Balance in MMO Economies?

February 9, 2008 11:48 AM

Game Design: In an MMO, how do you build an economy that will 'balance' itself? Obviously, having solid inputs and outputs are one thing, and balancing those inputs with the demand for the item, but what other things need to be taken into account?

I can't seem to find anything that's *written* on economy design in an MMO game, but I know that it's always been a big issue and has caused the downfall of a few -- E&B to name one. So as I start to experiment with code in this genere, I'd like to try to get this right the first time. ;)

I suppose this is true for any artificial economy, and I don't know much about the microeconomics that are necessary to write all the business logic.

posted by SpecialK to computers & internet (23 comments total)

I feel like it would involve a drop rate (or supply rate) for inputs contingent on the average number of players to ensure that inputs required would be available. Drop rates would be a baseline for value, but then the benefit of output would skew value in other directions. I feel like this would be more easy to tweak in testing than to explicitly model because of how subjective a lot values of an item are; Why does somebody want this item? because its useful, cool, required for something.. etc.
posted by dobie at 11:57 AM on February 9, 2008

Besides well-thought out considerations of scarcity, utility, and scale? I think a lot of well-balanced game economies come from how well they engage the player base. Assuming that some sort of player time input (i.e. labor) is needed to make the economy function, the more time you can get players to spend exploring the economy (ie. testing it), the more structurally robust it will become. I would be surprised if there wasn't someone at DiGRA presenting papers on this. It might be worth your while to check their archives.
posted by mrmojoflying at 12:08 PM on February 9, 2008

Designing Virtual Worlds by Bartle talks a bit about designing economies. I've only skimmed it, but it is a well regarded resource.

Castranova's Synthetic Worlds hits on many of these topics as well. As an economist, he has done some of the best work on understanding economies in virtual worlds.
posted by i love cheese at 12:22 PM on February 9, 2008

You may have to do a little digging, but I believe that the escapist has written a bunch of articles on MMO economies.

In my opinion, EVE has an economy that has a lot of impact on gameplay, making it important to have a well functioning one. I remember reading that they hired an economist to help them with its design and stabilization. Their Dev Blog has posts on trends they see. I know this isn't a direct answer to your question, but I think a good example of a complex MMO economy that has a lot written about it.
posted by Ctrl_Alt_ep at 12:37 PM on February 9, 2008

I'm not sure you could ever get it right the first time... I think the closest thing you could do is concentrate on making your system as flexible and as easy for you to manipulate on the fly as possible... maybe even a system that crunches data about the gear/gold ownership of players and tweaks supply in reaction.

For every quantifiable, there are factors that you will never be able to predict because of the human element. A weapon with inferior stats may be more popular than a superior one because players think it "looks badass" or is funny in a way that will break ice with the other players. Something that's basically useless -- the Orb of Deception in WoW comes to mind -- may be fiendishly sought after for humor value or sheer coolness. How do you quantify "funny"?

When you're balancing an economy, you're really balancing challenge and frustration for the playerbase to keep them challenged enough to be interested and not frustrated enough to quit. I don't think you'll ever find a mathematical formulae that truly works for that, because every player is different. Not only that, but when different players aggregate, you'll get servers with completely different economies... and some of those economies will be crap. Always.

I don't actually believe that an economy has ever sunk a game all by itself. There's always some other factors at work. If people are pissed off because there's not enough materials for tradeskills, it's not just a drop rate problem. It's a tradeskill balance problem, it's a "the quests in the zone where those materials drop are not fun enough and players are going somewhere else" problem, it can even be a "the starting race that levels in the zone where those materials drop is so ugly no one wants to play it" problem.

So yeah -- I'd say concentrate on making your system easy to tweak. The ability to easily go into your item database and bump some sword from a rareness rating of 10 to 11 and have that automatically tweak its drop rate, the amount of money a vendor will give a player for it, etc. throughout the world will save you seven billion more headaches than trying to get the rareness of it right the first time.
posted by Gianna at 12:45 PM on February 9, 2008

In its simplist form. The goal in MMO economies is to create and remove resources in the game at a balanced rate. This is on top of trying to make it feel believable and enjoyable within the context of a game (doesn't have to be perfect...and on the flip side no one would enjoy playing through periods of severe inflation or deflation other than the die-hard economist types).

The idea is, since you are essentially creating virtual resources out of nothing (through player action), you need a way to remove those resources to prevent surpluses of items that progressively lower in value. The true goes for the opposite as well...you don't want the items to be so valuable that they are used in place of the game's currency for assigning worth.

As far as resources/discussions to look at, I've found the following to be interesting reads:

Eve-online's economy blog
Discussions at gamedev
Proposed economy for Infinity Online

Other than that the actual mechanics don't really matter as long as it tries to balance sources with sinks. Some implementations are better than others, some acheive balance through placing extra restrictions (eg. soulbound items in wow...or degrading items in DAoC). The idea is to always have a constant flow in and out of the game...to keep the demand constant.
posted by samsara at 12:53 PM on February 9, 2008

Charlie Stross touches on these issues quite a bit in Halting State, which is well worth a read.
posted by monju_bosatsu at 12:54 PM on February 9, 2008

I suspect a lot of it has to do with intelligent sinks.

In second life (leaving the land economy out of it) this is primarily done through upload fees. Uploading textures is needed for the player-created objects (clothing, skins, animals, guns, etc) which are bought and sold and fuel the economy.

In A Tale in the Desert, it's donations of raw materials to "universities" needed to unlock new technologies, with which players make things to do a variety of things, including opening up more advanced technologies.

I'm not a big fan of PvP games but if you're interested in games with well-balanced economies outside that model, spend some time in A Tale in the Desert, which despite sucking as a game, has the most well-rounded and well-planned economy I can think of.
posted by DarlingBri at 1:05 PM on February 9, 2008

The new MMO Pirates of the Burning Sea purportedly has a rather deep player-run economy. I played the beta, and found the whole game yawn-tastic, but one of the developers had an interesting take on why they chose a blind auction model for conducting player transactions.

Of course, now I can't find it.

*googles furiously*

OK, that's not it. But it's close.

Dammit, this is not it, either. But it points to this, which also isn't what I'm looking for.

But hey! There it is!

EVE Online has had a staff economist for about a year now, I think, as mentioned above. I play EVE, and a large part of my time in-game is spent making economic decisions: what to manufacture, where to buy raw goods cheap, transportation costs and risks (20 jumps through low-sec space to buy low might be worth it, or I might lose my ship and all its fittings)...and, yeah, it's really tedious but still somehow fun.

The great thing about EVE's economy is that, with the exception of skill books and basic blueprints, everything in the game is player-created -- from the lowliest ammo to the most gargantuan battleships. So prices can fluctuate wildly across regions, and wily traders (with better ships than I can fly) can make mucho moolah off of those shifts.

EVE also limits economic information regionally -- this means you only see prices in the region you are in. Asteroids, which are mined for their minerals, are also inequitably distributed. So part of what makes EVE's economy work is asymmetrical information and resource distribution.

This is a fascinating question, and I look forward to seeing what others have to say on the topic.
posted by BitterOldPunk at 1:28 PM on February 9, 2008

The problem you'll encounter is that the twin goals of building a balanced economy and putting the "massive" in "massive multiplayer" are incompatible. To make a balanced economy you need stuff like inflation if the amount of gold (or whatever you use) gets too high. But if you do that, you price casual players out of anything but the shittiest gear. And "casual" players, for large values of casual, are what make up a lot of your income stream.

Look, if 2% of your players eat, breathe, and sleep the game they're going to be pulling in so much gold and loot that a casual player could never keep up in a million years. So do you say, "fuck it" and give up on a balanced economy or do you inflate away the value of the loot the hardcore players are pulling in?

If you choose the latter, how do you retain the casual players who will quickly see that they will never, ever be able to afford good gear and the longer they play, the further behind the curve they get?

You can make a decent economy (see Eve), but you won't have nearly the playerbase (read: income) if you do.
posted by Justinian at 2:01 PM on February 9, 2008

In WoW, the concept of items being "soulbound" (already mentioned by samsara) seems to be pretty important. Most good items can't be traded off a character once being equipped, or sometimes even once picked up. This keeps them from being devalued by having the market absolutely flooded with them.

There also need to be sinks that remove money from the economy completely. Repair costs in WoW would be a good example of this.

You might want to check into Kingdom of Loathing, either the forums or the radio show, where they seem to discuss market balancing issues pretty often. In that game, they give you a rare item in exchange for a cash donation that can be sold directly for "meat" (the in-game currency), so they also have to deal with the issue of market inflation resulting from real-world cash basically being used to create more money in the game. Part of their solution is to hold raffles for extremely rare items, which remove a lot of cash from the economy.
posted by dixie flatline at 2:11 PM on February 9, 2008

Here's an example of the kind of tradeoffs you've got to make:

There's a desireable item you can buy at a shop that makes life a whole lot easier. Let's call it widget alpha. The more widgets you have, the better able you are to survive. widget alpha costs 1000 gold. Lets say it heals you up fully when you use it, but the widget disappears. When the game is first released and everyone is hunting and skinng sheep for 1 gold a skin, this works well.

Now fast forward a month. Casual players have moved on to hunting goblins and selling their gear and can afford a widget every so often which they use in extreme emergencies. Fine. But the hardcore players have leveled up insanely fast and are taking down minotaurs already, pulling in maybe 50x as much gold as the casual players. They can afford as many widgets as they want without blinking. The cost doesn't register anymore.

The hardcore players having an infinite amount of widgets sorta breaks the game. It makes things much easier than intended. They can heal themselves up at will and thus are far more survivable than they should be, which lets them hunt higher level stuff than they should be able to, hence bringing in more gold, faster, than intended. Which lets them buy higher level stuff in a positive feedback loop.

What do you do? Pull the widgets? The casual players will scream (and some will quit) because they need these widgets for emergencies since their shitty gear doesn't protect them well. Peg the value of the widgets to the total amount of gold in the game? Then the hardcore players will be able to afford widgets occasionally, as intended, but the casual players are priced out. Same result. Leave it alone? Then the game is broken for the hardcore players who are pulling in gold at an accelerating rate. They see the game is broken for them and bitch about it. And they're your most vocal players even if a much smaller subset than the casual players.

So what do you do?

There aren't really any good answers if you want to appeal to a wide audience of both hardcore and casual players.
posted by Justinian at 2:14 PM on February 9, 2008

The In-Game Economics of Ultima Online is the original presentation/paper that launched this entire sub-field.

One other issue you need to take into account is how your economy will recover from the inevitable duping exploits that some clever players will find. The system may be reasonably well balanced most of the time, but there will be occasional sharp shocks as someone finds a way to run off a zillion silver chalices in a few hours and starts converting them into cash and from there into whatever they want. One-off events, as described in The Lessons of Lucasfilm's Habitat, can be useful, but you should also think about how aggressive you will want to be about tracing and confiscating dup'ed goods.

Good auditing tools are essential. You'll need to be tracking prices and various macroeconomic indicators, and you should have in place some of the basic data mining tools you'll need to take a closer look when you get the sense that something isn't working right. The Ultima paper is very good here; read with careful attention the parts where they tried to figure out why players were unhappy.
posted by grimmelm at 2:32 PM on February 9, 2008 [1 favorite has favorites]

Justinian: I'm just guessing but it seems to me that what the EVE devs would do is nerf the widgets: instead of full healing, maybe they'll now only heal up to x number of health points -- still useful for the casual and mid-level players, less so for the hardcore.

Who, will, of course scream "nerf!" and bitch endlessly on the message boards and threaten to take their toys and go home. But most won't: they've invested too much time and energy into the game to leave. And now there's money to be made farming widgets for the lower-level players who still need them.

If the question is "how do you build an MMO economy that will balance itself?" I suggest the answer might be: you can't -- developers will always have to intervene and tinker to keep things balanced. This will inevitably piss off players. I guess the degree to which you can balance and retain (and grow) playerbase is still an art and not a science. I suspect a "self-balancing" economy would be too simplistic to be fun to participate in.

on preview:

From the UO paper's suggestions of game modifications:

Create enforceable contracts.
Create auctions.
Allow vendors to buy.
Deepen production paths.
Change from improve-by-doing to purchase-to-improve.

Interesting that since this paper came out in '99 all of these have been implemented in different ways in different games.
posted by BitterOldPunk at 2:56 PM on February 9, 2008

A few years ago, the place to go for this kind of discussion was the Mud-Dev mailing list. My subscription imploded a couple of years ago and I didn't resubscribe, so I don't know the current state of the list, but ten years or so ago it seemed like every major MUD (and later MMORPG) designer turned up there sooner or later. I suggest tracking down the archives.
posted by Leon at 3:46 PM on February 9, 2008

Just as a side note to the discussion: having played both Eve and A Tale in the Desert, I'd say that Eve has a far more complex economy model than A Tale in the Desert. ATITD is largely a barter economy (I know you can make currency using the mint, but last I checked practically no one used currency for anything and Egypt didn't have a unified currency anyways) and has, at most, 1500 active players. Eve has a much larger player base, manufacturing is a much bigger deal (to make anything beyond simple items like ammunition and basic ship modules, you pretty much have to venture out of empire space and start mining moons and ice fields), and there is not only a standardized currency, but also a stock market and a built-in auction/classifieds service.

About the only thing ATITD has that Eve doesn't is a deep crafting system; it's been a while since I've played Eve but as far as I know, advances in manufacturing only really get you reductions in build time or resource usage. In ATITD, you actually have to hammer out your own metal tools (axes are a big deal) and the crafting mini-game is hard enough that many people can't even make a usable one, let alone a decent-quality one. As a result, extremely high-quality axes that give you more wood per chop can fetch extremely high barter prices, while guilds often toss noobs middling-quality axes for the cost of the raw materials, or even for free.
posted by chrominance at 4:18 PM on February 9, 2008

This is out of my depth, but this kind of thing seems to cry out for some artificial-life simulations.

If you could come up with one set of economic conditions, and a bunch of agents that model certain economic behaviors, you could wind 'em up and watch 'em go. Repeat with different economic conditions. You might be able to do this using a stub of the game engine, wired to interact with computer-based agents.

Obviously creating the models would not be trivial; just coming up with useful models for the economic behavior of agents might require some genetic algorithms itself. You could probably get some fascinating data out of the experiment.

Shoot, you might be able to turn the model for your economic conditions into a genetic algorithm responding to real-life gameplay. I'm just spitballing here, but if you envision the Invisible Hand as a computer-run "player" in its own right, well, that seems like it could lead to some interesting outcomes.
posted by adamrice at 4:44 PM on February 9, 2008

About the only thing ATITD has that Eve doesn't is a deep crafting system; it's been a while since I've played Eve but as far as I know, advances in manufacturing only really get you reductions in build time or resource usage.

Actually for the larger stuff its quite similar...for example, to build capital ships you first need to erect a player owned station in lowsec or 0.0 space, then use a capital building array. This station would of course be at risk of attack so you would need resources to protect it. Aside from that all the components have to be made from scratch minerals and datacores (or they can be bought outright). The crafting system is quite complex actually...for someone that was barely into it I had spreadsheet upon spreadsheet on mineral conversions.
posted by samsara at 5:02 PM on February 9, 2008

I am not a game designer, I simply enjoy MMOs, but here are some tips to point you in the right direction. First off, don't look at Second Life or Project Entropia as good examples because they both rely fairly directly on the player pumping cash into the system to get any sort of buying power, and the more your game is an actual game (and not SL's glorified 3D chat room) the more alienating this will seem to the player. Secondly, realize that while Eve is a good example of a robust economy, the economy in that game is a fairly major portion of the gameplay which the player can not safely ignore, and also that Eve is viewed as fairly cutthroat and intimidating place to get a foothold in.

So what makes a good economy generally speaking? Well, the day to day items that people consume during the course of gameplay need to be fairly easy and cheap to acquire or else the whole thing grinds to a halt. You also need good basic and cheap weapons and armor that will be suitable to sustain the player against monsters as they progress in levels but not give them any real edge. A good game will present the possibility that a hardworking player can get to the high levels simply with items they farmed if they aren't in a rush to get there and they don't want the very best equipment. Without these base level items the game will be frustrating and the player won't have anything to compare the nice items against. I feel like I'm overstating this, but once an economy starts dying (try playing the current version of Star Wars Galaxies for an example) this problem becomes very apparent.

There should be a middle tier of items that are craftable and take some work or skill to acquire the parts for, but they give a decent advantage and anyone who really wants to can get one eventually. There should be high level items that are rare, useful, and not everyone can own. And there should be sinks in place to skim currency off at an adjustable rate to control inflation. Usually sinks occur at both the player and the guild level, but be careful about being too aggressive because some people simply won't opt in no matter what advantage you give for the sink.

To get an idea of economies, try generating a list of games that you're interested in and then searching for them on Ebay every day for about a week or so. Ebay tends to remove listings for games on a regular basis, but over time you'll gather an idea of which games are having a fairly high volume of trading.

My personal suggestion is to look the Kingdom of Loathing as a fairly good economy. It has its problems like all do, but the high level items are mostly monthly items that cost $10 each in substitution of an account fee. Since these items aren't reintroduced, they slowly become rarer and reward investors, while the constant rate of new high level items being designed keeps it a possibility that new players can become competitive over time. Many of these disappear after use giving the player a skill or familiar, and having rare items tie to the player and not float around forever keeps them from ever going down in value. There are also ultra rare items where only about 5 or so are created a day with no possibility to farm them, so regular players who get lucky can transfer this into an economic advantage. There is a middle tier of consumable items that drop at regular infrequent intervals and seem to sell quite well. And finally they have high level equipment that can not be traded and has to be earned, sometimes with a finite window of opportunity to earn it, and this give players the opportunity to opt out of the marketplace and still be able to show off because everyone works for those items the same. The game is balanced so that you can earn the in game currency for one of the rare items with about a month of farming, which seems to work. Most of the Kingdom's woes come from currency bugs that have cropped up from time to time.
posted by CheshireCat at 4:04 AM on February 10, 2008

ATITD is largely a barter economy

Actually, I think it's entirely a barter economy. Currency never really took off as far as I could ever tell. But virtually all trading is based on the trade value of two staple crops. Even if you never grow them yourself, and are trading forged iron for goblets, the value of that trade can be traced back to wheat or flax.
posted by DarlingBri at 5:48 AM on February 10, 2008

Clive Thomson wrote an amazing article a few years ago for a weird magazine that talked a little about economics of games and also talked about how MMOG are little petri dishes for economists and political scientists.
posted by Mozzie at 3:34 PM on February 19, 2008

I would note that Kingdom of Loathing has a history of occasionally needing meatsinks- activities which pull meat (the currency) out of the economy. This wouldn't be so interesting, as currency sinks are hardly common, but what gets me is how creative the KoL devs are at coming up with blatant meatsinks, obviously designed to get you to throw money away on nothing but which people still throw meat at. A classic example was the Penguin Mafia Protection Racket. You could buy Protection Contracts from the Penguin Mafia for meat, and you could also order hits on people. Every so much money paid for the hit wiped out one Protection Contract, and you could order different kinds of hits that had different effects on the victim (assuming you paid enough to burn off all their contracts). The thing was, there was no benefit to calling in a hit on someone- just the joy of mildly annoying them (particularly given the near-total lack of death penalty seen in Kingdom of Loathing).

So in summary, the KoL devs managed to come up with a currency sink that provided about a week's amusement at calling out hits on your friends and enemies and which drained millions of meat from the economy in both hits and contract fees. Back when I played KoL, there was a cool meatsink like that maybe every couple of months. They've hit on a terrific means of controlling their currency supply by persuading people to simply throw the stuff away, and that's neat.
posted by Pope Guilty at 7:00 PM on February 19, 2008

What you're neglecting, Pope Guilty, is the reason those meatsinks had to be introduced. There were some bugs in KoL's code that allowed some users to gain huge amounts of meat. Bugs like this aren't actually uncommon on MMORPGs. I remember WorldsAway had some problems with people who could get negative numbers of tokens, which causes the bits to wrap around and is essentially billions.

When all this money gets put into the system what can the developers do? Usually all transactions like this aren't logged, so it's not even always easy to tell which users did it. They could roll back the database but then all transactions that happened in the meantime evaporate. It's not necessarily the case that they can simply ban users with 2B+ tokens, since they could have distributed some of that money. (In fact, in WorldsAway, I think there was organized distribution specifically to make it hard for the devs to figure out who was the original source and forestall immediate retaliation.)

In KoL's case, they handled the problem through a combination of community spirit (Jick is very well-liked among the playerbase in general, and KoL is meant to be not-too-serious, which meant he got some money turned in just by asking for it--he rewarded those people with items and titles, I believe) and those meatsinks. Most of the sinks were sales of items that cost truly outlandish sums. There was one item, Mr. Exploiter, that if memory serves sold for a billion meat. A KoL item is mostly just a gif, some variables, description text and a name. If there's no special function requiring coding, then it's really simple to add a new item, yet players still perceive them as being something they can own, and people who fancy themselves collectors like to own ultra rare items for the social status they afford. What Jick really did was leverage how easy stuff is to make relative to the perceived social value.

Anyway, to the overall question:
I think it's actually possible to make a self-regulating MMORPG economy. The key is to have the money introduced to the economy (through item sales and periodic payments, if any) and sunk out of it (through system vendor sales and other non-circulatory payments) be close to each other, with a bit of leeway one way or the other to account for playerbase growth/shrinkage. That's not perfect (it doesn't account for player saving/hording), but the one great advantage of doing an MMORPG economic system over real economics is that the computer can know everything. It can count up all the money in the system, how much is in savings, how much is in general circulation, and so on. If prices are dynamically tied to a good overall formula, it should be possible to have a controlled economic system the likes of which Greenspan could only dream about. This is cheating of course, but if you just want a workable economic system to keep the game going then realism can be much more trouble than its worth.
posted by JHarris at 6:28 AM on February 26, 2008