Sinks & Faucets: Lessons on Designing Effective Virtual Game Economies
For a virtual in-game economy to maintain economic equilibrium — the relative stability of its core currency and items — it must balance the rate of asset issuance with the rate of asset consumption/demand.
Stuff in ~= Stuff out.
To do this it has a couple of options:
- Continuously grow the rate of growth of new users to allow older players to sell to newer players.
- Continuously introduce newer, harder, and more resource-intensive end-game content for existing players to expend more resources.
- Balance the creation of gold and items (faucets) with the removal of said gold and items from the game (sinks).
If this balance is mismanaged, the player experience and retention suffer.
Luckily, virtual economies have existed for decades and there are many precedents of failed experiments and hyperinflated games that we can learn from. We hope to introduce how previous virtual economies have deployed sinks, along with available data on the efficacy and popularity/reception of different strategies. We also comment on new balancing tools available to community & culture first p2e games.
Table of Contents / Tl;dr
Part 1: Learnings from Previous Games
- Utilities, Rare Collectibles, and Gambles
- Taxes on Trade and Land
- Staking, rotating High-Velocity Cash to Consumables/Items
- Unlock more Game Content
Part 2: Strategies Uniquely Available to Web3 Games
- Seed mods and forks as token sinks
- Community-based Burn Mechanisms
- Build and Own Critical Infrastructure
Conclusion: Conditions where Sinks are Effective
Part 1: Learnings from Previous Game
Solution 1. Utilities, Rare collectibles, Gambles
Charge players for small quality-of-life upgrades or for the use of public goods as a token sink. This taps into a player’s desire for accomplishment and her desire to avoid unnecessary time expenditures.
Some examples are
- Flight paths in WoW and mounts/upgrades for faster travel (magic carpet rides in OSRS)
- Armor repair (OSRS, Elder Scrolls) — notably this replicates real-life depreciation costs
- Tolls for accessing areas (Al Kharid gate), including instanced dungeon costs
- Salons to change your appearance or name (Salon Durem in Gaia)
- The ability to pay to reset your stats/skill trees
- Buying and decorating houses (Elder Scrolls, Final-Fantasy)
Most of these utility-type sinks are accepted as a staple of games — a tax so frequently paid that its pain becomes negligible. On the other hand, there have been examples of utility tax sinks so egregious that they drove away many players. (Note responses to poll above)
For instance in Aion Classic, enchantment stones gave additional stats boosts to your weapons and armor. A maxed competitive end-game character, both for PvE and PVP required at least +10 boosts in armor and weapons, while the maximum possible enchantment was +15. Upon every failure past +10, the weapon/armor returned to +10 enchantment. Players not only had to grind hundreds of hours for stones, but they failed incredibly often and required large amounts of cash per enchantment attempt.
Phantasy Star Online 2 also uses the same upgrade mechanic. For the highest tier (11 or 12-star weapon) the odds are 0.0014% to upgrade it to +10 in one go, and each weapon requires +10 three times over to fully upgrade.
Takeaway: Grinding is not inherently evil. It fosters a sense of mastery and advancement, but only up to a point. Tough utility taxes can be useful gold sinks, but those that require grinds that are more frustrating than rewarding will result in player attrition.
Periodically auction/sell rare collectibles in in-game currency. This can be done consistently through a luxury merchant NPC, as part of a special event, or as a gambling game. This is driven by a player’s desire for social influence and ownership.
- Collectibles in Neopets
- Neverwinter’s Sergeant Knox auction
- Condensed gold, sold by an NPC in Runescape, required to make the (literal) Gold sink
As per our poll above, this is one of the most community-friendly gold sinks.
Gambling games are designed to have a negative monetary expected value, but players are incentivized to play because this is the only way to obtain rare cosmetics, is an easy way to obtain extremely rare items, or is fun. See gacha games. Each time a player participates, they must pay in the native token. Because expected payout is negative, this is a token sink. This is driven by the player’s desire for ownership.
Expected Value = Fun - EV(monetary loss) > EV(monetary gain)
From our poll, gambling was by far the most hated form of gold sink. We conjecture that this is because many gambling gold sinks created pay-to-win avenues.
Takeaway: Don’t create pay-to-win highways through gambling games. Embed rare collectibles into the lore and experience or make it a meme. People love flexing their wealth.
Solution 2. Taxes, trade, and land
Levy taxes on auction houses, marketplaces, p2p trades, and on the possession of land/houses (a Georgian Land Value Tax). This requires first for the game creators to control a marketplace.
- Auction house fees in MMOs (WoW, Diablo, Runescape). Of particular note is that the auction house fees (Grand Exchange Trade Tax of 2%) in Runescape are used to buy back items in the GE. This can be algorithmically adjusted to target items that are undergoing bursts of additional supply. The tax also positively encourages player-to-player interaction, adding social depth to the game.
- Taxes on idle resources. Lars Doucet outlines Henry George’s Land Value Tax applied to Eve Online Factories, and how that successfully combated the problem of rampant factory speculation. Star Atlas also will take this approach (see pg 6. Of their Economics Paper).
- The 4.25% of tax on Axie Marketplace sales.
Transaction taxes have historically been the largest sink for ISK (Eve currency) for the past 2 years. The broker’s fee (another marketplace tax) comes at a close third.
In a game where trade becomes a larger portion of gameplay in later stages — most MMOs would fall under this category — adding a heavy transaction tax effectively takes currency out of the oldest and richest players’ coffers, taking currency out of the game where it is the most abundant while preserving holdings of new players going through the onboarding process.
Applied to Axie these mechanisms look like the following:
Takeaway: Taxes come at an unfortunate third in the most-hated gold sink poll, but they are very effective. Tax marketplaces, trade and idle resources, and use tax revenue for additional burning or community oriented activities. Perhaps experiment with new taxation models taken from the large history of optimal taxation literature to find one that works even better than those tried before.
Solution 3. Crafting
Crafting can act as both an item and a currency sink. Players use lower-level items to create higher tier items/consumables/cosmetics which consume component items and at times in-game currency. This is driven by a player’s desire for mastery, curiosity in discovering new items and experiences, and creativity in finding new combinations.
- Guildwars2 has a crafting mechanic used as an item sink. Crafting functions as a black hole for both lower and higher-level materials.
- Another example is the Construction skill in Runescape. Leveling up construction requires repeatedly building and then destroying furniture, consequently destroying the materials consumed to make it. Each plank required to construct furniture costs money to create, while nails and other components must be created or purchased. In the existence of a taxable marketplace, high dollar-volume trades incentivized by difficult-to-train skills can be a powerful gold sink. Planks are some of the most heavily traded items on the Grand Exchange. Note that to reach the maximum level, construction requires at least a 100m GP investment (nearly 400M for the fastest method) which is a lot of GP for a casual player.
Crafting an item sink only works if the item being produced via crafting is not widely available in the marketplace or by other means. If an item is obtainable through methods that take less time and money than crafting, and/or players can easily buy up raw materials then sell the crafted product for a profit, this system may end up generating even more gold in the system.
Takeaway: Used well, crafting can be both an addition to the experience, lore and social aspects of the game while providing both a healthy source of a sense of mastery/advancement and a currency sink.
Solution 4. Staking/Locking up Currency
This is much more familiar for those in crypto, but similar solutions exist in web2 games.
Managing Miscellania is a minigame in Runescape. A player can deposit up to 5–7.5M gold (depending on quest completions) in coffers. Every day, up to 50–75K is taken out of the coffers and paid to “workers” who harvest resources, which are then claimable by the player. The rewards tend to always be profitable but are resources that are widely required and consumed while skilling. Players deposit gold into the coffers because of their desire for accomplishment — advancing their skills.
Managing Miscellania effectively locks and burns millions of coins and swaps them for lower velocity items or consumables necessary for progressing in the game.
Takeaway: Rotate high velocity currency to low velocity items or consumables.
Solution 5. Unlock More Game Content
Instead of requiring fiat to purchase additional game content, membership, DLCs (downloadable content), sell it for a large amount of in-game currency to players. The quality of content and player experience will be the key driver in affecting the player’s desire for unpredictability (curiosity) and accomplishment (to finish all of the game) and enticing them to purchase additional content with in-game currency.
Runecsape bonds grant 14 days of membership but can be bought from another player for 4.7M GP each. 1 month purchased through fiat costs 11USD. In just the past hour of writing (Weekday ET 9pm) 553 bonds have been bought on the Grand Exchange, for a total of approximately 2,500,000,000 GP. Pro-rated this means bonds are approximately 1.2% of daily GP traded a year. 2,500,000,000 GP is also around 250 hours of gold farmed using the highest gp/hr method available in-game. Check live data here.
If Jagex were to instead sell bonds for GP, this would generate an extremely healthy gold sink.
Takeaway: Rotate pay to play to skill/play to play. Considering web3 games don’t monetize through purchases of the game from the producer, this can work especially well.
Solution 6. Donations/Events
The most community-friendly gold sink according to our Reddit poll and in my personal experience is donations and events. Incentivize people to give away/burn gold in goodwill or for rare collectibles and titles by enticing their desire for meaning or social influence.
- Gaia online’s 2009 Bailout event. To accompany the remodeling of NPC shops, Gaia hosted a timely Bailout 2009 event, where each shop was purported to owe taxes to the “GRS” and players needed to donate to each shop to make sure it survived the update. One criticism of the event was that the outcome was pre-determined regardless of the total amount donated, and that larger donors didn’t receive additional benefits.
Because it is easy to automatically return funds if a threshold is not met in crypto, it becomes much easier for game developers to create funding-dependent updates as community events.
- The Well of Goodwill in RS3 took donations in items and gold. Jagex, the parent company of Runescape, then matched GP donations at a ratio to GBP and donated to charities.
Those who donated above a certain amount to the Well of Goodwill received rare titles — perhaps the rarest and most coveted of items within games.
The well allowed those that didn’t have a bank account or had trouble wiring money to donate in a way that felt as if it were low-cost basis (via accepting in-game currency), fun, and community aligned. Those who donated more than 5B GP (a truly STAGGERING amount for those who’ve never played the game) were awarded the title “The Billionaire”. Big Flex.
Despite the well’s immense popularity, the well will not be returning the game.
In line with the Reddit poll, I personally really enjoyed this type of sink. It bred a community experience that integrated with the lore, and updates to the game in the case of Gaia. In the case of Runescape, it gave great bragging rights and added an element of fun and achievement to donations and gold burning.
Takeaway: Integrate gold sinks directly into community events/lore. Create fun experiences around gold sinks and give people/guilds visible bragging rights dependent on the amount of gold burned.
Part 2: Solutions uniquely enabled by Crypto
In the previous section, we explored balancing solutions tried and tested by traditional games. Web3 games, however, come with multiple additional opportunities to produce organic sinks and balancing mechanisms. These include seeding mods and forks to community-led sinks. Both rely on the open-source, interoperable, and community-first nature of web3 games and only work for truly decentralized community & culture first p2e games.
Solution 1. Seed mods and forks as token sinks
modded versions of games have always been at the forefront of pushing gameplay to new frontiers and genres. In fact, some of the most popular games of all time were inspired or directly created in/as mods of other games.
Day Z for instance was at first a mod of Arma 2. Dean Hall took the best aspects of Arma 2 and transformed it into a survival game with zombies. Soon there were spin-offs of spin-offs of DayZ, including a mod developed by a then-unknown “Player Unknown” named Day Z: Battle Royale. This not only popularized the battle royale genre but later evolved into PUBG, one of the most popular video games in history.
Because Web3 games can monetize through secondary sales and taxes on core infrastructure, and thus benefit from the expansion of the entire game ecosystem, they have strong incentives to seed new mods and player creations which in turn creates organic demand for their ecosystem tokens. This is a natural counterforce to the innovator’s dilemma and can create ever-evolving and thereby longer-lasting games.
This is the strategy taken by Sandbox, Aavegotchi, and Sky Mavis. The core team, in all three cases, have built infrastructure, deep liquidity, and an initial ownership base. Next, they encourage ecosystem entrepreneurs to build mods and subgames atop their platform and collect revenue in the form of an activity tax. This is why Axie is branding itself as a “nation”.
You can think of Axie as a nation with a real economy. The holders of the AXS token are the government that receives tax revenues. The inventors/builders of the game, Sky Mavis, hold ~20% of all AXS tokens.
From the Axie Whitepaper.
Aavegotchi minigames are community-developed games that are playable only by those who have an Aavegotchi. Because each game is different, Aavegotchis with different traits are beneficial for different games. In many cases, an Aavegotchi that has a lower base rarity score performs better than one with a higher rarity score in certain minigames creating demand for a wide variety of Aavegotchis. For designated periods of time, playing the minigame can also earn you XP, giving you a higher chance of winning a leaderboard GHST prize. According to the Aavegotchi wiki:
Each mini-game will use different Aavegotchi traits to enable fair gameplay for a wide variety of trait distributions. For example, a HYPER-AGGRESSIVE Aavegotchi may perform well in an Aavegotchi Fight Club, but may not be well-suited to a cake-baking mini-game.
But in order to use the strategies above, games have to decentralize (can’t rug access) while incentivizing talent building on the platform (see the Roblox DevEx program). They have to also allow modders and ecosystem developers to take a cut of the economic output they create (through royalties and contribution-correlated rewards).
Takeaway: Allow community devs and game designers to create experiences that drive organic demand for your tokens and infrastructure.
Solution 2. Community-based Burn Mechanisms
Jon Huang estimates that approximately 98% of Axie DAU are scholars. Considering Axie has a DAU of 2M and YGG as the largest guild has 10K scholars, we can estimate that the number of scholars in guilds probably does not exceed 5–10%. If this number increases, or if we factor into account small managers, then introducing burning mechanisms at the guild-scholar level becomes very important.
An example of such a mechanism is to fairly rotate scholar-held SLP (presumed to be high velocity) to guild or protocol-held SLP. This could happen via guild-level games where scholars can stake/trade-in SLP for a chance at better cards or items, though the profitability of such a strategy to a guild is uncertain. Alternatively, the Axie Treasury could allocate AXS rewards for guilds performing community supporting initiatives which might include community-led burn mechanisms.
An example of an existing community burn mechanism is YGG’s commitment to staking its SLP treasury for RON or previously dedicating its SLP entirely to breeding new Axies.
Takeaway: Seed and support communities with a culture of long-term supporting the game
Solution 3. Build and Own Critical Infrastructure and Collect Fees as Sinks
This is an extension of taxes from part 1.
P2E games can build, govern and monetize through owning core infrastructure and levying taxes on the usage of core infrastructure. Sky Mavis owns its own chain, marketplace, and DEX (Katana) and charges a gas cost (in RON) and a marketplace fee (4.25% of ETH) to use these protocols. These protocols all generate fees, which can then conduct buyback and burn, or taxes can be directly burnt as well.
It is arguably much easier for a p2e game studio to develop protocol infrastructure than try to inorganically bootstrap an ecosystem of games (strategy 2). If P2E economies are protocols, then the more protocol scope and coverage they can cover the more value accrual can emerge.
But in an open-source project and in the long term, high take rates are likely unsustainable. Only if the P2E game delivers the highest quality core infrastructure, incentivizes the best operators to contribute to their infrastructure (over building their own) can they justify critical infrastructure as a token sink or as a large revenue source for their treasury operations.
Takeaway: Create the best critical infrastructure and collect fees as sinks or use protocol revenue for buyback and burn operations.
A balance of faucets and sinks is required to maintain a healthy equilibrium of prices within a virtual economy. However, a sink never guarantees that any game-economy token will regain its ATHs. In fact, sinks only work if people use them.
Web2 games benefit from the fact that they are closed economies where once money enters it is difficult for it to leave (no in-game currency to USD exchange that is easy, transparent, and legal). This means they also benefit from a larger degree of separation of their currency from fiat in the minds of players. This means players
- Have no choice but to spend the money inside the platform/game
- Are less likely to conduct highly mathematical calculations in returns when engaging in any activity and are therefore
- More likely to use currency for random things or for fun
A p2e game does not enjoy these benefits. To encourage players to participate in sinks in an environment where the USD value of your inputs and outputs is extremely visible, it must offer either additional currency for burning (which further increases supply) or some other form of additional utility.
And thus, in an open economy, only if the following equation holds will sinks be effective.
Motivation + additional utility from using sink > cost of using a sink
What determines the efficacy of the sink is whether people use it, and people will only use it if they see value in doing so.
We are just at the beginning of understanding how to build out in the open. Our mission at 1kx is to become the most helpful investors to builders looking to build the next big play to earn ecosystem.
If you’re building something in this space, we’d love to chat~