Welcome to Raph Koster's personal website: MMOs, gaming, writing, art, music, books.
Welcome to Raph Koster's personal website: MMOs, gaming, writing, art, music, books.

The whole Web
Raph's Website

These are full-blown essays, papers, and articles.

Slideshows and presentation materials from conferences.

Interviews and Panels
Reprints of non-game-specific interviews, and transcripts of panels and roundtables.

Excerpts from blog, newsgroup, and forum posts.

The "Laws of Online World Design" in various forms.

A timeline of developments in online worlds.

A Theory of Fun for Game Design
My book on why games matter and what fun is.

Insubstantial Pageants
A book I started and never finished outlining the basics of online world design.

Links to resources on online world design.

This is excerpted from a Usenet thread in which we were discussing many, many topics. Along the way, we talked about UO's economy. I think it was in the context of how much Trammel had affected things. Anyway, I gave a brief historical summary of my take on the UO economy at various stages of its development. For a more analytical approach, I highly recommend Zach Simpson's excellent paper on "The In-Game Economics of Ultima Online" which you can find on the links page. The whole debate started because I said that UO's economy was subsidized after the introduction of Trammel, to which the retort was, "all online game economies are subsidized..."

The evolution of UO's economy

Let me rephrase, it's a radically subsidized economy. I'm not speaking of removing all subsidization, I'm talking about not flooding the game with gold.

NPC shops are a subsidy, obviously, as they buy stuff for which there is zero real market. In a non-usage game, we could say they subsidize the gameplay of adventuring (which gives you lots of useless stuff) or pursuing your chosen playstyle (like say, crafting stuff that nobody wants, because it's how you have fun).

In a game like UO, which is usage based, the NPC shops are rewarding macroing and the pursuit of skill points. They're not even rewarding having fun necessarily. They're just a gold spigot.

Model one

The original UO system had shops adjusting their prices and stock based on what sold and what was bought. They would acquire stuff from players speculatively. Then they would attempt to sell those goods. If they ended up with too many in stock because people did not want the goods, they lowered their purchase price, eventually to zero. If they had too few, the price went up, and they might spend their operating capital on making more of that good themselves. They would cash in goods that did not sell for more operating capital (subsidy) but they would not restock them themselves, preferring to stock items that sold.

The data which they used was supposed to be based on the local market, so that price differences would be exposed between different shops and different cities.

What actually happened in the first model, of course, is that distances were trivial, so no local markets. But more egregiously, there were two huge issues.

One was massive overproduction. You see, the incentive to produce wasn't money or commerce. It was macroing advancement. Every single crafter in the game went into the hole immediately, because they gladly traded gold for skill points. This didn't get them anything except the satisfaction of a label next to their name, and made no economic sense on their part, really, but they did it anyway.

Shops all quickly has a surfeit of absolutely everything, and purchased nothing. And the players rebelled, resulting in a system whereby all data tracking for a given shop was cleared on a periodic time basis.

Model Two

The other thing that happened was hoarding. There were not nearly enough drains in the game economy, even with item decay. This manifested badly because gold was in a closed loop; we had a fixed amount of currency in the world--it was a tweakable number, of course, so we could "print it" when we needed it, like a central bank. But the issue was that we had spawns that injected currency into the system.

What happened was a collapse of liquidity. The gold crafters had was all spent on turning it into skill points. That money all went to the resource miners. From there it went to the hoarders. And there it stayed. There wasn't anything to spend it on.

Hoarding caused all sorts of disastrous problems, of course, culminating in special measures like the Clean Up Britannia campaign. We broke the gold loop, and instead let currency float to the equilibrium point with the actual drains we had in the game.

And there it stayed, stable for quite a long time. Massive hoards were broken up as players quit and houses decayed leaving stuff to decay--or when they decided to break up the hoard to give to friends, many of whom lost large sums. Much wealth was redistributed via jerks, actually, and to a degree they were a positive economic force because they were so rich already they let lots of people's wealth decay away too. :P

In the end, Model Two ended up with a stable exchange rate of 1 dollar for 200 UO gold from around three months after launch until well after I was no longer on the project.

Model Three

UO Renaissance removed that wealth redistribution via jerks (which was a great move in other arenas, but sure didn't help the economics), in the process made wealth gathering from the spawn spigot much easier, and shortly after UO currency began to collapse. Said collapse has been ongoing and last time I checked (a while ago), a UO gold coin was worth a hundred times less than it had been during the days when the economy was stable. I haven't been graphing the exchange rate the way I used to (see, I did fall in love with stats and data) so I don't know if it's stabilized yet.

The radical subsidy is that the game now pays you UO gold for, well, logging on. I expect to see another dip in the currency value as the effects of this "no need to refresh your house" thing go into effect.

Does this all matter? After all, it does find equilibrium. The answer, sadly, is yes, because there's countless pieces of static data involving currency scattered throughout the whole game. The reward you get from a critter. How expensive it is to buy a castle. What the living expenses are for a mage. Revising all that so that players feel like they are getting adequate reward in the face of currency devaluation is a massive pain in the butt, and if you increment it, all you're doing is contributing to the devaluation since you're increasing what players get...

The typical end of this cycle on an online game is a wipe of all equipment and goods. They usually leave the characters intact. (Characters are wiped over a different but very related kind of database deflation, covered in that citation I gave you for that post about Lady Vox).

BTW, this issue was first noticed on Habitat in 1985, cf "The Lessons of LucasFilm's Habitat."

Taken from Usenet posts in early 2002. The "Lady Vox" post can also be found on this website in the snippets section. The reference to "usage based" is referring to the fact that UO your character advances via repetitive use of a given skill. Therefore, the advancement encourages the production of goods which nobody wants.

Child's Play

A Theory of Fun
for Game Design

Cover of A Theory of Fun



Buy from Amazon

After the Flood

Cover for After the Flood CD

Available on CD

More stuff to buy

Gratuitous Penguin 2006 Wall Calendar

Gratuitous Penguin 2006 Wall Calendar

Receive CafePress Updates!


click here to visit the Legend website

"The world the way they thought it was..."