Ata Kuyumcu's Blog

Balancing an MMO economy with a black market

Balancing an MMO economy with a black market

Albion Online has a cool design constraint that sounds absurd when you say it out loud: every item in the game is player-crafted. There are no NPC vendors selling gear. The world produces nothing on its own except raw resources. The whole sandbox is built on top of that constraint.

But Albion also has mobs that drop loot. You go fight a heretic, you sometimes get a T4 bag. So somebody has to have crafted that bag. Where does it come from?

The answer is the Black Market merchant, an NPC sitting in Caerleon (the central PvP-adjacent city) that exists to broker between crafters and mobs. It’s the cleverest piece of economy automation I’ve come across in an MMO, and the design choices it embodies generalise well past MMOs.

Conceptually:

  1. A mob has a loot table that lists items it could drop.
  2. When the mob dies, the game checks the Black Market’s stock for those items.
  3. If the item is in stock, roll the drop chance, deduct from stock, and give it to the player.
  4. If the item is not in stock, the Black Market creates a buy order for it. The order sits there until a crafter walks up to the Black Market, sees the price, and decides to fill it.
  5. The crafter delivers the item, gets paid, the Black Market’s stock goes up.
  6. Periodically, items in stock decay (“corruption”), reducing total supply.

So crafters supply the Black Market, the Black Market supplies the mobs, and mobs supply the players. The player-to-player economy and the PvE economy are the same economy, just with a buffering layer in between.

Three things make this work better than it has any right to.

1. The closed silver loop

This is the bit that takes a second to appreciate: the Black Market never adds silver to the economy.

When a mob drops silver, a portion of that silver is diverted into the Black Market’s budget instead of going to the player. That diverted silver is what funds the buy orders the crafters fill.

Without Black Market:
  Mob drops 100 silver + items
  → 100 silver enters the economy.

With Black Market:
  Mob drops 80 silver + items
  → 80 silver enters the economy.
  20 silver goes into the Black Market's budget.
  That 20 silver pays a crafter for the item the mob just dropped.
  Net new silver in the economy: 80. The other 20 moved internally.

Most “player-driven economy” pitches I’ve seen handwave where the silver actually comes from. Albion’s answer is in the diagram above: existing silver gets routed through a crafter on the way to the player, with no net new silver entering the economy.

Robin Henkys (the game director) put it pretty plainly in May 2017 announcement video on the system: the buy orders the Black Market generates always correspond to a percentage of the silver dropped by the mob, which is what keeps the silver supply stable. SBI can dial it up or down based on what they observe in telemetry.

2. The price curve is a PID controller in disguise

The Black Market needs to set prices on its buy orders, but it doesn’t have an oracle for the “right” price. So it does what any reasonable control system does: it adjusts the price based on an error signal.

The exact curve isn’t public, but community testing over the years has converged on something close to this:

Buy order opens at: base_value * tier_multiplier
If unfilled for 1h:   price *= 1.05
If unfilled for 2h:   price *= 1.10
If unfilled for 6h:   price *= 1.30
If unfilled for 24h:  price *= 1.80
Cap: ~3 to 5x base price.

If the price is too low, no crafter fills the order, so the price goes up until somebody does. If the price is too high, lots of crafters race to fill it, demand is satisfied, the Black Market stops opening new high-priced orders at that level, and the effective clearing price falls.

The error signal is “is this order filled?”, the control output is price, and the plant is the crafter population’s collective willingness to craft and transport. It’s a P controller, basically, which works fine here because the plant has its own natural delay and nobody actually needs sub-second tuning of bag prices.

The thing I find satisfying about this is that SBI didn’t have to predict the right prices. They had to predict the right mechanism for discovering them. Much easier problem.

3. Corruption is a progressive sink, tuned per tier

If items go into the Black Market and stay there forever, eventually you have a stockpile and crafters stop being able to sell. So items in stock decay over time. SBI calls this “corruption.” Estimated rates from player observation:

T1-T3: ~30-40% per tick   (tutorial tier; would be free otherwise)
T4-T5: ~15-25% per tick   (mass-market; everyone can craft these)
T6-T7: ~5-10% per tick    (mid-to-high; specialised crafters)
T8+:   ~1-3%  per tick    (few crafters; demand is what holds value)

A flat decay rate would be easier to implement but wrong. T1 items are functionally free; if you decay them at the same rate as T8 items, you either over-destroy T8 (collapsing the high-end market) or under-destroy T1 (drowning the low-end in trash). By making the corruption rate a function of tier, SBI gets to set a separate equilibrium for each tier. The newbie crafter making T2 gear can still sell to the Black Market because there’s always a buy order at minimum price, but most of what they sell evaporates and never reaches a mob loot table. The T8 crafter, by contrast, can be reasonably confident their item will actually find a player.

The system targets the tiers that would otherwise accumulate to worthlessness, while preserving the tiers that need scarcity. Same idea as a progressive tax, applied to item supply.

Caerleon sits at the centre of the map and is the only city with road connections into the Outlands, which are the full-loot PvP zones. The Black Market is in Caerleon, and only Caerleon. There is no Black Market kiosk in the safe blue-zone cities, which means: if you want to sell to the Black Market, you have to physically transport your crafted goods to Caerleon. Through PvP zones, in many cases. Carrying a backpack full of valuable items. Thus, the Black Market’s location single-handedly creates:

  • A category of player whose entire game loop is “craft and transport.” (Risky shipping-runs for profit.)
  • A category of player whose entire game loop is “intercept transporters.” (Gankers patrolling the choke points around Caerleon.)
  • A category of player whose entire game loop is “escort transporters.” (Guild caravans, hired protection.)

If the Black Market were available in every safe city, crafters would never leave the blue zones, transporting risk would disappear, and a huge chunk of the PvP content would simply not exist. The economy would still function, but it would be much less interesting to actually play.

Phrased as a design principle: economic infrastructure placement is PvP content design. Where the market is determines where players go, which determines where fights happen. All of that gank content comes for free as a side effect of where the market sits.

Two pointers if you want to go deeper. SBI’s 2017 announcement post is where Robin Henkys originally walked through the mechanism; it’s the primary source for the 20% diversion figure and the corruption-trader framing. Matt Woodward’s hour-long GDC 2017 talk Balancing the Economy for Albion Online sits one level up, covering the balance methodology this all sits inside: anchors, target silver-per-hour rates, telemetry feedback, and constraint-driven design.

There’s a lot more in Albion’s economy worth writing about: the silver-to-gold exchange spread as a soft sink and a way to capture the RMT market, the territory upkeep loop that connects guild PvP back into resource demand… I might get to those in the future. The Black Market alone is enough to chew on for one post.