Tabletop RPGs and Level Editors
by Ian Schreiber
For the longest time, I had a hard time even thinking of how one would evaluate a tabletop RPG (say, comparing the new D&D to the old one… or in my case, grading a student-created d20 supplement). With board games the obvious answer is “playtest it, DUH” but to an extent, that fails with RPGs: so much of the experience comes down to the players. A great GM and great players can work with a mediocre system to produce some really fun moments, while an inexperienced GM and lousy players can ruin what’s otherwise a wonderful system. How do you separate the players from the system when it is the combination of the two that produces the end result? How do you tell the difference between d20 and GURPS when you can pretty much run the same campaign in either system?
It occurs to me that this is the same problem that would be faced by a designer trying to create a game where all (or most) of the content was player created, like Second Life or NeverWinter Nights or Unreal Tournament or LittleBigPlanet or WarioWare DiY. Anything where the game = the level design tools.
The answer, I think, is that you evaluate based on what the players are going to spend most of their time doing, regardless of how good a level was designed. With Unreal the player is going to be doing a lot of running and shooting. With WarioWare DiY the player is trying to figure out what to do and then executing a plan in a very short space of time. With Dungeons & Dragons, the players spend most of their time rolling dice in combat situations. Sure, you could run a D&D campaign with no combat and just a lot of talking and political intrigue, just as you could make a turn-based strategy game in the Unreal engine, but that’s really not what the system was designed for.
Designing these kinds of games is hard because it’s a third-order problem. You’re designing a system, which someone else is going to use to create content, which someone else is going to experience. The trick is to decide what you want the end players to spend most of their time doing, and then design systems that encourage such behavior. For example, suppose you wanted to design a tabletop RPG about bragging, where each player’s character is trying to convince the others that he’d win in a fight, and the actual combat takes only a few seconds compared to the hours of pre-combat taunts. In such a case, you’d want to make combat resolution into a single die-roll, while designing multiple successive die-roll contests for speaking (rather than a system like D&D which is the opposite). Or suppose you wanted to make a video game about physics interactions between objects; then your level editor had better have all kinds of objects that interact with one another in clever ways (weights, ropes and pulleys, springs, bouncy balls, etc.).
How do you then deal with the playtesting problem, where players might be having fun (or not) in spite of your systems, simply because there is a really good GM or level designer? The answer is that you don’t just look at whether the players are having fun or not; you look more closely at how they are spending their time, proportionally, at each task in the game. If players are spending more time in your FPS engine walking and looking around and not enough time actually shooting at each other, or if players in your conversation-based RPG are spending half of their play session resolving a single combat, then you need to go back and check your systems to simplify the non-critical actions while complicating and extending the core actions of your game.