« Patent Arcade / IGDA SIG | Main | Chinese Fatigue Regulations »

Aug 23, 2005



At the very least it seems to be a useful mechanism to resolve cash distribution problems in long-lived worlds. I believe Eve-Online had problems with players at certain periods of the game's evolution (earlier) accumulating large cash reserves. This issue is not a unique problem. How to smooth away cash holdings imbalances over time once the actual resource "drop" imbalances have been addressed?


It's quite simple - inflation is the natural result of economic expansion. Since it is well known that it is much better to live during an economic boom than a bust, games that are fun will inevitably cause inflation over time.

One way to offset this inflation is a rapid growth in population. Since the rate of population growth for most games drops off significantly after a certain period, this stops working, resulting in the trends you will see in the charts.


In a game with unlimited content, inflation is bound to occur. Players aquire items through killing mobs, making items or by doing quests (at least in the majority of this and next generation games). For the killing group, mobs drop cash, items or nothing, but almost always give exp. In order to curb inflation, mobs would suddenly have to stop dropping items/coin, merchants would have to stop buying items when inflation gets too high, or mobs would have to be numbered and not unlimited. After a while, killing mobs becomes much less fun when you aren't rewarded with anything except experience.

Less dropped items/coin would also make crafted items more expensive, since resources are more scarce, demand is higher and theres less money out there from killing mobs.

Quests can just give less money or less monotary rewards.

In any case, the amount of coin going into the system must equal the amount going out. Implementations such as attuned/no drop items or item decay IMHO reduces the replayability of games and adds unnecessary frustration to the populace.

So my roundabout answer is "yes, it is possible" but it simply hasn't been done yet. I believe an advancement system based on character development (as in skills) rather than item aquisition is the first step, because then all the item and coin talk isn't so much a priority.


Talk of MUDflation always seems to leave out the question of data. There's just not enough information available for anyone to be able to make informed decisions (not even game companies, I'd wager). Population growth, money supply, some kind of CPI -- all this needs to be more transparent if we're to draw any real lessons.

Anyway, calling it inflation isn't completely accurate, it seems to me. What's represented in those charts is that game currencies are being devalued against the dollar. If we take a game economy to be a separate economy, like that of a real-world nation, inflation would imply that the prices of (in-world) goods and services were going up. Is that happening? The charts don't take in this information, so we can't tell from here.

It would be interesting to develop some kind of in-world CPI for various virtual worlds. In most places, of course, prices aren't as free to fluctuate as they are in the real world. But having that metric would help isolate what drives down the value of game currencies.

At this point, there's no way to tell what drives that phenomenon. The flood of people into a virtual world probably helps, but it could also be adjustments to the economy that we don't get to see, including some mentioned above, like increased loot drops. Content and engine patches could also have a significant impact on "productivity." If your fireball hits for 10 percent more damage in the new patch, that's a not insignificant improvement in productivity, in virtual economic terms, that would no doubt have an impact on the money supply. In a relatively free economy, that would drive up prices (inflation). In an economy where prices are more or less fixed, it would drive down the currency's exchange value (devaluation). Even a CPI in dollar terms of goods and services sold on eBay and elsewhere would be useful. That's a lot of time for someone to spend putting numbers together, though. In any case, I think it will be even more time (never?) before game companies start giving out the kind of useful data that would clear up the picture.


Inflation would be disregarded as 'no big deal, really' if it were not for one thing. The nature of inflation is that your long-time residents have more and more money, and open market rise to reflect that reality. Meanwhile, any newbie you actually convince to come and try out the game are still getting the lowly monster drops everyone else got on day one. Meaning that newbies are priced out of the auction market, and face a long, uphill climb in order to get to a place where they can compete.

If someone came up with a good solution for this Inflationary Moat (tm), I think most other problems related to MUDflation are either solvable or ignorable.


I think the problem, if it really is one, has a pretty obvious cause. Everyone who plays a MOG is essentially their own mint, and they can print up as much currency as they want to by bashing orcs.

What would be more interesting, for me at least, is a chart of the value of in-game items in USD. We know that as the amount of gold goes up inevitably supply will outpace demand and RMT prices will fall. Items in-game which have a fixed value, such as mounts in WoW, inevitably will devalue in terms of USD as time goes on. What about items on the auction house however?


I would think that without some drastic economic drains (extremely expensive items that decay rapidly, taxes, and the like), longtime players will stockpile funds because they a) are at the endgame when loot per kill is maximized, b) have fewer things to spend money on, and c) their high level characters never expire. In such a scenario, players cannot help but collect money.

To manage the economy 'properly' would require some combination of greater expenses or character deaths, to get in-game cash out of circulation. As the previous poster mentioned, the Inflationary Moat is both a hurdle and deterrent to newbies, and contributes to the feeling that MMOs have a definate life cycle - a hyperinflated auction market feels analogous to a red giant star, in my mind.


Actually, this is a slight perversion of the term "MUDflation". This originally referred to the fact that items kept growing in power as things were added to the game. This was a big problem because when new wizards added to the game, they wanted to have the biggest and the best items in the game so that people would visit their new areas.

As for the inflation of currency, Damion hit the nail on the head. Inflation means that newbies get the short end of the stick. Their purchasing power becomes insignificant, and the game becomes much harder. Auction halls in newer games make this all the more apparent. Stuff I used to be able to get for just under 1 gold now costs several gold as people are charging more for items WoW. High level characters are paying lots of money to outfit their twinks, making a character without a sugar daddy a frustrating thing to be.

It's interesting to note that MUDflation actually does the opposite in terms of how it affects newbies, because as the equipment gets more powerful the lesser equipment everyone prized in the early days becomes cheap castoff gear for newbies to buy. This makes the game a bit easier for them, although they still have to work to catch up with the high levels.

But, is inflation fun? Anecdotally, I know that curbing inflation isn't that much fun. I've worked to stabilize the value of currency in Meridian 59. I increased the cost of casting the more common and popular spells and adjusted the drop rates of certain reagents. I also added in universal controls to regulate drops from monsters. These changes weren't met with overwhelming excitement, especially when they've been used to hot-and-cold running currency for most of the history of the game. :) So, if you have a game with a "broken" economy, fixing it is not necessarily going to be widely accepted as a good thing by the players. Would a well-balanced economy from the start be fun? Is such a beast even vaguely possible? Interesting to ponder.

Some thoughts,


As for the inflation of currency, Damion hit the nail on the head. Inflation means that newbies get the short end of the stick. Their purchasing power becomes insignificant, and the game becomes much harder.

Wouldn't this only be true if we're talking about a special case - where wages (say "drops") don't keep up with price increases? If we're talking a more benign form, where price increases are more or less predictable, or cost increases can be easily passed on and built into the wages of the world ("cost of living"), then we're talking a different fish.


I don't think 'fun' in itself is a useful description, it's too coarse-grained - a fallacy of aggregation.

Inflation is fun when you're doing it - whee, I'm spending this much money and Not Caring, isn't this great? I'm getting the really cool loot (or crafting the really cool loot) and selling it for huge piles of gold, isn't this great? - but it's no fun when everyone's doing it and you have to keep up.

5 gold is the new 2 gold, 10k is the new 2k, &c., &c.

One thing that I've observed on the games I've played is that there doesn't seem to be a steady progression from "poor newlyn" to "rolling in end-game riches", but rather there's an inflexion in the curve - a point where it it switches over from struggling to get anywhere to having more than you actually need.

Perhaps one way to deal with it is in information management rather than Stuff management, on the basis that people only want what they know about, and if they can't see the edges then they won't want all of it. But this might involve explicitly chunking "the economy" into connected smaller economies. How many games are there out there with multiple currencies and restricted interconversion?

I know I'd find dealing with the additional level of complexity interesting, because it would (it does on my main MUD) provide lots of extra opportunities to use player skills and generally mess around to everyone's benefit.

If we were to design an economy for fun, how would we measure it? Time spent involving oneself with it isn't a metric if it's the only game in town. Depth of involvement in it, perhaps, by checking the number and variety of interactions, or travel distances of various marker items (this iron ingot was salvaged from armour dropped by a goblin in Alcatraz, made into a sword by Bob, sold to Charlie, looted from his corpse by the evil Diana, salvaged and re-forged by Edward, and ends up nailed to a castle wall in Darkfall).


By the way, RMT makes the inflation worse. When a rich high level character stockpiles money, it goes out of circulation, lowering market prices. When that gold is sold to a newbie for $$$, it re-enters markets and drives prices up again.

RMT prevents designers from using player end-game boredom as a money sink. Even if players quit, they don't take their riches away with them. They sell them back into the system, raising prices. Thus goldfarming begets its own demand and its own supply, in an upward spiraling loop: as prices rise, newbies must either buy from farmers or goldfarm themselves, in order to participate in the economy at all. And as each newbie switches from gameplay to money-grinding, prices rise still more, and the game gradually changes from an adventurous exploration into a tedious money treadmill.


newbies are priced out of the auction market, and face a long, uphill climb in order to get to a place where they can compete.

I can think of one way to alleviate this, that I can see in my current game (Runescape) - and I think, to a lesser extent in WoW. If a new player can collect resources that are in demand from the long-term, higher level 'rich' characters, they'll be able to sell for a lot more cash than they'd probably be able to get killing mobs.

In WoW, it's primarily from 60's twinking their alts, or changing professions. In Runescape, the raw materials are directly desired by the high levels. They are buying experience points... since the raw materials almost always cost more then the end product. Of course, this can leave the mid-level players in a bit of a quandry.... not enough cash to buy lots of materials, but not hurting for cash enough to lose the experience that selling gathered materials would entail. At least when you really need that cash to buy some wanted item... there's a way to get it.

The gathered materials themselves may bring more cash into the system (when the finished goods are sold to an npc, say)... but primarily this shuffles existing cash from the haves to the have-nots. As long it gives experience, it probably doesn't matter if it has any value set by the game or not.

Never go on an adventure without a hat!


Damion Schubert wrote:

The nature of inflation is that your long-time residents have more and more money, and open market rise to reflect that reality. Meanwhile, any newbie you actually convince to come and try out the game are still getting the lowly monster drops everyone else got on day one. Meaning that newbies are priced out of the auction market, and face a long, uphill climb in order to get to a place where they can compete.

So design the game so that players don't have to purchase things at auction from other players to compete. Fixed price NPC vendors for anything remotely 'necessary.'



What I see in the graphs is what I have termed the summer effect. Albeit, I only ever tracked UO Gold, but the over all effects look the same. From jan to the end of summer, you see gradual decline, over the next 6 months we'll see it start to climb a bit, but never really reach the peak from the prior year, atleast, not conistantly. It seems there are more players in the winter months than the summer months.

So, stock up on the currency while it is cheap :) jk


Buy online currency futures now, Rich, and short them in the Spring? ;)


Maybe that's part of the teaser for your book, but I'm intrigued by the idea that designing a non-inflationary MUD economy is "eminently doable". I guess there's two ways of asking the question. "Is inflation fun?" questions whether it's something we'd like for its own sake, versus "Is inflation worth it?" which questions whether we'd have to sacrifice too much in other areas to get rid of inflation.

I do think there are some things essential to "standard" MUD design which lead to devaluing stuff and/or currency. There's kinda economic growth, which is standard because everyone wants to get their character richer. But there's not many channels for this economic growth, so once the game has been well explored that growth can only take the form of "more of the same". In the real world, we largely look to "new and different" for economic growth, but "new and different" isn't something players have much control over in most games. To put it in beardy terms, a great deal of economic growth in games is exogenous (comes from the outside via game developers adding content) while in RL it's all endogenous (comes from the inside via people innovating).

There's very good reasons to keep players from serious innovation, since the possibility of real innovation includes the possibility of ruining the game for everyone else. And it doesn't seem like people want to give up on the (play game)->(get richer) model of economic growth. Those two things working together are bound to cause some amount of inflation (or devaluation, depending on perspective).

I'm actually curious how much this applies to games with a large PvP element, however. I can see such games acting very differently.


Nate Combs wrote:

> Wouldn't this only be true if we're talking about a special case - where wages (say "drops") don't keep up with price increases?

You've got that backwards. It's the drops that increase prices, primarily the higher level drops. This is a variation on what Raph terms the "hula-hoop problem" where players expect that they should be able to sell hula-hoops they craft for a profit, regardless of demand. Likewise, players expect that a level 50 monster is going to drop more goodies than a level 10 monster. If not, then they will simply bottom feed and kill level 10 monsters fast enough to make money. WoW saw this in the early days with the hippogriff feathers and broken electro-lanterns; farmers would hunt these monsters tirelessly even though they were lower level than the fighters because you could make a lot of money off the items they dropped.

So, there's the problem. The big issue is as Sam Kelly mentions above, that the increase in drops and decrease in having to buy "necessities" means that higher level players have a lot more money to spend past a certain point. You could increase the newbie drops, but that'll only encourage people to bottom feed. Increasing drops across the board just makes inflation worse. Lowering high-level drops also encourages bottom feeding. The best policy seems to be to put high-end money sinks into the game to entice players. However, that has a double-whammy: It can encourage people to buy currency to afford the big ticket items, essentially trading money for time. It can also just make the problem worse; once you buy your epic mount in WoW, for example, the only thing you can really do with the money is pour it into an alt. Or, spend stupid amounts of money for high end crafted stuff that gives relatively small marginal bonuses.


An alternative policy that never seems to appear is closing the economy. Meaning hoarding money causes drops to lessen. I'm not sure what effect that would have.


It seems there are more players in the winter months than the summer months.

Or, perhaps, it's supply that changes, not demand. Over summer, I would expect there to be more school aged players with free time to farm gold, and reason to want extra real-world spending money. What's more fun than getting a job over summer? Playing WoW and selling the gold!

No, I have absolutely no evidence to back this up. :)


If forced to guess, I'd say inflation is more common because:

1) maintaining price stability is hard, and
2) deflation is even less desirable than inflation from a player happiness POV.

There's also the matter of players using money as a marker for success. Because money is a collectible, players will collect it on the theory that whoever collects the most "wins." So they'll find the actions that generate the most cash for the least effort, and do them over and over again to put massive amounts of money into their personal inventories.

And then they do the virtual equivalent of stuffing it under a mattress, which (as noted) does funny things to an economy.

That "sitting on money" effect strikes me as one of the several economic problems virtual worlds currently face. For what it's worth, I've had a few ideas for attacking this problem of hoarding:

* A robust system player contract system would not only increase the velocity of money by letting players give each other work, it would allow a controlled form of "player-generated content."

* Create a system whereby all money disappears unless used -- set the drain rate according to whatever factors you like, perhaps on a sliding scale. (I'm not a big fan of this approach as it penalizes effort, but there might be some way to make it work.)

* Let There Be Consequences. OK, you can collect all the money you want... but the bigger your bank account, the more beggars and salesmen you'll attract; the more likely you'll be to have your pockets picked; the more likely it becomes that the kinds of people who prefer dark alleys and rooftops at night will find their way into your home and make off with the silverware. In other words, the in-game consequence of hoarding money will be attracting the attention of more people (PCs and NPCs) who want to take it away from you. The more you just sit on it, the more guards you'll have to hire... and they don't work for free... and how do you know they won't succumb to the temptation to stuff their codpieces full of your cash?

I like that last approach best (obviously). If you're going to let money be a collectible, at least let collecting massive amounts of it have some kind of interesting in-game consequences.



There's a different idea. I'm not familiar with graphical MMORPGs in general, but I'm guessing that newbies start out with a standard set of equipment (including money) from game launch until the game has a financial apocalypse.

Why not increase the overall value of this equipment so that, relatively speaking, it stays the same? So maybe the newbies during the first month of launch get rusty swords, but the newbies during the fifth year end up with broadswords +1, because they're common, cheap, and comparatively worthless.

Anyone give me some feedback on what they think would happen in that scenario?

(I call this the "We don't raise our kids on clubs and stone knives anymore, and fire's no longer our most precious commodity; we've got private jets and cell phones." phenomenon, because a long sentence like that is cool.)


Brian's right about what MUDflation means.

What I call "database deflation" (note the prefix) is the effect whereby all the effects noted above result in formerly nice items becoming first common, then junk, thereby rendering everything in the game easier, thereby forcing new content addition only at the top. This then results in "the game doesn't start until level X," a lack of accessibility to newbies, and so on. It's basically a process of content obsolescence. Merely doing spot fixes (such as fixing the newbie gear) doesn't help because it's a losing race versus the ravenous playerbase.

Timothy Dang--it's been ages since I saw that name. Where have you been?

Closed economy as Michael describes--yes, of course it's been tried. It succumbed to hoarding, of course. Play quickly grew unfun when no spawns appeared. :)


I still say someone should make a game in which there is only 10,000 gold coins in the entire game, and players have to learn to barter.

Maybe on holidays the "king" will mint some more coins to expand the currency-based economy.



It would be interesting to see an MMORPG with inflation, as it is none do. The costs of rent, transportation, NPC sold and purchased items all are constant; change them so they are a percent of the currency in the hands of the players and then you will start seeing inflation.
As it is you just have standard supply/demand and greed; and about the only fun thing about it the laugh you get from what people think they can acutally sell stuff for.

As for solutions to on how to keep it so that low level characters can acutally get stuff they need at a reasonable price EQ2 comes rather close to solving a huge portion of it. In EQ2 they break every 10 levels into tiers, so for most items if you are level 24 you only are going to need or use stuff in that tier. This has lead to a kind of bell curve where higher prices reflect where the majority of the players have thier levels. So if the majority of the players at in the 30s stuff in the <20 range is back to being fairly cheap, stuff in the 30s can range from cheap to expensive(based on supply/demand) and some stuff in the 40+ range is cheaper then the level 30s items they are replacing.


Erm, isn't MUDflation a process of deflation not inflation? Since the value of items in the economy go down faster than the value of money goes down?


Factory > Erm, isn't MUDflation a process of deflation not inflation? Since the value of items in the economy go down faster than the value of money goes down?

Yes, exactly. Inflation refers to prices, not exchange rates. The industry refers to the phenomena being discussed above as inflation because it's a convenient term that everyone's familiar with, but if you really want to design more robust economic systems, it's going to be important to get these things straight.

The fact that the value of game currency is dropping against the dollar doesn't mean prices are rising within a given game. Yes, this does happen on auction houses, but it's not described in the charts referenced. Inflation rarely happens with good sold by NPC vendors, except in games with more open economies like UO and Eve Online. Anecdotally, inflation happens, but until someone cooks up even a rough on-the-fly CPI for a game (or games), it's going to be hard or impossible to track this kind of thing.


What is the purpose of "money" in an MMO game?

It seems to me that there is very little discussion about this. We just start with the assumption (just like stats) that we "have to have money".

Not that we don't, necessarily, but I think it is important to understand what function it plays in our games--which is not necessarily at all the same function it plays in offline society (an abstract substitute, which has little or no intrinsic value, but acts as a convenient, time-and-place-shifting substitute for actual trade of valuable goods).

I think the assumption that money is money, leads to failed attempts to apply "real" economic theory to our games, where money may be something else entirely.

Why have money in a game, anyway?


It would be interesting to develop some kind of in-world CPI for various virtual worlds. In most places, of course, prices aren't as free to fluctuate as they are in the real world. But having that metric would help isolate what drives down the value of game currencies.

The closest thing we have is EQecon at http://www.eqecon.com/ (http://www.wowecon.com/ is still in the works and is just a plugin right now).

EQecon which has been talked about a bit here before (http://terranova.blogs.com/terra_nova/2004/02/more_economudtr.html) tracks the prices of almost every item in EverQuest that goes up for sale in the bazaar. This way you can see "inflation" by both server-specific and an average throughout the game as a whole.

You could then create data from the average price increase in all items over the last six months and come up with a decent figure for inflation?

You could even pick out key items in EverQuest and create a sort of Consumer Price Index (for such items as +stat increasing food?)

Or just track the price changes of precious gems/platinum bars in the game?

Then again some changes will be due to the fact that the game world is changing and just as our technology becomes outdated, MMORPG items become outdated (and almost as fast as our technology). The Top Items/Gear in EverQuest 4 years ago are a joke today, just the same as carrying around a 3 pound cell phone.

The exchange rate of Gold/USD is a whole different issue. It is controlled by such factors as competition and the supply from Chinese “shops” and how cheaply they are willing to sell to you. In fact IGE’s default “buy” system on their website for regular players and not major suppliers is dynamic; it changes the prices on you depending on how much stock they have (prices which are still ridiculously low no matter what). Then there are such factors as a major banning that wipe out both currency supplier and resellers inventories causing a spike in prices (although have not seen this happen as often as it use to, must be IGE’s new Public Relations firm the Bohle Company :) http://www.bohle.com/clients.html who has such clients as IGN, Alienware, and Gravity Interactive’s U.S division). You may also remember the days of when Yantis and IGE were fierce competitors and went through a series of price wars bringing down the EQ1 Plat/USD exchange rate.

Now a day you have a number of Chinese shops who are selling directly to the consumer for highly discounted prices that would really throw off any Gold/USD figures. Just run a google search on World of Warcraft gold and see the range of prices you come up with along with the poor English on the sites with the lowest prices many of the times. To compose a more accurate index you would need to get quotes on the prices from ten to twenty of the top currency suppliers who supply the resellers (and even those prices change depending on personal relations and bulk purchases). Firms such as IGE have huge markups using their bulk and size to buy large amounts at cheap prices and using their dominance and name brand to sell at some of the highest prices in the market.

Also found an interesting (although old) powerpoint on the UO economy (www.mine-control.com/zack/uoecon/uoecon.ppt ) and how the original system of a closed economy where each server had a set amount of each resource in a “bank” and those resources would flow into the world from the bank, and as the items decayed, broke, or were sold they went back into the bank to be distributed out again. This was suppose to prevent inflation through conservation of resources but was in fact foiled by players hording resources (which Raph has talked about before).


galiel> Why have money in a game, anyway?

I think you said it yourself -- exchanging monetary tokens is more convenient/safer than lugging around actual goods for barter. (Assuming you're not a Yap Islander.)

Virtual world money has the same kinds of effects on human action (within the virtual world) as real world money does on real world activity because both are systems for achieving the same purpose. A collective agreement to treat tokens as objects of real value enables more economic activity. In other words, tokenizing wealth allows a greater rate of exchange of wealth.

That's useful in a virtual world because it adds economic activity to other kinds of activities, and expanding opportunities for action makes the world more interesting (and if it's a game, more fun). Yes, some people abuse enhanced economic abilities. That damages the value of economic gameplay, but IMO not enough to make such abilities not worth having. (I'd say the same is true in the real world.)

I think designers build standard economic systems for a more personal reason as well: control. Common economic systems have commonly understood emergent behavioral effects (supply & demand in particular), which means that virtual worlds can be designed to apply those effects to achieve desired goals. So economic design offers a way to control (or "guide," to view this more charitably) the behavior of players. The alternative is to design a world without some equally well understood means of regulating player behavior. That might be interesting... but it's also scary.

All this said, it's not wrong to question conventional wisdom on economic systems in MMOGs. Along with their benefits, all economic systems have pathologies, but maybe there's an alternative that would work as well as a token-based economic system with fewer of the disadvantages. But so far, the advantages of a standard system based on money seem to outweigh the disadvantages, so I expect we'll keep seeing changes rung on that approach for some time to come.

I'd enjoy seeing some alternatives tried, though. Interesting failures can be useful.

Note: What I find really entertaining is the idea that money in the real world (in societies that have adopted electronic banking) has to a certain extent become as "real" as plat or credits in a virtual world -- your cash wealth now exists only as ones and zeroes on a computer. Maybe the real world is becoming more like some virtual worlds...?



exchanging monetary tokens is more convenient/safer than lugging around actual goods for barter.

Why? Is this an inherent limitation, or is it just the result of other real-world conventions we have brought into the virtual world--or, the opposite, artificial game design constraints we have imposed to address one need that conflict with another?

Virtual world money has the same kinds of effects on human action (within the virtual world) as real world money does on real world activity because both are systems for achieving the same purpose.

It is interesting to me how existing conventions in virtual worlds are often defended as working "just the same as in the real world", at the same time that proposed innovations based in real-world research are often attacked as "not applicable to virtual worlds".

This is definitely NOT meant to be a personal comment, Bart, but just a general observation of the broader ongoing discussion.

I'd enjoy seeing some alternatives tried, though. Interesting failures can be useful.

I could respond "you mean like current MMO economies?", or I could respond that presuming failure is a rather reactionary and stodgy way of approaching innovation, seemingly intended to thwart it and preserve the status quo, but I won't.

I'll just say that the point is not to push alternatives, necessarily or solely, the point is to examine our assumptions.

And, finally, that the alternative is not necessarily a system not based on money, but perhaps an alternative approach to managing money within the systems we design.

Edward De Bono talks about one approach to problem-solving that involves entertaining an absurdity or counter-intuitive notion in order to knock loose innovations blocked by our assumptions.

I can't find the specific example, but he talks about a real-life case where a (I think candy) product was breaking when it fell from one conveyor belt through a chute to another, because of impacts against the walls of the chute. After trying and failing repeatedly to reduce the collisions, they took a seemingly absurdist approach: what if we *increased* the collisions with the wall? They came up with a series of baffles within the chute, against which the candy hit as it fell--each fall too small to create damage, and each hit taking some energy from the fall.

What if, for example, instead of viewing database inflation as a given we have to deal with, we instead tried to think of ways to design it out of the system? I don't know if we would succeed entirely, but we might end up with an improvement that reduced maintenance by, say, 15% per year. Would that not be a worthy exercise?


Raph> Timothy Dang--it's been ages since I saw that name. Where have you been?

Lurking, here and MUD-Dev, playing a few games, distracted by a very distracting now-2-year-old. My interest in MUDs waned some because my efforts to cajole economic data were unsuccesful. Now that Prof. Castronova is demonstrating that there may be academic prospects, I'm putting MUD economics back on the table as I spend the next month or two threshing out my dissertation topic.


Why? Is this an inherent limitation, or is it just the result of other real-world conventions we have brought into the virtual world--or, the opposite, artificial game design constraints we have imposed to address one need that conflict with another?

It's not an inherent limitation, just like the dragon you fight is not inherently a dragon. Sure, we could put a box there, and you use a number of different codes to try to break it, with each one reducing the overall difficulty on the box. Oh, and the box is called "dragon". But if you put in this... you end up putting in that...

Money appears in any system with actual economics; it's a matter of time. The notable case was ATITD, where a barter system was in place, but one player decided to print money.

Remember that money has THREE purposes: (1) medium of exchange, (2) store of value, and (3) unit of account. It's less meaningful for two people to compare with by saying, "I have a thousand sheep" and "I have a thousand cattle", when one likes lamb and the other likes beef. It's difficult to say, "I'll give you this sword for a dozen gems." It's hardly impossible; indeed, it's not as if barter isn't used. But where you have goods and services, money eventually emerges because it's more convenient as a medium of exchange.

The difference is that developers typically make the automatic assumption of using fiat money, enforced by code; ATITD's case was a form of commodity money.

This information was brought to you toda-- night by Wikipedia. (Yay!) http://en.wikipedia.org/wiki/Money



What we're really discussing here is the 'value' of these things to the players and how inflation effects that. In economics, there is an old argument about whether the true value of something is the material put in it (which emphasizes the importance of capital) or whether true value is the labor (time) put in it.

Since economics is about scarcity, I think value is determined by whatever you have the least of at the time. If you have plenty of capital, you use it to leverage your time. If you have little capital, you put in extra time to use the capital efficiently.

Applying this to virtual worlds, a balanced economy should have elements that appeal to this entire range of needs. Gold farming is the primary way to convert time directly into money. The problem with that is macro-farmers who try to convert near-unlimited time into near-unlimited money. The result is rampant currency inflation. The worst effects of this is on those who farm gold (real players and bots) and on veteran players who horde gold. Players who engage in resource farming or who have a high percentage of fixed assets benefit from higher prices. In the real world, this is comparable to an economy that prints money constantly.

Resource farming in a game does not lead to currency inflation, but instead causes object inflation (possibly currency deflation). In real-world terms, this would be the accumulation of wealth by an economy. The those hurt by object inflation are resource farmers and players with large fixed assets. Those who gain are players who gold farm or hoard gold. In the real world, this is comparable to an economy that accumulates wealth.

Since it is the purpose of all economies to create wealth (& health), object inflation seems to benefit the general economy the most. In fact, it is a necessity, otherwise the economy cannot support a growing playerbase and players end up warring over the same 3 crystal swords forever (it might make for a harsh PvP game).

On the other hand, deflation can spiral downward easily. That's real world economies create minor currency inflation on purpose. As stated earlier, money needs to be store of wealth to be effective, so RW economies try to balance the downward pressure of object inflation by adding the upward pressure of currency inflation to ensure price stability. Too much object inflation causes cash hording and price deflation. Too much currency inflation, and cash is no longer a proper store of value.

Again, how do apply this lesson to MMORPGs? I think one reason why designers err on the side of currency inflation is to apply fiscal stimulus, especially for new players. I think newbs are far more likely to engage in PvE trading than veterans (who else is going to buy the 'broken club' they picked up?). Removal of that stimulus would cripple new players growth.

Balancing the inflation/deflation pressures is likely to be impossible in the presence of macro-farmers. They allow the supply of labor to be near-infinite, so the value of labor will eventually approach zero (though I think they'd quit first).

If the level of macro-farming can be controlled, then it's just a matter of balancing the inputs and outputs. I think a good design would be one that pushes money out through newb activities and resources out through veteran activities. However, it is difficult to fine-tune the level of money inflow since it depends on the ever-changing level of player activity. On the output side, creating large currency sinks for the veteran players can have a dramatic effect on the money supply. Last month's release of EVE-online had some currency sinks that absorbed probably 500 billion ISK in currency. Some of the more sensitive commodities have gone from a 10%/month inflation to 10% deflation. Doing this in a more measured form by controlling the number of sinking opportunities would allow a game to have fine-control of the money outflow.
A truly excellent design would do all this automatically, like putting in a rare drop or a currency sink opportunity every time 1000 gold is farmed.


The difference is that developers typically make the automatic assumption of using fiat money, enforced by code; ATITD's case was a form of commodity money.

Can you talk more about the relative advantages of each--that is, currency imposed by the developers, vs. naturally evolving currencies (both of which, I would argue, are always present in our games, no matter how hard developers try to eliminate the latter and control the former)?


I realize there have been few explicit answers to the questions Edward posed in his post (not least because I hijacked the thread to talk about the nature of money).

If we designed the real world economy to be fun, what would it look like?

I suspect the answer you are looking for is analytical, but the answers you will get will be political/ideological.

For example, the first thing that popped into my head was eliminating artificial scarcities. In the real world, we have economies around genuine scarcities, like energy; it is generally considered a desirable thing, by people across the political spectrum, to "solve" that problem and tap into essentially limitless sources of energy (whether your flavor is solar or hydrogen or fission isn't relevant here).

But, it seems to me, most of contemporary corporatist economics revolves around artificially manufactured scarcities. These serve the interests of corporations, abstractions that has little in common with the interests of the communities they operates within and affect (who generally do not own meaningful (as in, enough to sway decisions) shares of corporate stock, and who generally do not benefit from tax revenues, since the tax structure increasingly serves corporate rather than social interests, so most large profitable corporations either pay no taxes, or at least pay no local taxes to the community whose services they benefit from).

Again, IANAE, but it seems to me that this matters, because non-scarce items generally reside in the commons, while scarce items do not. As more and more of our daily experience is governed by artificially created scarcities, more and more of the commons is robbed for the benefit of corporations.

Nontheless, if I say something like "eliminate artificial scarcities", I bet the first thing that pops into many people's heads (at least in the games developer community) is "Commie!" (I've been called that by at least three Terra Nova posters).

I also think it would be much more fun if people weren't dying of hunger every second of every day, given that we produce as a species more than enough food to feed everyone. I suspect that didn't even make the list of other people pondering your question. So I'm not sure we can have a rational discussion about ideal real world economies, the stakes are too high.

Which is why I find much greater appeal to your second question:

More concrete: If we managed the synthetic economies 'properly' and got rid of the inflation - which is eminently doable - would the games be less fun or more fun?

First, I am really intrigued by the notion that getting rid of inflation is eminantly doable - is that a product of the "synthetic" part, or are you proposing that inflation, which I have experienced as a cyclical, recurring problem, is also a solveable problem in real world economies?

Second, why would inflation ever be desired in an online world? It creates the necessity for developer intervention and maintenance, which is a cost, and, particularly in the unrelenting way it manifests in online worlds, it subjects players to consant erosion of their position - not due to fun interaction and competition with other people, but do to what looks and feels like an artificial, imposed burden from the developers (even though it usually isn't). So I would love to hear why it would be fun.

Final point:

The most interesting thing from my standpoint would be to see how many developer's models of a fun real world would resemble their current virtual world designs.


galiel's right that we haven't really answered the question (of course the asker has remained silent since the asking, so there hasn't been any further clarification...).

I can't see how inflation could be considered fun in RL. Without getting too deep into any theories of what fun is, it seems like fun involves making decisions. RL inflation doesn't increase in any meaningful way our ability to make decisions. It does make some decisions more challenging, in that it involves more mental computation at best (steady inflation) and more guesswork at worst (unpredictable inflation). There will be some tiny segment of the population which will find that additional work entertaining, but I really do expect it's tiny.

In MUDs, inflation is different and tied into a lot of other things, so again while inflation itself might not be fun in MUDs, it might be worthwhile in that you'd have to give up other things to stop inflation.


galiel> Why have money in a game, anyway?

While it's also true, as mentioned above, that money serves as an important medium for exchange, etc... The answer to this one, for me, is very simple. One of the primary reasons people play these games is to indulge the fantasy of becoming wealthy. The way you implement that fantasy in game mechanics is to make something that you call "money" and then enable them to get a lot of it. This relates to my answers to Edward's questions below.

Edward> If we managed the synthetic economies 'properly' and got rid of the inflation - which is eminently doable - would the games be less fun or more fun?

Short answer: In a treadmill game, the answer would be "less fun". In other types of games, the answer is less clear.

Long answer: About all the big hit online games I know of, except for Habbo Hotel, are very clearly descended from Dungeons and Dragons. While there have been evolutionary changes in game mechanics, settings, monster types, skills, etc. the basic elements of appeal haven't changed too much. Like D&D, many (though not all) of the sources of appeal are various forms of escapism, offering players things they can't get in real life. There's being in a world that has magic, castles, wizards and dragons. (Or substitute high-tech, space ships, robots and aliens. Or super powers, etc.) There's having more freedom to decide what you do with your time than you might have in real life. But I think of primary interest to a lot of players is not just the fantasy of becoming both powerful and wealthy (in both gold and in shiny magical goodies), which is after all, acheivable in real life... And not just the fact that most people can become a lot more so in-game than in reality... But it's the relentless inevitability of the equation "time put in = more money and power gotten out". In the real world, I can put time and effort towards a goal and make some progress, no progress, or even end up worse off than when I started. Not so in most online games. It's the "Inevitability of Success" paradigm. (If stated as a principle of online game design, I'd word this in some form like "Players almost always want their own success to be inevitable" or "Players want to always win". People who want things like "permadeath" and/or other ways to "lose" want it so that they can both win AND feel their win has more "value" because A) they could have lost but didn't, or B) other players lost but they did better than those guys.)

This "success is inevitable" paradigm is strongest in the single players games (Wizardry, Ultima, Bards Tale, Might and Magic, and their descendants). In Bartle terms, they're heavily acheiver focused, with explorers also having something to do. The online games, by adding "other people", have the killers and socializers brought into the mix - sometimes (though not always) adding some richness for the explorers as well through the occasional adding of new content (or frequent adding, in user created content games). But the treadmill games still are based on, at their core, single player game "acheiver" game mechanics with chat commands and maybe some guild support grafted on. Even the killers and explorers need to go on the acheiver treadmill to be able to do what they want. (Sometimes even socializers - a lot of people have no interest in talking to a level 1 character!)

In a hypothetical game where NPC shopkeepers didn't exist and most of the needed resources had to be purchased from other players, the problems mentioned above could break, or severely impede the "success is inevitable" formula. But in existing games, player trading/selling and auction houses are a secondary tier, layered on top of merchants who, in most games, will always buy and sell items for the same fixed prices for all eternity. Just like in most single player computer RPGs. These are related to the drop rates and values of items from monsters in a way that ensures any player can totally ignore the potential shortcuts of auction houses or RMTs and still max out their character's level and power & end up with enough money to buy most of the items available in the game. Which is the power and wealth fantasy that almost every game from D&D to the present was set up to indulge. You can certainly get on a game like World of Warcraft and level up a character from 1 to the level cap of 60 in a matter of months without having help from other players or visiting the auction house. Interaction with other players makes the game more fun, and can shorten the time to reach level 60 by a fair amount. But it's not required.

Given that a fundamental core part of why many people play a game like WoW is to reach level 60 and feel like they've beaten most of what the game has to throw at them, they've become powerful, and they've become wealthy, throwing obstacles in the way that slow this down, or increase the number and likelihood of opportunities for failure have to do a LOT to add "new fun" in order to make up for the amount of existing fun they'll be draining off.

Putting more "friction" in by increasing weapon and armor repair costs, rents, indexing NPC vendor prices to the rising money supply rather than leaving them fixed, etc. basically slows the players down so they don't "win" as fast. But most players WANT to win faster, not slower. (Except for the hardcore people who look for game boxes that promise 200 hours of playtime before you win. Most people with jobs and families look at those boxes and say "I don't have enough free time to make the 'committment' to win this game." More people play Windows Solitaire by orders of magnitude - it's quick.)

If on the other hand, rather than evenly distributing the drains that keep the game from being inflationary, you let skillful/lucky players max out just as fast, but you put in enough perils for the less skillful, less lucky, or just momentarily careless players that they lose enough resources to balance your economy... Then you've destroyed the "Inevitability of Success" that's such a key factor of what makes people like treadmill games as an "improved alternative" to what they experience in real life. I remember John Taylor of Kesmai telling me about a player who wanted his Island of Kesmai character back after it'd been killed by some monster. He said "I've invested over $7000 in this character!" (Remember, this is back in the 1980s, when the early commercial online games cost $6 to $12 an hour to play - sometimes as much as $24 an hour during business hours!) He'd always tell this story and say "You INVEST in the stock market, you PLAY our games." While that's amusing, and there's some truth to it - I have to wonder if that guy ever played again, or if faced with the prospect of spending a comparable amount just to get back to where he was before... Plus knowing that he could potentially lose his virtual power and virtual wealth again the way he did the first time... Wouldn't he likely have just quit? Personally I think anybody that spends that kind of money is a "valued customer you must strive to keep around". Especially in those days, when Kesmai derived a major part of their income from a few dozen high-spending players who paid for huge amounts of online time.

Today's situations usually aren't quite as extreme, thankfully. But I don't think the removal of inflation adds as much fun as the fun it gets in the way of. The second tier, player auctions and such, might gain a little fun for some players - while simultaneously, probably losing some fun for others. The RMT business would probably benefit. But that's something that is outside the actual game and gameplay, in most ways - unless you consider the excitement of knowing "I might make $200 for this weekend of hard work in-game" to be a major game element. I think it's not, in a WoW/EQ/UO/DaoC/AC type game, because inevitably only a tiny percentage of the players can make significant money - it's only adding fun for a very, very few. So I don't think bolstering RMT raises the "fun experienced per player" total much, and it can reduce fun for others. In an environment where inflation had been totally conquered, the impact on reducing fun for others is probably less severe. But the gains are still too small to offset the fun reduction caused.

In a nutshell, people want to be "virtual millionaires" so strongly that the best strategy discovered for treadmill games so far is "Hand out money like crazy to everyone, but make people work for it just enough that they feel like it's still worth lots and they're rich, and slow down rampant inflation just enough so they feel like it's still worth lots and they're rich." The main trick these days lies in "How many months/years can you get them to buy into this fairytale before they've got everything, get totally jaded, and no longer care about your silly claims that they're rich in that game world". Though many of them will then go chase the same pot of gold at the end of the rainbow in a NEW game once they tire of the last one.

The fantasy of wealth is one of the most powerful human motivators - even though wealth is mostly a new, man-made concept. In early hunter gather societies, about all wealth might be able to mean is "I have a lot of food right now", and that's pretty fleeting until we got into the agricultural age and were able to store large amounts of grain for long periods of time without it spoiling. These days it's all about wealth, power, and sex (at least we always had those other two). State lotteries sell a miniscule number of people ACTUAL wealth. And they sell hundreds of millions of others the fantasy of wealth, for a brief time, for as little as a dollar. This fantasy is so strongly desired that this industry dwarfs computer and video games - in 1997 the lotteries brought in $34 billion in the USA alone. I imagine the
total is higher now.

For non-treadmill based games, I think the question is harder. I'd venture to say "In some types of games, they will be more fun without inflation." The question then becomes which types are more fun, and which not, and why?

I think it may be the case that attention-economies are inherently non-inflationary. I certainly hope to test that hypothesis with my work. The basic observation I came up with years back was that if a game goes up in population by a factor of 2, or 10, the amount on attention collectively desired by the playerbase goes up by 2 or 10. But so does the supply of "people paying attention to stuff". So it's scalable. But what of inflation, the variation of supply and demand over time? Well, as the population rises and falls, the ratio of supply to demand should stay reasonably stable. Though one might argue "quality" of attention of a given player may decay over time if they get bored with the game. I think it's possible, though, that the level of concentration of wealth in the hands of a few might become a lot steeper as the population rises. This could tend to produce an inflationary pressure on the majority, as the amount available per person for everyone else is correspondingly reduced. I think game mechanics that reduce this tendency and keep wealth more broadly spread out may have some benefit in attention economies. Though letting a few people get super-rich fuels a fantasy of potential success that many others may enjoy, even if they'll never actually achieve that success. (Just like the lotteries.) I'm waiting to see if a game like Project Entropia or Second Life will eventually reach the point where their first real-world millionaire draws swarms of other players to the game hoping to duplicate that success.

Anyway I do plan to try a number of experiments with different types of "currencies" in Furcadia, including possibly setting up local "guild currencies" that players can manage on their own (and establish exchange rates between), and having an attention-generated currency (magical mana rather than money, so people don't complain that it acts so non-money-like) that vanishes every sunday at midnight, eliminating problems of hoarding and inflation & making people willing to spend it frivolously.

I'm sure I'll have some interesting new things to say on the subject of inflation in a few years. :)

Edward> If we designed the real world economy to be fun, what would it look like?

If we're to answer the question in a constraint-free context, I'd say "A lot like Star Trek: Next Generation". Basic needs, and even a number of luxury items, would be available freely and without limit through "replicators" or other advanced technology. But a few things (travel to far-away planets, time and attention from other people, unusual non-replicatable items, rare knowledge, military power, etc.) are limited, to give you something to strive for, trade, etc. Making everyone near-omnipotent and not needing/wanting anything from anyone else that they couldn't get by themself might be fun for many people, but it certainly wouldn't lead to a fun "economy". People wouldn't have incentive to trade if they could get it all without trading. In a Star-Trek-like world, I feel that people evolve a "gift economy", which I think is a much cooler and more fun thing than a "need based economy".

I'll note that I picked Star Trek because this is a question about the real world. In a virtual world, at least at this moment in history, it seems many people prefer a game with more scarcity and difficulty. Second Life and Furcadia come closer to a gift economy than games like Everquest, and Everquest has more players by an order of magnitude. (WoW is more like two orders of magnitude, which is probably why I chose Everquest as my example.) While I still have hopes that some radically different type of game with broader appeal will outperform WoW by an order of magnitude, this is, as yet, an unproven theory.

I think people want something more difficult in the virtual world because "pretending to lose" or "pretending not to be rich yet" isn't as much of a hardship as losing (or being not rich yet) in the real world. While many people might like to pretend to be Conan sweating and struggling for a month to become king in the virtual world, in the real world most people want to win the lottery tomorrow and live it up the rest of their lives. Or better yet, win it today. (As my business partner 'Manda always says - "I want my instant gratification NOW!")

If we're asking the question what kind of economy would be most "fun" in the real world given actual real world constraints - well, it's hard to know if there could be anything much better than we have now. I'd say "capitalism works kinda ok, sometimes". :) The thought of distributing food to everyone hungry sounds nice in theory. Though in practice, my mind is nagged by a quote from "The Book of Ebon Bindings", part of the source material for the old 70s RPG "Empire of the Petal Throne". (I think this was the second RPG that was published by TSR, right after Dungeons and Dragons.) The quote is, "If you give a beggar a coin, tomorrow there will be two beggars." If we gave everyone alive today enough food to eat, wouldn't they just have more babies until they had a shortage again? Many people would - we need to change that cultural element before we can end world hunger. I hear this is a particularly bad problem in India, where people feel the more children they have, the higher the chance one of them will make enough money to support them when they're too old and weak to work.

There is the old saying "A benevolent dictatorship is the best form of government." But to have a benevolent dictator overseeing the economy to make sure it's the most fun for the most people, we'd have to amend that by wishing for a benevolent dictator who is magically guaranteed to be succeeded by another after his death, and another, and another, for all eternity. Or maybe an immortal one. And he'd also need the Santa-like power to deal with billions of individual situations and decisions daily so he could help most everyone. Or else we'd need a network of many, many equally high quality benevolent sub-dictators working under him. But maybe this is just as far-fetched as Star Trek replicators. Or, given human nature, probably even more so.


Dr. Cat> If we gave everyone alive today enough food to eat, wouldn't they just have more babies until they had a shortage again? Many people would - we need to change that cultural element before we can end world hunger. I hear this is a particularly bad problem in India, where people feel the more children they have, the higher the chance one of them will make enough money to support them when they're too old and weak to work.

This is getting away from MUDs (and inflation), but the "tragedy of the commons" expression comes from a http://dieoff.org/page95.htm>paper by Garrett Hardin where he was talking about what he believed was the need to coercively control population. One could make the argument that allowing people to starve so that they don't reproduce is a form of coercion.

While the phenomenon of the tragedy of the commons is evident in a lot of situations, it seems less applicable to population growth. It's looking more and more like the best way to control population involves a degree of economic security for the poor and economic and educational opportunities, most particularly for women. Amartya Sen talks about this in regards to different states in India which have tried different approaches.

As to what you say about inflation, it certainly sounds sensible to me. It seems like the unstated goal for many designers is: (a) a player is gauranteed to get wealthy in time, but (b) the overall economy will not become so wealthy as to disturb the game. Are these incompatible?


It's looking more and more like the best way to control population involves a degree of economic security for the poor and economic and educational opportunities, most particularly for women.

I have to agree. It's been known for a long time that when countries reach a first-world level of prosperity, their poulation growth rates drop dramatically, to just a little above equilibrium levels. Various theories have been advanced regarding why exactly this is. But regardless of the reason, I think elevating the second and third world countries to that same level, or as close as possible, is the surest (and most humane) way to solve the problem in the long term. It's something that will take quite a number of years, though.


With so many different opinions of what inflation, deflation and MUDflation actually means in virtual worlds and RMT-wise in the real world, I wonder how Entropia Universe can claim a stable 10 to 1 ratio to the dollar. It really seems dubious to me.

Does anybody know how their economy works? Do they use the typical faucet/drain approach?

I assume inflation is not fun but a by-product of economic growth (which I assume is "fun"). So do the EU guys have it perfectly balanced and when this is the case, is it then the first booming economy without inflation? If so, then the answer to whether inflation is fun would be "not necessarily" as enough people seem to enjoy EU and "no" in case of those in for making money.

The comments to this entry are closed.