Original Post: Jan 13 2013
Let’s talk a little about game design. As a researcher and teacher, I sometimes find it difficult to express in words exactly what it is that game design is, or what I hope to teach to outsiders–even to gamers themselves sometimes. Capturing exactly what the difference is between design decisions and development process is difficult, since “good” design often fades into the background because it just works. As a player, one rarely notices all the individual rules and decisions that have been made, the experience occupies the conscious mind.
Discussing game design is difficult because it is a second-order task. A designer constructs the rules, environment, etc. but always in hopes that the player will experience something that is greater than the sum of those parts. When set in motion, the player(s), rules, environment and fiction all combine and create something that is not obviously discernible from a list of the game’s rules. Further, since it is the player whom we as designers are trying to affect somehow, their perception and personal interpretation of what is happening is more important than what the game design document says should be the result. (Not that players can’t be wrong or ignorant, but we still have to deal with their perception of our work.)
Finally, discussing the design of the kinds of games I like is very difficult, because there is very little opportunity for experimentation to verify any hypotheses about the way the game works. That is, I’m not often able to change one rule and see how it plays out. The games are sealed and finished, I can’t tamper with them myself very often. Even in franchise games such as Assassin’s Creed or Mass Effect, which are largely similar, many changes have been made between iterations, so it becomes difficult to nail down exactly what cause has led to which effect.
For these reasons, I’ve been excited by the learning opportunities I’ve had while playing the latest post-DayZ Arma mod to sweep the Internet: Wasteland. In general, Wasteland is much like DayZ: a semi-persistent, on-going mission set in Chernarus (or other maps now) where players must contend with each other, as well as manage a hunger/thirst mechanic. There are no zombies in Wasteland, only players. Much higher-grade weapons are generally available than in DayZ, mostly found in the trunk of the many vehicles which randomly spawn throughout the settlements. So Wasteland is, broadly speaking, a very large, variably-paced deathmatch. There are 3 factions, BluFor, OpFor and Independents. The Op and Blu forces are designated as teams who are to work together, while Independents are free to do as they please–team up or lone wolf.
From this basic premise, each server begins to differentiate. In a digital version of “house rules” many server admins fancy themselves amateur game designers (or at least developers, more on that soon), and so start to tinker with the code that defines the Wasteland mission. These small changes ripple outward in unpredictable ways, and interact with other changes, to affect the overall experience.
From here I will explore two examples of rules which are affecting the experience on certain servers that I’ve been playing on. The beauty of this server situation is that I can jump between very similar games, and compare the two sets of rules in a much more “controlled” experimental way than is possible in many other games.
The first issue is really more of a problem than an example of a rule-change between servers, but serves to illustrate my point about game design. In Wasteland, players can use various kinds of building-materials to create bases. These include sniper platforms, sandbag walls, and other ready-made base buildings. Some of these items are quite large, and the mechanics whereby a single player can walk along carrying a small building around is quite ridiculous. However, the problematic aspect is that a BluFor player can pick up a ten meter-long wall, and by walking or turning it around, can kill the other players using the collision physics built into the game.
Obviously, no cooperative team mate would want to do this–but this is a game played on the internet where people are jerks, so it happens all the time. Some troll will go to the BluFor base, as a BluFor member, and simply move a base wall, killing as many of his faction as he until one of them shoots him. The issue here is that the collision physics do not track who is responsible for the deaths–the system has no way to know that it was BluFor Johnny that killed his own team. Thus, the punishment mechanic, which is a way for players killed by their own teammates using weapons to remove team-killers, doesn’t work. In fact, it is the troll–once shot by BluFor–who is better equipped to punish his law-abiding teammates.
To me, this is more of a development problem or task, than a design concern. This is a sort of maintenance issue where the game is obviously not working “right.” The rules that define how the game is designed to work aren’t perfectly translated into computer logic, and we have a less than optimal result. That said, the rules, the design decisions, are already made: teams, base building, physics etc. We simply need to code them into the computer system better.
Finally we come to the meat of this article–a long enough build up I’m sure. Recently, one of the servers I play on has changed a few things in Wasteland. I’m going to report the situation backwards for rhetorical effect here: Users noticed the changes and started complaining. Mainly, the complaints where about the lack of vehicles in the spawn areas. People were spawning in and had to run a couple of kilometers to find a car. This was not the case as recently as last week; cars were generally easy to find. Cars, remember, also usually contain weapons. So people were complaining about being required to run around with just pistols for quite a while.
The obvious solution to this symptom is to increase the number or frequency of vehicle spawns in the game. However, this was not the actual problem. What the admin changed was where people were spawning. Players were spawning along the coast, a much narrower range of areas than previously. The result of this is that the people who spawn first get in a vehicle and drive away, die somewhere inland, and repeat. So eventually, players have organically moved all the cars that were at the coast to somewhere slightly inland. The perception was that there were fewer cars, when really the issue was that the players were moving them basically in one direction, away from the coast. Since no one was spawning inland, the cars that made it even to the first set of towns away from the coast were staying there. People weren’t driving any cars from inland back out to the coast, or at least not enough to make a difference.
By changing the spawning area, the admin did not just create a vehicle shortage. Because the sixty-or-so players were all spawning in about ten locations, rather than forty or fifty, there were far more shoot-outs in those spawning locations than when the spawns were spread out. These shoot-outs were often between pistols, since very few people had any access to higher grade weapons. Secondly, they are usually fatal, so at least one of the participants respawns again on the coast, exacerbating the problem.
So, there are no vehicles on the coast, there are many players, who continue to fight, and kill each other, and respawn again on the coast. Getting inland became doubly difficult. This has another knock-on affect, impacting the experience of being inland itself. Prior to these changes, the distribution of spawn locations and the high mobility offered by vehicles led to the entire map being “in play.” Unlike DayZ where certain obvious paths became well-worn between highly-valued locations, Wasteland was less predictable. One could expect an encounter at any moment, or could dig in almost anywhere and set up an ambush. After the changes, with a large percentage of the population stuck in a coastal deathmatch, the rest of the map is desolate.
I have asked, but have yet to hear a cogent answer as to the purpose of these spawn-point changes. I’m not sure what the admin was seeking to do, but I’m sure this isn’t the effect he wanted. The example is illustrative though–how one change can have such dramatic impacts, and how the symptoms (“Why aren’t there any vehicles anymore?!?!”) do not always point to the real problem. As I said above, game design operates at one remove: you design rules, but what you want to create is play.