I keep talking about the importance of considering what technologies (in the broadest sense of the word ‘technique’) can do to a society when writing a story. In many stories, the technology itself isn’t the purpose of the story that was being written. Inception is a heist film – in some sense, the dreaming technology is less relevant than the interactions between Cobb and his memories of his wife, and his quest to return to his children and home even if that means assaulting Fischer. This is the second of Asimov’s three kinds of science fiction: the technology is incidental to the adventure.
But what about role-playing and strategy games, where the techniques concerned are the main way that the players’ characters interact with the game world? Here it is essential to consider how the technologies interact with one another, in the form of possible combinations of the rules. Otherwise the game may end up essentially unplayable. Most good games don’t have that severe of a problem, but considering all of the possible combinations of rules becomes very difficult for complicated games where the full description of the rules may be hundreds of pages long. In particular, economics is very hard to do right.
If the game developers and beta testers haven’t found and fixed all possible problematic combinations of rules, there will be exploits to take advantage of. And given a large enough base of players, they will be found. There are very long lists of such game-breaking techniques, but here I’ll focus on economic ones.
In all of the versions of Dungeons and Dragons, there are exploits that allow relatively low-level characters to defeat any opponent. The most notorious example is Pun-Pun, a level-1 starting character in DnD 3.5 that has arbitrarily high power. But the exploits in DnD are easier than that. The purchase price of a ten-foot ladder is less than the sale price of the two ten-foot poles and shorter rungs that it is made of. A player can in theory drain all of the cash out of the local economy.
In the Dresden Files RPG, there is another simple exploit. A lot of the competitive balancing in the game relies on high-powered magical characters being unable to use complicated technology, particularly computers and other electronics. This is justified by the characters ‘magical energy’ damaging the electronics. They are walking techbane. But moving water is also established to block magical energy. You can still use a cell phone, a GPS, a computer terminal, and all of the fancy gadgets in a modern hospital as long as either they or you are encased in a thin layer of circulating water. Time to go shopping at a fire-fighter uniform supplier.
In Mage: The Ascension, some characters can magic the laws of probability and win the lottery. That may get a certain amount of unwanted attention, but you only need to do it once. To deal with this sort of thing, all of these games have one basic rule: the moderator is always right.
Things get a bit more problematic in games without a moderator constantly adjusting the rules to avoid or limit game-breaking. It doesn’t even have to require a large player base – AI programs can do the same thing. In 1981 and 1982, a challenge using the rules from the sci-fi RPG Traveller was twice won by an early learning program. It found that thousands of kamikaze ships would defeat any other solution.
A more recent example: Starcraft is very close to competitive balance, but in scenarios with nearly-infinite resources and equally-skilled players, the Protoss game race has a slight advantage over the others. They can assimilate enemy units and potentially have three times the army of anyone else (if the game lasts that long).
Game software can be updated. This is easier for online games. In World of Warcraft, there have been both positive and negative game-breaks. On the positive side, there was a bug that could be exploited to allow a single paladin character to do death by a thousand cuts to the hardest-to-defeat enemies in the game in one move (as opposed to the usual method of two dozen characters taking several minutes to bring it down). On the negative side, a programming bug caused the Corrupted Blood debuff to turn into a pandemic inside the game, killing off or wounding almost all characters. Those were all fixed by obvious rule patches in short order.
The Limits of Rule Patching
But rule patching can only go so far. The most complicated game breaks arise from interactions between many player characters, in effect a large synthetic economy. The World Of Warcraft internal auction markets are nowhere close to equilibrium, and arbitrageurs can make lots of in-game money. In some cases, trading bot programs have accumulated up to several times the total amount of money in circulation on any one game server. Blizzard deals with that by shutting down bot accounts whenever they are detected. But Matt Fisher at Stanford tells me that bot programs can be programmed to appear almost identical to a human player who obsessively trades on the market, so there is no way to fix that.
Perhaps the limits of rule patching can be excused. When game systems are so complicated that no-one can predict their outcomes, and when they involve the interactions of thousands of separate agents, fixing problems with them is as hard as fixing problems with the real-life economy. Doing significantly better than random chance would be impressive.