The Encounter Triad

In Magic Realm -- which I keep revisiting as the ur-type of exploration adventure games -- you'll find a bunch of standard fantasy archetypes like a dwarf, a wizard, a knight, and so forth. But there are also unconventional play options like a captain(?) and a pilgrim(?!) that would raise the eyebrows of a modern player. These guys don't make sense in the modern concept of a pseudo-RPG, because they're really not capable of winning most fights. They just wander around the board, sneaking past enemies and grabbing unguarded treasures. You can find a modern reviewer here who seems totally baffled by this play element, a flaw in what he obviously regards as an otherwise classic game.

What's happened between 1979 and 2013 is an alteration of the social contract of the game. In the 1970s, a central feature of what we now call "encounters" was the sorting operation of each encounter into three possible responses: fight, run, or talk. Being a good player meant being able to skillfully classify each encounter into one of the categories. Intentionally fighting something that necessitated a "talk" reaction or a "run" reaction was a failure of the player to play the game skillfully. A game that created penalties for neglecting to respect the triad was a good design.

A modern player coming into contact with this philosophy for the first time will be baffled by why (for example) the first edition of D&D used extraneous dump stats like intelligence and charisma to determine things like "known languages" and "reaction bonus", instead of doling out sexy combat bonuses. Or why the encounter tables had entries like "4d100 kobolds" without evincing the least concern about what your first-level cleric was going to do after bumping into 213 of the yappy little critters just a few miles outside the Keep on the Borderlands.

Today, 30 years later, the triad is mostly forgotten. Instead, the new social contract is that "fight" should always be an acceptable response, and that any attempt of the game to punish a decision to fight is a failure of game design. The ideal of game balance (which keeps an entire division of Blizzard employees working around the clock) requires that every player make a totally equal contribution to combat. If you said something like "my character's special talent is running away from fights", that would be a punchline to a joke. The turning point might have been the Roper encounter in the Sunless Citadel, where WotC tried to launch their 3rd edition with a module that featured an encounter that was totally unbeatable for a first-level party and suffered a broad backlash from the player base.

Let's think for a moment about avoiding a hostile encounter. Here's the original evasion table from TSR's Underworld and Wilderness Adventures, published in 1974.

This table is important for two reasons. First, it implements what I call a quadratic vs. linear scaling law. Note the size categories rise in a roughly quadratic manner, with bounding brackets that include values like 1, 4, 9, and 25. (Mysteriously missing: 16!) Size has a sublinear effect on evasion, with the decline in evasion chance falling roughly as the square root of the party size. This is a design element that would never pass muster today, where asking players to either (1) use a big table with ugly formatting, or (2) perform a square root operation. This would never make it past playtesting in today's math-phobic world. It's a compact example of diminishing returns, however, and so I've been using linear/quadratic scaling for my own design purposes whenever possible.

More importantly, though, it implements a sort of self-balancing procedure in the game. Early in the game, when you have a small scouting group, you'll have an easy time dodging goblin patrols. The more goblins in the patrol, the easier they'll be to dodge. As your group grows, you'll have a harder time avoiding their notice, and even some of the large warbands will start to chase you around the map. This kind of mechanism creates a protective in-game feedback loop that has some logical rationale: It's hard for a small group to run from a big one. More importantly, it moves the responsibility for balancing combat off the shoulders of the designer (or, in an RPG, the GM), and onto the shoulders of the players. You want a fair fight? Just keep dodging the unfair ones until you find it.

This style of evasion table survived in the RPG world for another decade before falling out of use. As it began to disappear, the concept of challenge level replaced it -- the idea that encounters automatically scale to reset their difficulty relative to player capabilities. Instead of finding huge camps packed full of orcs that you need to carefully circumvent until you can raise an army to conquer them, the new paradigm is that those camps just don't exist until your party becomes powerful enough to need them to exist. That approach has its place. But it doesn't feel right for an exploration game on an open map, which is by nature at the extreme sandbox-ish end of the spectrum. An excessive amount of dynamic scaling in a sandbox will place the entire game on a treadmill (cf. the Oblivion syndrome). Treadmills are the antithesis of exploration.

This encourages me to take the table above and clean it up (using a nice universal formula) for the company-scale exploration game I want to make.