016): assembles crafting as a wide ecosystem. Hundreds of recipes, thousands of materials, all bound to character relationships, the seasonal system, the farming system, and the festival calendar. Stardew's genius is to position crafting not as a standalone mechanic but as a part of life. GMTK's crafting-design videos walk through this approach in detail.
Moonlighter (2018): modernizes the Recettear formula and compresses an action dungeon and shop management into a single loop. Crafting acts as a middle layer where dungeon rewards are evaluated. Lesson from Moonlighter: if crafting alone is not enough, chain it with a primary mechanic (combat, trade, social) instead of trying to inflate it.
A practical conviction: don't pick one of these four games for your crafting study, play at least three in depth. Sketch out their loops on paper. Then sketch out your own. The overlap is probably genre standard; the divergence is probably your game's character.
Connecting Crafting to Narrative
A crafting system unhooked from narrative is a spreadsheet. The most powerful crafting systems frame the mechanic as part of a story. How? Present the crafting process from a character's perspective, in the language of a culture, inside the context of a story.
For Potion Rise Simulator, our protagonist is an alchemist taking over a small village shop. Every recipe of his comes from a city myth, an outbreak, an aristocratic family's habit. When the player makes a Healing Potion, it isn't just an HP-restoring item; it's a kindness for an old woman's grandson in the village. That context doesn't change the function of the mechanic, but it doubles the emotional weight.
A practical technique: add narrative flavor text to recipe cards. Beside the ingredients and the result, place a single line or two of story note. "This potion was first brewed by Eldred, a long-dead mage, who used a rare herb from his garden." That small addition lifts a recipe from a memorized formula to an inheritance. The Lost Garden blog covers the long theoretical history of this approach in game design.
Potion Rise Simulator's Approach
Now let's get concrete on how we apply these principles in Potion Rise Simulator. Our system is a four-tier supply chain, hybrid recipe discovery, three-level risk-reward, and narrative-bound recipes.
Supply chain: seed (15+ varieties, bought from NPCs or found while exploring) → plant (grown in the player's garden, yield varies by season and soil) → extract (drawn in the workshop through distillation, fermentation, pressing) → potion (extracts combined per recipe). Each tier averages 30 to 90 seconds and can run in parallel.
Recipe discovery: 5 base recipes are open at game start; the player unlocks 60+ recipes over the run through NPCs, found books, and story events. Every recipe has an optional "premium variant" the player must discover experimentally; the game won't tell you, but in-game NPC dialogue drops hints.
NPC tier system: Peasant (basic potion, 30-80 silver), Trader (mid-tier, 100-250 silver), Aristocrat (premium, 300-800 silver), Court (legendary, 1000+ silver). Each tier has its own aesthetic, dialogue, and patience profile. If an aristocrat refuses you three times, you can't sell to that tier for several days; a peasant will accept even a basic potion. This hierarchy forces the player into strategic tiering.
Dialogue choices: during a customer interaction, the player chooses one of three dialogue options (respectful, friendly, insistent). The choice affects the sale's success rate and long-term relationship. This is a more narrative-leaning version of Recettear's bargain. Customer interaction isn't just an item swap; it is a small story moment in the game.
If you're considering whether to build these systems with C++ or Blueprint, our Blueprint vs C++ article covers which logic suits which approach; data-table-driven systems like crafting tend to be quicker in Blueprint, while performance-critical concerns like gameplay tags get cleaner C++ implementations.
Frequently Asked Questions
How many recipes are ideal in a crafting game?
The ideal number depends on the genre; for a potion-shop simulation, 30 to 60 recipes is a comfortable scope for most players' first playthrough. Wider simulations like Stardew Valley can hold 100+ recipes, but most are optional and long-tail. The actual decision metric is not the recipe count but the meaningful variation between recipes. If two recipes use the same ingredients and yield equivalent outcomes, merge them; if a single recipe satisfies two distinct player needs, drop one of them.
How much RNG should a crafting mechanic carry?
Pure RNG-driven crafting tires most players; fully deterministic crafting turns the player into a spreadsheet operator. The ideal balance is controlled variation: the player gets the base result reliably, while optional components introduce yield, quality, or trait bonuses. Recettear's bargaining RNG and Stardew Valley's grown-crop quality randomization are examples of this model. The player knows the floor and enjoys the upside.
When is automated crafting risky as a feature?
Automation can be necessary in late game where the player's time matters more than the mechanic, but added too early it hollows out the central loop. If your game is designed around 'making potions' and you let the player automate that mid-game, all that's left is counting money. The fix is to position automation as a natural late-game reward, after the player is satiated on manual crafting.
Should I add a mini-game to crafting?
A mini-game gives crafting a physical feel and lifts the player out of being a passive clicker. Potion Craft's map-based liquid mixing turns the mechanic into something you can't perform on autopilot, and even carries the brand identity of the game. That said, when a mini-game becomes an obstacle the player has to push through every time, it tires them. The core question is whether the mini-game produces a 'feeling' or a 'friction'. The first is a success, the second is a tax.
Is using crafting as monetization (a paywall) a good call?
In a premium indie game, using crafting as a paywall is risky; the player expects a full experience for a single purchase, and gating extra craftables behind additional payments damages trust. That said, crafting maps naturally onto content expansions (new recipes, new ingredients, new locales) for DLC. In a free-to-play context, crafting paywalls can work, but the designer must hold a fair balance between 'wait' and 'pay'. Aggressive pay-to-skip burns out the community quickly.
Conclusion: Designing Crafting as a Story
Crafting is not a spreadsheet; it is a way of living. A well-designed crafting game turns the player into an alchemist, a tailor, a cook. That transformation doesn't come from mechanic design alone; it comes from positioning the mechanic inside a character and a story. Recettear's shopkeeper, Potion Craft's alchemist, Stardew Valley's farmer; each gives a crafting system a personality.
The system we designed for Potion Rise Simulator places the lessons of the genre's best games inside the life of one village alchemist. If you are working on your own crafting project, our advice is this: prototype the loop on paper for three minutes first, then add the narrative context. If the mechanic isn't fun on paper, it won't be in engine. You can find a Potion Rise Simulator crafting demo through our games page.