Games-Within-Games

Here's an important aspect of early D&D I've been meditating on lately: the propensity for it to be an ongoing construction of games-within-games. Let's consider a few exemplary examples that spring to mind, starting with D&D and some of my most-favorite computer games:

(1) Dungeons & Dragons. In some sense, OD&D can itself be thought of as the “discovery” that the CHAINMAIL rules contained an even more interesting sub-game with its fantasy combat at the man-to-man scale (not to mention its even more refined system for jousting competitions). In the initial “White Books” you had both the standard dungeon exploration, as well as separate and distinct rules for large-scale wilderness exploration, castle-building, aerial combat, and ship-to-ship naval engagements.

What do I consider some of my most memorable D&D adventures? How about module X10, with its unique strategic-level world-warfare game (in parallel with PC-based diplomacy/adventure scenarios – including possible sidetracks to other X-series modules). Or M5, with a points-based diplomacy roleplay between imperial powers at the adventure's climax. Or even module S3, with its special system for trying to manipulate high-tech artifacts (among other things).

(2) Sid Meier's Pirates! Man, did I play a lot of this game on my cousin's Commodore 64 one summer. In some sense I consider it to be the near-perfect game – and, a lot of my design efforts wind up looking like attempts at replicating this classic. One of the strengths is that it has a completely different sub-game for each skill you might perform in your career as a privateer in the Carribean. Strategic sail navigation, taking a sun-sighting, fighting by cannons, personal swordplay on the deck, invading towns, choosing crew and cargoes, puzzling over map fragments, and wooing the daughters of prominent mayors, are all simulated in distinct sub-games. And almost all of them are both flat-out wonderful, and interface perfectly with all the rest (to the extent that only at this late date can I recognize them as sub-games at all).

(3) Mechwarrior. The original Mechwarrior was another game I played and re-played a whole lot of times. It's the first game I played that had both (a) “sandbox” play, and (b) “plot” based threads. The “sandbox” allowed you to progress as a mercenary captain, taking randomly-generated combat missions, improving your team and equipment over time. The “plot” (for lack of a better word) allowed you to follow up on clues that you were the member of a deposed royal family, and potentially win back your family's home. Some great (and dare I say Gygaxian) aspects of this: (1) you could play the mercenary sandbox indefinitely, (2) it was actually fairly hard to discover that there was a “plot” based mystery to follow up on in the first place, and (3) you still had to do some random mercenary missions in order to build up the strike team you needed at the end of the plotted scenarios. The exact time and sequence of events is impossible to predict in a game of Mechwarrior.

Now, some of this should be well-known to players of current computer game “sandbox” designs (Grand Theft Auto, anyone?), but since I don't play modern consoles, I can't comment directly on those. The thing I want to emphasize is that we don't lose the willingness to allow games-within-games in our classic tabletop RPGs.

Consider a few other examples from TSR/WOTC. In the old Star Frontiers Knight Hawks space combat game (by Doug Niles, who deserves his own blog acclaim), there was a brilliant scaling rule: for 15+ ships, use the coarse, Basic rules for the game; for 5-14 ships, use the more detailed Advanced rules; for 2-4 ships, use the Advanced rules with the individual characters' piloting & gunnery skills detailed. In the more recent d20-based Star Wars game, the spaceship rules were entirely done by analogy to the stock character-to-character system – which I was rather appalled to see when I read it.

Post-2000, there's been a bit of an over-reaction by my left-brained brothers and sisters, often times feeling that all activities in a particular game need to be abstracted out into one single universal mechanic. While this might be nice in theory, in practice I consider it be an abject failure (see the Star Wars example above). Even AD&D is not immune to criticism – when it converted overland movement rates from hexes to miles-per-day (so as to be usable with any campaign map scale; compare DMG p. 58 to OD&D Vol. 3, p. 16), it should have been emphasized that each DM really needed to manipulate those numbers and turn them back into spaces-per-turn on their personal map scale. Unfortunately, it did not. Here we see how frequently the attempt at abstraction interrupts the gamesmanship that we need at any scale of action.

Hence we have a few criticisms of the current branding of D&D: Action at different scales should have different mechanics that support the distinct flavor appropriate to each. Likewise, character classes that represent very different approaches to adventuring (magic vs. martial arts) should have different mechanics supporting each. We lose a lot when the game is reduced to a single kind of action scale (6-second moves on 5-ft squares), and the willingness to include sub- and super-games is prohibited (such as castle-building, tactical mass warfare, etc.) And, we have even more reason to avoid fetishizing character development, because we have to be willing to lose those characters abruptly if we play out an encounter at a larger scale (see 3E's Tome of Battle for the mangled result of being unwilling to allow for this).

Much of the addictive beauty of the original D&D game comes specifically from its flexibility as a model of developing games-within-games, both above and below the “normal” scale of action. It's a more interesting and more challenging enterprise than writing either "story" or "sourcebook" supplements, which add nothing concrete to our gameplay. But likewise, we should avoid being dogmatic, and try to engage our expansion systems only when it makes sense to do so (perhaps taking Doug Niles' SFKH as a canonical, concise example).