Complexity in RuneScape
As a computer game, RuneScape is allowed to be as complex as it likes and still hide all that complexity from the player. In most cases, this complexity adds little benefit, but at even littler cost. The end result of so many small choices is a rich landscape of optimizations for loadouts, tasks, character levels, markets, and so on. With the repetitive nature of its game play, RuneScape is a min-maxer’s dream, allowing them to eke out marginal, fractional benefits from every decision. It’s also “transparent” enough that you’ll be fine if you don’t care for all that. Combat pures are an example of this, and while a combat pure has some advantages over a “main” account, mains can still play without optimization.
A screenshot of Devious MUD, glitched.
Now “complexity” is a broad term with several interpretations1 and by itself is neither good nor bad. In a fight where you can choose between closing distance for a powerful attack or staying safe for a weaker ranged attack, that’s a desirable choice! So including rules for ranged attacks, makes a game “more complex” and still ultimately better.
But every additional rule creates real overhead at the table, which it might not in a computer game. Complexity is the root cause of pausing to look up a rule in play, or missing an obvious choice. In emulating the best parts of RuneScape we may find ourselves slowly adding more and more “cruft” to the game, rules and subsystems that are used too rarely to know intimately or with too little impact to justify their use. So I have identified three general philosophical tools to manage this complexity: maximize its impact, minimize its overhead, and hide it with automation.
Maximizing Impact
Let’s say we want to use a novel dice mechanic that takes some computation. If the only place I use that mechanic is in an optional spell, for example, not only will the work put into the mechanic only pay-off rarely, but unless the mechanic is a really intuitive fit, it will need to be re-learned every time. To maximize the impact of the mechanic, we want it to be as central as possible to the game or as similar as possible to other mechanics. (This may mean that an otherwise great mechanic ultimately doesn’t fit nicely anywhere.)
We can also shift more work to the GM’s side of the screen. This means that only one player (the GM) has to interact directly with the system, but the whole table benefits from it. This is already the case for many of the most complex parts of a game: NPCs, locations, etc. Sometimes it can even be “pre-compiled” during prep, like preparing a random encounter in advance. But here we see the problem quickly, the GM already does a lot of work. This solution works well until it doesn’t, when the GM either quits, or more likely just abridges the system.
A screenshot of a fight with Tzok-Jad, glitched.
Minimizing Overhead
One way to minimize the overhead of complexity is to hide it in a list of options. No player can choose every option, so no player needs to know everything. In D&D, this is why spells take up half of the player’s handbook. In OSRS items and equipment are the bulk of available customizations, and the things that vary most from player to player and task to task. Consequently, much of the game is about procuring, maintaining, and choosing the right items for any given task.2
This is also the heading under which I include organizational and information design improvements. If you have to look something up every time, but it’s simple and quick to do so, then you still have the complexity of the system, but you’ve reduced its cost.
A composite screenshot, glitched.
Automating
Bringing automation to the tabletop could be called a type of Minimizing Overhead, but I’m considering it separately because it has distinct trade-offs.
When the 3.X branch of D&D started to become unmanageable, tools like PCGen were developed to help. But these introduced their own new types of complexity (installing and navigating new software), and were relatively inflexible (difficult to expand or modify). A more promising example of automation might come from the world of virtual table tops, where for some games, the basic die-roller has been given the ability to parse character sheets and rules. The difference between roll 2d20kh1+2 and roll athletics with advantage is that the former is the player translating mechanics for the machine to execute, while the latter is the machine executing the mechanics directly for the player. Potentially, this could allow us to hide some complexity, like random events.
At a physical tabletop, this could be accomplished with a “helper” app, similar to those used for Magic: the Gathering 3 or any number of board games4, but this risks becoming cumbersome or gimmicky or both. Some of the big features of a tabletop game are its physicality and its minimalism. “Pen, paper, dice” conjures a certain romance that not even “and 500-page rulebook” can shatter, but which “and smartphone app” might just.
A composite screenshot, glitched.
The Illustrations
The images in this post were produced by a genuine glitch! Some version mismatch between imagemagick and libmagick on my system broke support for Fourier transforms in an interesting way. The first two should have been round-trip images, and the second two were reconstructed from the imaginary components of different images.
Against my natural inclinations, I am not going to define sub-types of complexity here, as I think most of the discussion is applicable regardless. Smarter people than I have thought more deeply about different types of game complexity and their implications.↩︎
This leads to a unique form a choice paralysis known as “bankstanding,” staring at your open bank of items, unsure what the optimal loadout for a planned task is.↩︎
It’s not necessary to use a helper app for Magic, but in competitive play they’re now commonly used to manage tournament overhead tasks, like seat assignments and match reporting. Even in casual play they serve to keep track of life totals and clarify some rules.↩︎
This is increasingly common for a variety of uses. Search for Planet X uses an app to randomly initialize the game state. Mansions of Madness: Second Edition uses an app to replace the need for one “evil” player.↩︎