Virtual economy modeling for MMOs — currency systems, sink/faucet balance, reward schedules, inflation control, auction houses, and exploit prevention. Load when designing game economies, virtual currencies, loot systems, or player trading, or when the user mentions "game economy", "virtual currency", "sink and faucet", "inflation", "auction house", "loot table", "reward schedule", "RMT", "gold seller", or "player trading".
Design and balance virtual economies that sustain long-term player engagement without collapsing into hyperinflation, deflation, or real-money-trading dominance. This skill covers currency architecture, sink/faucet modeling, reward scheduling, auction house design, and economy telemetry. For server-side persistence backing the economy, load gamedev-server-architecture. For multiplayer transaction networking, load gamedev-multiplayer.
Every game economy starts with currency design. Get this wrong and no amount of tuning fixes the downstream problems.
Single-currency systems (one gold coin for everything) are simple but create a single point of failure -- if gold inflates, the entire economy breaks. Multi-currency systems isolate failure domains: PvE gold inflation does not affect PvP honor points.
Most successful MMOs use 3-5 currencies:
| Currency Type | Purpose | Tradeable |
|---|
| Example |
|---|
| Primary (gold) | General commerce, NPC vendors, auction house | Yes | WoW Gold, FFXIV Gil |
| Bound (tokens) | Progression gating, prevents buying power | No | Valor Points, Tomestones |
| Premium (gems) | Real-money store, cosmetics, convenience | Limited | GW2 Gems, Lost Ark Crystals |
| Seasonal | Time-limited engagement loop | No | Battle pass tokens, event currency |
| Social | Guild/group activities | No | Guild marks, raid tokens |
Bound (soulbound) currency cannot be traded between players. This is the primary tool for preventing real-money trading from undermining progression. Rule of thumb: any currency that gates player power should be bound. Any currency used for cosmetics or convenience can be tradeable.
Premium currency bridges real money and the in-game economy. Critical design rules:
| Game Model | Primary | Bound | Premium | Seasonal | Notes |
|---|---|---|---|---|---|
| F2P MMO | Gold (tradeable) | 2-3 token types | Gems/Crystals | Event tokens | Premium-to-gold exchange is critical |
| Subscription MMO | Gold (tradeable) | 1-2 token types | Optional cosmetic | Event tokens | Sub fee acts as implicit gold sink |
| B2P (buy-to-play) | Gold (tradeable) | 1-2 token types | Cosmetic-only | Expansion currency | No pressure to monetize aggressively |
| Mobile gacha | Soft currency (gold) | Hard currency (gems) | Paid gems | Banner currency | Dual-currency is the genre standard |
| Survival/sandbox | Barter (no currency) | N/A | Cosmetic-only | N/A | Player-driven economy, minimal NPC sinks |
The fundamental tool for economy balance is the sink/faucet spreadsheet. Every source of currency entering the economy (faucet) and every drain removing currency (sink) must be enumerated, quantified, and balanced.
List every way currency enters the economy, with per-player-per-hour generation rates:
| Faucet | Currency | Rate (per hour) | Gated By | Notes |
|---|---|---|---|---|
| Quest rewards | Gold | 500-2000 | One-time | Front-loaded, drops off at cap |
| Mob drops | Gold | 200-800 | Kill speed | Primary steady-state faucet |
| Daily login | Gold | 100-500 | Calendar | Flat injection, not skill-gated |
| Dungeon completion | Gold + Tokens | 1000-3000 | Weekly lockout | Major burst faucet |
| Gathering/crafting | Gold (indirect) | 300-1000 | Skill level | Creates items, not currency directly |
| Auction house sales | Gold (transfer) | Variable | Market | Not a faucet -- transfers between players |
| Vendor trash | Gold | 100-400 | Inventory | NPC buys items for fixed prices |
Critical distinction: player-to-player trading (auction house) is not a faucet. It moves existing currency between players. Only NPC payments and system rewards create new currency.
| Sink | Currency | Rate (per hour) | Participation | Notes |
|---|---|---|---|---|
| Repair costs | Gold | 100-500 | Mandatory | Scales with gear level |
| Consumables (potions, food) | Gold | 200-600 | Semi-mandatory | Raiding requires flasks/food |
| Auction house tax | Gold | 5-15% of sales | Voluntary | Primary endgame sink |
| Fast travel | Gold | 50-200 | Voluntary | Convenience sink |
| Crafting material costs | Gold | Variable | Voluntary | NPC-sold reagents |
| Cosmetic vendors | Gold | One-time | Voluntary | Mounts, transmogs, housing |
| Respec/reroll | Gold | One-time | Voluntary | Character customization |
| Guild costs | Gold | Weekly | Social | Bank tabs, perks |
| Durability decay | Gold | Passive | Mandatory | Items degrade, require replacement |
Target a steady-state ratio where total faucet output slightly exceeds total sink consumption at the median player. A 1.05-1.15 faucet/sink ratio means players slowly accumulate wealth, which feels rewarding. Below 1.0 creates deflation (players feel poor). Above 1.3 creates rapid inflation.
Daily Gold Generation (median player):
Quest rewards: 1,500 (assumes 1hr questing)
Mob drops: 600
Dungeon: 2,000 (one daily dungeon)
Vendor trash: 300
Daily login: 200
─────────────────────────
Total In: 4,600
Daily Gold Consumption (median player):
Repairs: 400
Consumables: 500
AH tax (est.): 300
Fast travel: 150
Crafting reagents: 200
─────────────────────────
Total Out: 1,550
Ratio: 4,600 / 1,550 = 2.97 ← DANGER: heavy inflation
A ratio of 2.97 means the economy will inflate rapidly. Solutions: increase sink rates (higher repair costs, higher AH tax), decrease faucet rates (lower quest rewards), or add new sinks (see Inflation Control below).
Define how long a median player should take to earn key purchases. Work backward from these targets to set faucet rates:
| Item Category | Target Time-to-Earn | Price (Gold) |
|---|---|---|
| Basic mount | 2-3 days | 5,000 |
| Rare mount | 2-3 weeks | 50,000 |
| Endgame consumables (daily) | 30 min | 500 |
| Best craftable gear | 1-2 weeks | 30,000 |
| Cosmetic set | 1 week | 15,000 |
| Housing (basic) | 1 month | 200,000 |
If your faucet/sink math puts a basic mount at 2 weeks instead of 2-3 days, faucet rates are too low or prices are too high.
Inflation is the default state of most game economies. Players generate currency faster than they spend it, prices rise, new players cannot afford anything, and the economy stratifies. Controlling inflation is the primary ongoing challenge.
The direct approach: remove currency from the economy.
| Sink Type | Effectiveness | Player Perception | Notes |
|---|---|---|---|
| Repair costs | Medium | Negative (feels punitive) | Scale with content difficulty |
| AH listing fees | High | Neutral (cost of doing business) | 5% list + 5% sale is standard |
| Consumables | Medium | Neutral (necessary expense) | Must be meaningful, not just gold tax |
| Cosmetics | High | Positive (aspirational purchases) | Gold-sink mounts, housing, transmog |
| Limited-time offers | Very High | Positive (FOMO, but be careful) | Rotating vendor with expensive items |
| Crafting failures | Medium | Negative if too punitive | Enhancement/upgrade systems that consume gold on failure |
| Progressive repair | High | Mixed | Repair cost scales with server-wide gold supply |
Limit how fast currency enters the economy:
Apply DR to repeated activities within a time window:
First 10 dungeon runs/week: 100% gold reward
Runs 11-20: 50% gold reward
Runs 21-30: 25% gold reward
Runs 31+: 10% gold reward
This preserves the experience for moderate players while throttling power-farmers.
Items cost more as server-wide wealth grows. Track total gold in the economy (see Telemetry section). When gold supply exceeds targets, NPC vendor prices increase proportionally. This creates an automatic stabilizer -- as inflation rises, sinks deepen.
| Symptom | Primary Fix | Secondary Fix |
|---|---|---|
| Steady inflation (2-5%/week) | Increase AH tax, add cosmetic sinks | Reduce top-end faucets |
| Rapid inflation (>10%/week) | Emergency: cap daily earnings, add gold-sink event | Find and fix the broken faucet |
| Deflation | Reduce repair costs, add faucets | Inject currency via events |
| Wealth inequality (high Gini) | Add catch-up faucets for low-level players | Progressive sinks on wealthy players |
| Stagnant market (low velocity) | Reduce AH fees, add new desirable items | Seasonal resets |
How and when players receive rewards determines engagement, retention, and spending patterns. Reward schedules are behavioral psychology applied to game design.
| Schedule | Pattern | Player Behavior | Best For |
|---|---|---|---|
| Fixed Ratio | Reward every N actions | Steady grind, predictable | Crafting (make 10, get reward) |
| Variable Ratio | Reward after random N actions | Compulsive engagement, slot-machine effect | Loot drops, gacha |
| Fixed Interval | Reward every T time | Login spikes at reset, dead time between | Daily/weekly quests |
| Variable Interval | Reward at random times | Constant checking behavior | World events, rare spawns |
Most games combine all four. Daily quests (fixed interval) keep players logging in. Loot drops (variable ratio) keep them playing. Crafting (fixed ratio) gives a sense of control. World bosses (variable interval) create excitement.
Structure resets to create natural play sessions:
Stagger resets across time zones. A single global reset creates server load spikes and punishes players in unfavorable time zones.
Players who take breaks must be able to return without feeling hopelessly behind:
Never make catch-up trivial (undermines current players) or impossible (kills returning players). Target: a returning player should reach current-content readiness in 1-2 weeks, not 1-2 days or 1-2 months.
Progression speed varies by game design intent:
| Curve | Shape | Feel | Use Case |
|---|---|---|---|
| Linear | Constant XP/level | Predictable, can feel grindy | Sandbox games, skill-based |
| Logarithmic | Fast early, slow late | Quick hook, long endgame | Most MMOs |
| S-Curve | Slow start, fast middle, slow end | Tutorial buffer, satisfying mid-game, prestige endgame | Story-driven MMOs |
| Exponential | Slow early, fast late | Feels increasingly rewarding | Idle/incremental games |
For MMOs, the S-curve is typically best. The slow start teaches mechanics. The fast middle provides dopamine. The slow end creates aspirational long-term goals.
For any random reward system (loot boxes, gacha, boss drops), implement a pity timer that guarantees a reward after N failed attempts:
Pity timers prevent the statistical tail where unlucky players never get the reward. Without them, ~1% of players will have an experience 3-5x worse than median, generating support tickets and churn.
Player-to-player trading is where economies get complex. A well-designed auction house (AH) is the backbone of a healthy player economy. A poorly designed one enables exploitation.
Two primary models:
Bid/Ask (Order Book): Players post buy orders and sell orders at specific prices. Transactions execute when a buy price meets a sell price. This is how real stock exchanges work.
Simple Listing: Sellers list items at a fixed price. Buyers browse and purchase.
For most games, start with simple listing and add buy orders later if the economy matures. EVE's order book works because EVE players self-select for complexity.
Fees are the primary currency sink in a trading economy:
| Fee Type | Rate | Purpose |
|---|---|---|
| Listing fee | 1-5% of list price | Prevents spam listings, penalizes overpricing |
| Sale tax | 2-10% of sale price | Primary gold sink, funds NPC economy |
| Cancellation fee | 50-100% of listing fee | Prevents price manipulation via cancel/relist |
| Deposit (refundable on sale) | 5-15% of list price | Discourages unrealistic pricing |
A combined 5-10% effective tax rate (listing + sale) is standard. Below 3% and the AH is not a meaningful sink. Above 15% and players avoid the AH, trading directly instead (which you cannot tax).
Prevent abuse without destroying the trading experience:
Real-money trading (gold selling) is the persistent threat to every game economy:
You cannot manage what you do not measure. Economy health requires real-time monitoring with clear alert thresholds.
| Metric | What It Measures | Healthy Range | Alert Threshold |
|---|---|---|---|
| Gini coefficient | Wealth inequality (0 = equal, 1 = one player has all) | 0.3-0.5 | > 0.7 |
| Median wealth by level | Whether progression feels rewarding | Monotonically increasing | Flat or decreasing |
| Velocity of money | Transactions per unit of currency per day | 0.1-0.5 | < 0.05 (stagnant) or > 1.0 (panic) |
| CPI (item price index) | Track prices of a basket of 20-30 key items over time | Stable or slow increase (1-3%/week) | > 5%/week (inflation) or < -2%/week (deflation) |
| Currency creation rate | Total new gold entering per day | Within 10% of design target | > 25% above target |
| Currency destruction rate | Total gold removed per day | 70-90% of creation rate | < 50% of creation (sinks failing) |
| Active trader ratio | % of players using the AH | 30-60% | < 15% (AH irrelevant) |
| Median time-to-earn | Hours to earn key items | Within 20% of design targets | > 50% deviation |
Track wealth in brackets:
Level 1-20: Median 500g, P90 2,000g, P99 10,000g
Level 21-40: Median 5,000g, P90 20,000g, P99 80,000g
Level 41-60: Median 25,000g, P90 100,000g, P99 500,000g
Max Level: Median 100,000g, P90 1,000,000g, P99 10,000,000g
When P99/Median ratio exceeds 100:1 within a level bracket, wealth concentration is extreme and likely driven by exploits or RMT.
Select a basket of representative items and track their prices daily:
Compute a weighted price index (CPI-style). Plot weekly. An upward trend beyond 3%/week signals inflation that needs intervention.
Set automated alerts for:
The most catastrophic failure. A bug allows players to duplicate currency or items. Within hours, exploiters generate billions of gold. Prices skyrocket. The economy is destroyed.
Prevention:
Response:
Sinks are too aggressive. Players hoard currency because spending feels punishing. Market activity drops. New items are not listed because sellers cannot recoup crafting costs. The economy stagnates.
Prevention:
Response:
Professional gold sellers do not just farm -- they manipulate markets:
Prevention:
Players find an exploit that generates currency faster than intended: a quest that can be repeated, a vendor buy/sell loop with positive margins, a crafting recipe that produces more value than its inputs.
Prevention:
Premium currency is the bridge between real money and the game. Design it carefully:
| Tier | What It Sells | Player Perception | Revenue Impact |
|---|---|---|---|
| Cosmetic-only | Skins, mounts, housing | Positive (fair) | Moderate, stable |
| Convenience | Inventory slots, fast travel, XP boosts | Neutral to mixed | High |
| Time-saving | Level boosts, crafting speedups | Controversial | High, risky |
| Power (pay-to-win) | Best gear, stat boosts | Strongly negative | Short-term high, long-term destructive |
Stay in cosmetic and convenience tiers. The moment paying players have a statistical combat advantage over free players, the game loses competitive integrity and community trust.
A seasonal battle pass creates predictable revenue and engagement:
Loot box regulation is expanding globally:
| Jurisdiction | Requirement |
|---|---|
| Belgium, Netherlands | Paid loot boxes banned (classified as gambling) |
| China | Must disclose exact drop rates |
| Japan | Kompu gacha (combining random items for a prize) banned |
| UK | Under review; age-gating likely |
| US | FTC scrutiny; state-level bills emerging |
| Australia | Senate inquiry recommended regulation |
| South Korea | Must disclose probabilities, government audits |
Safe practices regardless of jurisdiction:
| Anti-Pattern | Why It Fails | Better Approach |
|---|---|---|
| Single currency for everything | One inflation vector destroys the whole economy | 3-5 purpose-specific currencies |
| No sinks at endgame | All faucets, no drains = hyperinflation within months | Endgame cosmetic sinks, progressive repair, AH taxes |
| Selling power for premium currency | Destroys competitive integrity, drives away non-payers | Cosmetic and convenience monetization only |
| Fixed NPC prices with rising player wealth | Players outgrow NPC economy in weeks, vendors become irrelevant | Progressive pricing tied to server wealth metrics |
| No trade restrictions on new accounts | Gold sellers create accounts, dump gold, and delete | 24-72hr trade lockout, level minimums, volume caps |
| Flat drop rates without pity | 1% of players never get the reward, generating rage and churn | Pity timer guaranteeing reward after N attempts |
| Manual economy balancing only | Humans react too slowly to exploits and market shifts | Automated telemetry, alerts, and progressive sinks |
| Ignoring secondary markets | Players trade outside the game (Discord, forums) to avoid AH tax | Keep AH fees below 10% to keep trading in-game where you can monitor it |
| Seasonal resets without carry-forward | Players feel their time was wasted when progress evaporates | Let seasonal currency convert to permanent (at reduced rate) |
| No separation between earned and bought currency | Cannot distinguish organic economy health from whale spending | Track earned vs premium currency flows separately in telemetry |
| Skill | When to Load |
|---|---|
gamedev-server-architecture | Server-side persistence, database design for economy state, transaction processing |
gamedev-multiplayer | Networking for trade transactions, auction house real-time updates |
gamedev-mmo-persistence | Saving player inventories, currency balances, transaction history |
gamedev-2d-platformer | Simpler economy design for non-MMO games (shop systems, upgrade currencies) |