Wednesday, November 8, 2017

How to Write a Rule: Deconstruction

Deconstruction is the thing.  It makes sense to start with a rule that works, deconstruct that rule and see what we can learn from that.

Demonstrably, the most successful rule set in D&D has been the combat system, though many people argue that from the view that it is boring or that it fails to meet an imposed simulationist "standard."  Here I'm going to resist getting into the details of the various systems that have now resulted from multiple rule changes, concentrating on the rule system's fundamental precepts.  And let me just take a moment to add that these precepts do not just apply to D&D, but to most table-top role-playing games that seek to resolve a physical confrontation between combatants, regardless of the tech of weapon.  That is because the principles are universal: to hit something, there has to be a determination of hitting; once a hit has occurred, there has to be a determination of effect; because things are attempting to hit, or attempting to avoid being hit, there is movement, so there must be a determination of location and a comparison between these locations that awards opportunity; and because the world is cluttered, there must be a determination of obstacles, surfaces and physical restraints on movement, and therefore on the opportunity to hit, the chance of hitting and the effect of hitting.

I expect most of my readers already know this, but for those who are younger, who are perhaps less inclined to read, who just have not taken the time to explore the issue, I'll make the point that the combat system came before the role-playing idea.  Gygax, Perrin and others developed a set of rules they called Chainmail, enabling them to play a strategy game involving armor and medieval weapons.  Long before they published, they had hundreds of hours of experience with their own combats, with representing cardboard chits as game units.  Despite this low-level immersive quality, they could not help noticing themselves that when certain chits managed to survive battles with unusual frequency, they began to do something very human: they anthropomophized those chits.  The "survival" was nothing more than the argument I made a couple of weeks ago: that an audience full of standing people flipping coins, sitting down when they roll a tails, will eventually produce a phenomenon where one random audience member will remain standing, flipping head after head with astounding consistency.  This is called a statistical anomaly and happens with terrific frequency when a really large number of variables collects.

The combat system of Chainmail produced enough variables that it seemed unlikely to the participants that one particular combatant could survive so many battles, when random numbers seemed to indicate a more likely death.  The participants began giving these chits names and of course personalities ... and this in turn led to the creation of rules that would enhance the legitimacy of those personalities, a process that culminated in the crude, simple rule systems that produced the first role-playing games.

This is why I say that the rules surrounding the combat rules are demonstrably the best rule concept in the game, as no other rules that have come since then have succeeded in producing a similarly independent, wide-spread game culture out of the RPG phenomenon.  Some might argue "role-playing" itself, except that this is a result of the combat system and remains dependent on the combat system to support the consequences of in-game conflicts.

Why, then does the combat idea work?  It is sometimes argued that combat is based on a negative/positive result: one either wins or loses, based on the die roll, and that is a weak rule idea.  I have argued as much myself, on many occasions.  However, this is a gross simplification of the combat system as it stands.  Success does not rely on "a" negative/positive result, but upon scores of said results, as many as 20 to 50 results per round, depending on the size of the combat and the complexity of the given system.  This multiplicity of negative/positive results produces a statistical normality, in which anomalies occur that themselves produce unlikely and therefore exciting effects.

Let's look at the four points about all combat systems that I touched briefly:

  • Combatants are located in time and space; this location offers opportunities for strategy in the way they are free to shift, approach, collect into groups in order to improve their tactical superiority, back away, and play with how much time they have to prepare before actual combat occurs.  Since preparation of equipment and powers is an important feature in how combat is resolved, more time, won through careful movement strategy, greatly increases survivability.
  • Combatants must obtain an opportunity to attack defenders, whether through closing quickly and enabling the cut off of preparation by opponents, or using weapons that can be employed at a distance, so that defenders or would-be attackers can be eliminated at a safe distance.  Opportunity is mitigated by the ease of movement over the battlescape, or obfuscated by solid features or movable debris, so that the actual problem of obtaining opportunity when it is wanted is a strategic goal.
  • Once opportunity is obtained, combatants are forced to resolve the "wager" of attempting combat by actually rolling dice, a matter that can be modified by preparation and opportunity, but which is ultimately subject to the statistical probability and anomaly of random numbers.  Wagers are paid off by successful hits, while losses are applied to the reduction of further opportunity and the fact of giving the opponent a chance to determine if their wager to hit might pay off.
  • The effects of winning wagers, where a hit occurs, are then widespread and extraordinarily varied, challenging the struck combatant to survive the hit, have the opportunity to return the hit again and make the decision if "breaking off" from the combat isn't the better strategy.  Scattered along a line of a dozen combatants, each particular combatant's response to the effect of being hit has great potential for creating emotional immersion [as does the winning of successful wagers to hit].
To these we can add a fifth effect, those who will take this collection of combat results and choose to react immersively to these results, shouting that "I'm going to kill him!" or "Fuck, one more like that and I'm dead!"

This is the combination of rule-creation that we're vying to achieve ~ but not to worry, no one expects this sort of success, not even the original makers, who more or less stumbled into this because they had access to a number of technological improvements in the late '60s that enabled this breakthrough.  The combat game mechanic from D&D worked because a) it was logical in its use; b) it returned an emotional/visceral effect; and c) it was easy to adjust and expand, as desired.  It still is.

If we want to learn from it, what are the takeaway lessons?
  • Incorporate time and space into the rule structure, in a manner than enables the player to control these factors to some effect, without this being a random roll.  Ensure that the players have an opportunity to prepare in some meaningful way, that will promise an adjustment to the wager they will eventually have to roll in order to see if they succeed in what they're doing.
  • Where possible, produce more than one possible path towards success.  Just as combat includes elements such as missile weapons, spells, the use of animals, the structure in which groups interact together and so on, in addition to stepping forward and swinging, ensure that the rule system you're creating enables the player to create a strategy that doesn't rely on one single obvious course of action.
  • Minimize the effects of die rolls while maximizing the number of rolls, so that life/death or success/failure depends upon a statistical collection of results, rather than a single flat roll.  Obviously, in many cases, there should be a natural limitation to how many rolls are practical (or how many details can be meaningfully be rolled for); the goal is to find just enough rolls, with mitigating wagers, that make the activity interesting.
  • Make the effects of winning and losing wagers interesting.  Just as varying forms of combat has the chance of reducing hit points, ability scores, potential for movement, consciousness and location, seek to remove points from various stockpiles in the player's possession, including wealth, status, health and associates.
  • Produce a reward that encourages the players to return for the promise of that reward again and again.  The rewards of combat are varied and drive the entire game.  Even if your rewards are that phenomenal, try to make them as meaningful as possible for the player's experience.
  • Always direct your rule systems towards immersion.  If the player does not feel like they are actually experiencing the effects of the system as if they were real, to at least some degree, then your rule system needs more work.
To do this, you will need to construct a theoretical framework.  This is where we will begin with the next post.


Drain said...

Excellent post, I agree on all that you expound here.

Transplanting this design rationale onto other facets of play, let's say Stealth for a non-innocent example, I noticed from reading the rules at the time how you took pains to make space important rather than resort to a vague "opposition is just ahead that a'way, roll stealth to bypass". Good design choice.

Could this concrete definition of space then enable a greater variety of approaches, making for a better rule? The very limited sampling of our two Juvenis examples were rather straightforward, but not noticing doesn't mean it couldn't happen.

You obviously had to infringe on the "avoid single decisive rolls" in favour of the "enabling immersion through mechanics" point. Again, the rigth choice was made.

Alexis Smolensk said...


Yes, you're getting it.

Regarding the stealth roll, I don't think I've seen many cases where players have properly perceived the potential set up the roll allows. Mostly, I've seen parties revert to two standards: either most commonly, a) send one character ahead and let them scout, then bring everyone else forward, or b) just move forward and take our chances.

(a) provides data, since usually the recon isn't discovered, and is free to retreat, but it doesn't give the party much superiority if they then approach as a group anyway. (b) of course just relies on luck and party strength.

I'd like to see some experiment with the rule concept; I have some ideas myself, but I don't wish to explain them, as that might ruin the experience of another player, or ruin their chance of thinking of something better than what I have.

Ozymandias said...

I apply these concepts to my wilderness exploration rules with an interesting result: I need to expand options for rewards.