« Maslow Short-Circuited | Main | Terra Nova Welcomes T. L. Taylor »

Oct 14, 2003



He told Elizabeth Kolbert, not Elizabeth Koster.


Sheesh, next you'll be telling me that the Chief Creative Officer at Sony Online Entertainment isn't Raph Kolbert...

Sorry about that. I'd edit my post, but I'd hate to give the impression of knowing what I'm talking about.


So, to the questions at hand:

1) Simple economic systems causing grief? The only real sources of grief I'm aware of are the more "complex" systems some games try to implement. On one server in UO, a dedicated group of players used to go buy out reagents from vendors as soon as they'd get more in stock. This meant that this one group of players were able to effectively monopolize the reagents available from vendors. While this a viable form of economic activity and interesting to observe, it simply wasn't very fun for the people caught having to pay inflated prices to the monopolists. A "simpler" system that had infinite amounts of reagents available for purchase would have eliminated this issue.

2) Does it really matter? No, I don't think so. This requires a bit longer answer.

One problem with the game-type worlds and economics is that economics don't always support "fun". Everyone wants to "win", no one wants to be left short on funds because "sorry, the economy won't provide for your needs today." It's an extension of the hula-hoop problem Raph talked about in the early days of UO; people expect to get money as a result of their action (labor). Violate this, even if it's economically "realistic", and you get pissed off players.

Consider traditional single-player RPGs. I recently replayed Might & Magic 7 again from start to finish, and legitimately ended up with over 2 million gold pieces at the end. I earned this money through legitimate use of the in-game economy. Obviously if this happened in the "real world", there would be serious economic effects. Yet, inflation doesn't add to my sense of fun; I like the ability to go into an expensive magic store and buy whatever I want on a lark without the storekeeper hastily raising his prices because I've scrimped and saved to an extreme degree.

No offense to any of the bloggers here, but I think a designer can get a "good enough" economy without being an expert in the area. I think this is true for most areas of the game; I have an interest in linguistics, but that doesn't mean I expect every game designer to hire a Tolkien-level writer on staff to construct a believable language for every race/species that's in the game. While it would add to immersion, it also adds to the costs of developing worlds that already cost way too much to develop as it is in terms of time and money.

As I said, a firm grasp on the basics should provide enough background. Thus, we get "simple" economies in games based off of systems like the faucent and drain model. Really, I think this system works just fine for most game worlds *as long as you have a human administrator overseeing it*. Trying to come up with a self-balancing system through code is almost always going to end in crying (if you're lucky, bugs and exploits if you aren't). A human can step in and take necessary measure to rebalance the system quicker and better than any algorithm. Too little money in the system? Increase money drops from monsters. Too much? Lower money drops, and maybe increase item drops to compensate; now players have to choose what loot they want to bring back to sell for cash, or what they want to put into their vaults. WAY too much cash in the system? Quick! Schedule an event where you auction off some rare items in the game!

My thoughts and experiences,


Well, would it sound like bragging to mention that the Camelot economy actually worked as designed? Yes, it was a faucet->drain economy on a large scale, but it worked as it was intended to. If it had a fault, it was that it put people on too much of a forced upgrade cycle, but that was more of an effect of the underlying game systems I had to deal with than anything I designed in. It even managed to hold itself together when the major high-end drain, Spellcrafting, got delayed until 9 months after release.

Closed economies don't work. Not surprising, the real economy isn't really closed, but depends on the fact that rises in prices lead to increases in supply, increased supply leads to wasteful consumption. The real world is a faucet-drain economy.

Anyway, I think I can solve the floating-prices problem, it really comes down to "building in" moderators on supply and demand so that no matter what the players do, prices can neither climb nor drop excessively.

Will it work? I think so, but I'm so far outside of the normal bounds of economic theory, it's hard to predict. But I'm hanging a lot of gameplay hooks off that line, so it had better work.



Question 1 and 2 are deeply related, and together they suggest something that is profound (to me anyway).

1. Are the simplistic models that cause grief?
2. If they do, does it matter?

"Grief" can mean on of two things: "isn't fun" or "not working efficiently according to the standards we apply to the real-world economy." Dave and Brian point out that, from a game developer standpoint, interpreting grief in the second sense is pointless. Only the first sense matters. And guess what, all of economics is devoted to analyzing economic behavior from the second standpoint. Which means that there are no economists in the world who are in a position to advise these practicioners about all this.

In other words, to my knowledge, there are no specialists capable of designing economies that are fun.

Actually, if there are any, they are all in the game community, and what is breathtaking about this is the possibility that fun may actually be the thing we SHOULD be shooting for in the real economy. In that case, we should have game designers sitting in the Council of Economic Advisors.

Example: economists say "Inflation is bad. If we could get by with zero inflation, we should. However, it seems that a little bit of inflation oils the gears of production a bit [for reasons we have never understood, I might add. E.C.], so we can have a bit of inflation." Now along come MUDs. Every durn one of them has MUDflation out the wazoo. Currencies that trade at millions to the dollar. Newbies grunting under the weight of the uber gear that somebody just gave them. People fully capable of paying hundreds of thousands of gold pieces for a house or a horse. And guess what, Alan G.? IT'S FUN!

Dave is right, there are no lessons out there that have been learned. Heck, we haven't really come to grips with the what the objective of economic policy is. We're in the stage of dumping the red goo into the blue goo and hoping it doesn't explode.

So, to ask whether there are broken systems, well, if the game sells, the system isn't broken. But does it matter? I would say, yes. The economy really is a place where people can have fun. It's a motivator to get people talking face to face. It's a source of self-esteem - look how proud Julian is about his hoard.

So I am really interested in the Wish economy. What's wierd is, the Alanator should be too.

PS. I've had this thought for a while - import/export merchants could be used as a lore-consistent moderator for price swings. Set up exotic-looking folks by boats at the edges of the world, who suddenly get new shipments of a good from abroad when its price gets too high, and just as asuddenly get a hankering to buy up the excess of a good when the price gets too low.

PPS. That thing in UO about the guy with 10,000 shirts. I wonder why there aren't more storage fees imposed. I don't mean for newbies, but for these people with massive hoards. Like Julian. Make 'em pay (implicitly) for storing those bits in the database; they are chewing up resources.


"Anyway, I think I can solve the floating-prices problem, it really comes down to "building in" moderators on supply and demand so that no matter what the players do, prices can neither climb nor drop excessively."

I'm intrigued... I understand you can control the inflow of resources as well as the consumption rate (breaking, using up, waste, etc) of basic resources based on the prices you see those resources take. But how would you control the prices of player-crafted items in player-to-player trades??


So the SWG economy is an interesting beast to compare to UO's.

UO's has been examined exhaustively and in detail. One thing I'd comment about the UO economy is that I DON'T think it is proof that a closed economy cannot work; if modelled in its simplest form, one can think of it as "if you cap the size of the basin [eg, player hoarding], then pour in enough water, you will always have the excess that does not fit in the basin circulating." I also suspect that selective exclusion of resources from the closed loop would also make a world of difference (eg, renewable resources).

But that's neither here nor there. SWG uses a faucet/drain but with the interesting wrinkle that the only game-set prices in the entire thing are for a few drains, and a minor (and statistically insignificant, at the moment) source of income from junk dealer shopkeepers. All goods worth selling are made by players; all purchasers worth selling to are players.

The market that resulted has been remarkably stable, with only the expected oscillations. The currency is in a spiral of losing value, but a large amount of the playerbase lives hand to mouth, and the value of goods has simply adapted. There also have been no manifestations of the typical 'sploits with buying low and selling high reliably to NPCs, etc.

Basically, we got a faucet/drain economy like the classical mud model, but with LESS knobs to twiddle. It's actually sort of refreshing. That said, we have, as usual, far too much money coming into the system and not enough exiting, so there's mudflation going on.


Yeah I think SWG has really done the best job of capitalizing on the division of labor / specialization / gains from trade ideas of Smith (1776). Every player has a niche role in the economy. It was cool to run around as a newbie scout, harvesting these tiny quantities of bone and hide, throwing them up on the bazaar, and watching them actually sell. I kind of couldn't believe it at first.

And you know, there were plenty of zero-worth items, like your newbie weapon. I admire the fact that SWG calls junk "junk" - you really can't turn that stuff into gold.

Regarding MUDflation - you either have to make it harder to mint new money, or you have to have as many NPCs sucking money out as as there are NPCs pushing money in.


The discussion has, as always, been illuminating for me. I wonder about the conclusions of many which seem to be summed up by Brian "Psychochild" Green: "I think a designer can get a "good enough" economy without being an expert in the area."

I'm sure this is right for standalone virtual economies. Faucet-drains work fine. When theey go out of control a tad, just tweak the settings, nerf some objects, and accept that mudflation is a fact of life.

But what happens when these economies are tied to real world markets, as in There or even as seen recently in the SWG credit deflation on eBay spiral (see http://terranova.blogs.com/terra_nova/2003/10/swg_300_and_plu.html). I think the latter one isn't going to trouble Raph and SOE much. Their attitude is going to run something like: "if you are ebaying then we hate your mangy guts and aren't going to do anything in our economic model to help you out."

But what happens if There suffers rampant mudflation, and my REAL US DOLLAR which I've turned into 1787 Therebucks suddenly is worth less than a kopek? Surely this is a problem for There or any other would-be metaverse owner?


Does There both buy and sell at that rate?


Last time I checked they were only selling. That is, they don't directly provide any means of translating Therebucks back into USD. You buy the Levis, they're yours and that's the end of it.

In the long run, of course, this won't work. Imagine a situation where I'm unhappy with my Levi's or the game I just played or whatever. There (or their third party provider) will have to provide some exchange mechanism. And this is when mudflation really bites you in the arse.


I think the answer must be no, Ted, for several reasons. Bruce Boston... paging Bruce Boston...


I'm just off to class, so I can't post now. But in case Bruce doesn't answer Greg's page, people might like to have a look at Bruce's $US<->$T arbitrage site called First Bank of There: http://www.firstbankofthere.com/

You might also like to have a look at a competitor, Tbux: http://www.tbux.com/home.php


"I'm intrigued... I understand you can control the inflow of resources as well as the consumption rate (breaking, using up, waste, etc) of basic resources based on the prices you see those resources take. But how would you control the prices of player-crafted items in player-to-player trades??"

I'm having a busy day, so I'll have to make this brief. Forgive me if I gloss over too much.

The core problem with supply-demand driven game economies has come from two sources: Singularities for supplies with no demand (and hence prices that trended towards 0) or demands with no reliable supply (and hence yielding hoarding behaviour); and NPC prices that were recalculated at discrete time intervals. Players could manipulate the system by buying low-volume goods in one time increment, driving up their price in the next time increment, at which point they would sell, driving the price back down in the following increment.

Wish addresses both, first by having an assumed NPC supply and demand for each settlement, to which the player-->NPC transactions are added to get price multipliers. Commodity materials come into the world first as bulky, "raw" forms, which must be sold to an NPC to be converted to their refined, compact versions (and a 10% overhead plus House taxes are extracted from these transactions). Theoretically they could be transported and sold elsewhere in their raw forms, but the weight differentials run 3-20 to 1, and pack animals will require fodder (using players to do the packing will not, but they have comparatively low encumbrance limitations).

The materials required to make finished goods is implicit in their object definitions, so the NPC price for a finished item is taken from the commodities it is assembled from (killing two birds with one stone, since this also gives us crafting recipes for every item).

The House does not tax Player --> Player transactions, it seemed counter-productive to try since a variety of means could be used to avoid the tax (such as leaving the town limits). But Player --> NPC transactions are all taxed, controlled by the simple measure of simply not placing any NPC vendors outside of towns.

This can (and is intended to) lead to significantly different base values in different towns, opening up the possibility of making money simply by moving materials. Towns where a material is expensive will be profitable destinations for traders from towns where it is cheap. Although every town has an incentive to encourage cheap materials (so they can make equipment cheaply), the only way to ensure this would be to not allow anyone from outside the House to buy materials (and prices being low, there would be little incentive to import). So to keep their prices low, they'd have to gut their revenue.

It's going to get considerably more complicated than that in practice, I'm sure, but I've been working on this post in snatches all afternoon in between doing other things, and have to go.



"The currency is in a spiral of losing value, but a large amount of the playerbase lives hand to mouth, and the value of goods has simply adapted."

I obviously don't have access to your data, but in once sense the currency has been stable and gaining value. Specifically, the credits / unit of resource has been on a steady decline since the launch. Again, this is only from my own observation of the Bazar behaviour, but I have been closely following it. I would have expected if credits were becoming easier to gain, people would be willing to pay higher prices for resources. Further, if credits were becoming easier to gain through adventuring, said adventurers would be less likely to bother selling their resources on the bazar, so would demand higher prices to make it worth their while.

I guess you can argue that this is a result of Mudflation on both credits AND resources. However, as they have remain relatively tightly linked (I've seen them go from 5cr/unit -> 3 cr/unit, with a large drop at Labour day) I'd say it definitely proves that you have created a stable economy.

And, I can definitely provide anecdotal evidence of most players living hand to mouth. It is impressively how easily credits arrive, and how easily they disappear :>

The EBay crash of SWG credits I personally consider meaningless at this point - it is inevitable that at the early stages of the game there will be an overvaluation of the in game currency from people eager to jump ahead. I'd wait a year before considering that a sign of MUDflation.

- Brask Mumei



There doesn't buy back currency but it has been suggested several times, by myself more than a dozen.

But for now there is a very strong secondary market for both buying and selling Therebucks. The main two as referenced by Dan FirstBankofThere.com and Tbux.com.

Tbux basically specializes in selling Therebucks that player developers have earned by selling clothing and vehicle designs. The FBT both buys and re-sells therebucks from developers that would rather not sell them themselves or that want a quick payout.

Its still very early in the progression of the economy, but I think they way the currency is designed we should have a very stable currency. In fact, our business plan depends on it.

One of the things that keeps the currency strong is the fact that the main way to get currency is to buy it. There sells it for 1787/US$. The secondary sites sell it a bit cheaper, but are basically selling recycled currency that was once purchased by someone. Many members are buying from the secondary sources as any discount is a discount. Buying currency from There at T$1787 or buying it from Tbux at T$2000/US$ is an easy choice. And because Tbux and the FBT can eventually sell all they may have at 1900-2200/US$ there isn't much of an advantage to sell it cheaper except to get US$ faster. Even then we just run out of currency faster and eventually the exchange rate returns to 1787 and members resume buying from There.

My hope would be that we see the currency hover between 1787 and 2200, 90% of the time, with the exchange rate returning to 1787 every 30-60 days.


At the same time while the currency may be set-up will some built-in fail-safes, we are in no way exempt from MUDflation. If fact, we already have good evidence that it exists. If I understand MUDflation, its a mix of low-end deflation and high-end inflation.

We have a very simple skill reward system that gives out little gifts. Because so many members get these, in many instances it doesn't mater how cool they may look, these items have almost no currency value.

At the same time Player Developers have been creating 'limited edition' items and some very nice ones at that. Some of their designs are going for upwards of T$50k-T$80k or about US$30-US$45.

So I think our string of MUDflation will be somewhat 'item centric'.


"The materials required to make finished goods is implicit in their object definitions, so the NPC price for a finished item is taken from the commodities it is assembled from (killing two birds with one stone, since this also gives us crafting recipes for every item)."

When it comes to players they do calculate with a good degree of accurracy the cost of materials used in obtaining an item (including failure rates, quality, etc), but they also add a "profit" factor to account for the relative scarcity of their trade skill, the demand/quality of the items they're crafting and the total time it took them to gather materials as well as put the item itself together...

So while the NPC market you describe has complex (compared to other games) logic behind it, is it supposed to affect directly (in some way I missed) the player-to-player market for player-crafted items? Or does it affect it solely by raising and lowering the supply and demand of basic resources?


It's all indirect, my experience with trying to excercise direct controls on player --> player transaction is that it never works out the way you planned. But I got some really interesting and subtle effects with the way Camelot handled driving PC prices through raw material costs, and I think the different prices in different locations is going to add some interesting wrinkles.

Partly, I think our markets in these games have been *too* efficient, their efficiency and predictability were coded in and much of the fun that comes from the subtleties of trying to outguess and outwit the inefficiencies was lost. The "merchant game" has been strictly an excercise in marketing and economies of scale.




"It's all indirect, my experience with trying to excercise direct controls on player --> player transaction is that it never works out the way you planned."

Agreed 100%. I did want to hear what your thoughts were on this since at first it looked strange.


I realize I am months behind on the activity of the discussion on in-game UO Economics but I do want to give a comment.

Back when UO first came out, the most intriguing thing to me was the Economy of the game and how the natural resources were tied to it. I had the unfortunate experience of playing during the scarce months...were if you had 100 feathers they were fetching hundreds of gold...Of course there seems to be lots of discussions already as to why and most point to the fact that having a closed-economy won't work in-game that the faucet-drain approach works better. However, I'm not certain I agree.

Take the US Economy for example. True its not exactly "closed", the federal government manipulates money supplies as they see fit, increase lower interest rates, its a world economy and many other factors. But there are a couple of things that occur in the US economy which I think are the most important in terms of money flow, which seems to be lacking with an in-game economy. I believe (at least I would like to open it for discussion) that the following approach will work to keep a closed in-game economy flowing.

The general idea stems from the current way the US Banking system, Mortgage System, and other such programs work. The current US Economy would go into a standstill without them. Here's how it works. Lets say many people start taking on a mortgage. Now more and more start taking on mortgage, sooner or later all the money in the US would get tied up in mortgages, there is none or little money to go around. That’s a problem. What fixes this in our economy? Mortgages are "wrapped up" behind the scenes and sold back to institution (Charles Schwab, Fidelity, Morgan Stanley, Banks, etc) and they become investment instruments. People buy them, money goes back to the lending companies so they can give out more mortgages. Confused? Good. If you are, just know that ultimately your mortgage will get turned into someone’s investment (like a bond). You pay your mortgage payment, in many cases it goes to your mortgage company who pays one of the government sponsored agencies who pays the institutions they sold the "wrapped-up" mortgages to (called Mortgage-Backed Securities) who pay the Security holders (which can be individuals, companies or whoever).

The money you have in your bank account works in a similar way. The bank is only required to keep 10% of its deposits in the bank. The rest of the money goes back out into the hands of people in the form of loans.

Now what if, an in-game economy used this same idea but all the way down to the resource level. Sticking with the UO example, I could put anything in the bank; gold, shears, shirts, plate armor, reagents. Sometimes I would have 5 sets of plate armor sitting in the bank. Which is taking up precious real world resources, say 50,000 iron ingots (i forgot the actual approximation). Or the guy who has 10,000 shirts, he's taking up 100,000 balls of yarn, which is taking up bunches of wool. Or the people hoarding 1,000,000 black pearls. If these items sitting in the bank "lent back" resources to the world just as your money does in the real world bank, I believe a balance could be achieved in having a "closed economy".

True its not "realistic" that 1 Plate Mail Armor in the bank could give back 10,000 iron ore to the mountains west of Britain, but it does model the real world cycle of money (resources). Your $1,000 real world deposit can become a piece of a loan to small business selling chairs, which you buy. Done properly you would never have a problem making a withdrawal of your 1 Plate Armor or 10,000 shirts just as you don't notice when you go to the real world bank and ask for $1,000 in cash, they just give it to you.

It works in the real world (at least so far) so why not extend the model into an in-game economy? Thoughts? Arguments? Constructive criticism is appreciated.

The comments to this entry are closed.