The Elegance of d20 and D&D

[Odhanan]

By the same token, simple does not mean elegant. Call of Cthulhu has very simple mechanics. If you roll your skill or less on d100, you succeed. Yet, it falls down as soon as you try to add difficulty factors, opposed checks, and specific skills into the mix. Dodge leaps to mind - my ability to avoid your attack has nothing to do with your skill. If I have a maxed out Dodge skill, even Great Cthulhu can't touch me!

If I may, I'm a huge fan of the Basic RPS for its adaptability to emulate very different playstyles, even though it is a system that is created on a humanocentric scale (since a human can have 100% in a skill, and having skills beyond that mark becomes more of an abstract value than anything else). That's actually my main gripe about it, as opposed to a system where you have difficulties to beat. Stuff to roll, either dice or bonuses, can go up and up, and difficulties can be tougher and tougher, the system will be fine. Get two gods together using BRPS and get them to fight, and that gets really boring after a while.

Anyway, not the point. Point is, you refer to something like the Dodge Skill, Mike. Since the beginning of the BRPS this has been houseruled by modifiers on Skill checks according to the difficulty of the task. +20 to the score because it's an easy task. -20 because it's hard. For the Dodge Skill, you can apply staking maluses of -5 per dodging attempt for instance, because people tire themselves after a while, and then sooner or later, Great Cthulhu will hit you! If you add modifiers/difficulty factors, the example you're talking about is negated.

Same think about opposed checks. This actually has been modified under Hawkmoon - Nouvelle Edition in French, by some people of the zine Lance-Feu, in which you can use Skill percentages on the Table of Oppositions instead of ability scores to determine a percentage of success for the active character.

I know that's not the topic of the thread, but I thought some people might be interested by these remarks about the BRPS. Mike's original point remains though (or maybe that's what he meant all along and I missed it): by adding these modifications, the system becomes less streamlined, so less simple. So simpler doesn't mean it's more elegant, indeed.

 

[Prime_Evil]

Elegant does not mean simple. A wristwatch is an elegant device, yet it is far from simple.

Agreed. But isn't there a 'sweet spot' on the curve where it is possible to achieve a high degree of elegance with the minimal additional complexity?

Personally, I think that a truly elegant system is one that makes the inevitable trade-offs between complexity, playability, and flavour in a skillful manner. The best game designers balance these competing demands in a way that can only be described as elegant.

A computer program can be elegant, though it consists of millions of lines of code.

True...but in game design as in software engineering there is always the danger of unnecessary bloat. Sometimes a computer program needs to be a few million lines long in order to achieve a specific design objective...and sometimes only a dozen lines of code is sufficient. I believe that true elegance is achieved when the rules are no more complex than they absolutely need to be.

Although there is no direct link between complexity and elegance, I suspect that there is a direct link between complexity and robustness. The more complex a rule system is, the greater the risk that different subsytems will interact with one another in ways that the original designer did not anticipate or intend. Complex rule systems often have emergent properties that only arise during actual play and are as much a surprise to the original designer as they are to the players who are discovering them for the first time. The more complex a set of rules become, the more important it is to thoroughly playtest new subsystems before admitting them into the canon of 'official' rules.

IMHO, the d20 system has reached a point in its evolution where it will be increasingly difficult to add new features to the official ruleset without breaking existing features.

 

[Gez]

I was simply taking exception to some of the exampled ineligancies. For example the HD=CR=LA. There's simply no way to do it.

HD=CR=ECL. LA=0. I said I knew what I was talking about, didn't I?

And there is plenty of way to do it, otherwise all the character classes would be the same. In essence, what I said correspond to no longer consider the creature types (Aberration, Animal, Dragon, Construct, Elemental, etc.) as types, but as classes. Classes that would be open to a lot of customization, of course.

Then just keep the options roughly balanced, like for feats.

It would also simplify monster advancement and monster tweaking. And also monster "regression" à la Savage Species to play a powerful type of monster from level 1 onward, gradually getting all the powers of the creature. And it would open the way to a more elegant method for creating templated monsters such as half-dragon ogre or half-fiend minotaur: monster class multiclassing.

Sheer elegance.

 

[Hussar]

HD=CR=ECL. LA=0. I said I knew what I was talking about, didn't I?

Ouch, do you realize the massive amount of work that would require? For each and every creature, you would have to build it from the ground up. Essentially, each DM would be required to reinvent the wheel with each and every monster he used. Unless the new Monster Manual has twenty levels of every monster currently IN the monster manual.

In other words, to build a hippogriff, you would need to break down exactly what monster level it is, when it would get wings etc.

To me, this would be the opposite of elegance. By making monster creation extremely detailed, you would lose the ability to plug and play monsters.

I think Prime Evil said it very well:

I believe that true elegance is achieved when the rules are no more complex than they absolutely need to be.

 

[genshou]

HD=CR=ECL. LA=0. I said I knew what I was talking about, didn't I?

And there is plenty of way to do it, otherwise all the character classes would be the same. In essence, what I said correspond to no longer consider the creature types (Aberration, Animal, Dragon, Construct, Elemental, etc.) as types, but as classes. Classes that would be open to a lot of customization, of course.

Then just keep the options roughly balanced, like for feats.

It would also simplify monster advancement and monster tweaking. And also monster "regression" à la Savage Species to play a powerful type of monster from level 1 onward, gradually getting all the powers of the creature. And it would open the way to a more elegant method for creating templated monsters such as half-dragon ogre or half-fiend minotaur: monster class multiclassing.

Sheer elegance.

I'm going to strongly disagree with this. If the amount of other abilities is strictly tied to hit dice (and therefore hit points), creatures like the nymph, dryad, and the frail nivian of my homebrew would not exist--they would have to have more hit points and be as tough as a grizzly to have all their nature-oriented special abilities. What I think is more elegant is simply CR=ECL, as Upper_Krust has done. It allows for creatures with a lot of cussitude but little or no special abilities (colossal monstrous spider), as well as creatures like the nivian with almost no hit points but still possessing a lot of abilities.

 

[jmucchiello]

I completely disagree with this. The best games have mechanics that make you stop and think, "Damn, that's a cool rule."

Last night, I played a card game called Coloretto for the first time. There were at least three moments where players noticed a strategy or move that nobody had picked up yet, and each time they used it everyone at the table had a better understanding of and appreciation for the rules.

The limited structure of a strategy card game like Coloretto is such that you generally must have "cool rules" or nobody plays those games again. But RPGs with cool rules rarely get called out for the rule having been awesome. Usually the result of applying the rule is the front page story. When you find a game that lets you jump off the balcony, swing across the room on a chandelier and knock the evil giant into the fireplace without once having to make up or figure out a ruling. You cheer the jumping, swinging, and burning. You don't stand back and reflect on the elegance of the system. (You as a designer might do this but the general public won't.)

Still, having a better appreciation for the rules because you found a quirk within it lends to the rules beauty, not to its elegance. But the difference between beauty and elegance is fairly esoteric. A beautiful rule can be admired on its own. Elegance is the system of rules taken as a whole. The D20 DC system is beautiful. D&D is not elegant.

Every time that a D&D player gets to use Cleave, he's happy that he has those rules. Every time someone makes an attack against impossible odds and rolls that 20, they're exulting in the cool possibilities inherent in the rules. The most elegant rules are those that are so compelling that we forget they're rules and assume they're just cool parts of the game.

I still don't believe that elegance and fun are related. Even Chess has its fiddly bits (castling, en passant, 3-move stalemate) that have nothing to do with elegance and yet people still find chess fun.

Elegance and balance should be related and there lots of threads on the rules forum about whether Cleave (and Great Cleave) are broken or not. This is probably something of thread derail though. I'm not picking on Cleave specifically. I'm just pointing out that where you see elegance in Cleave, others don't necessarily see it as "fun". Another example where elegance is not related to fun.

As for feats, the structure of feats is very elegant. That doesn't mean that every specific feat is elegant.

I said this too. D&D is made up of specific feats that are inelegant. People still have lots of fun playing D&D.

I think that's changing. I think that in today's environment, feats need to be shorter, simpler, and easier to use. I agree that those feats are difficult to use and learn, but I think the solution is to build a better design model for feats.

Needing a better design model for feats sounds like it should be my argument that feats are inelegant. Maybe I lost the train of the discussion. I'll skip to a different point:

Elegance doesn't have anything to do with exceptions-based mechanics. Magic is an incredibly elegant game, and it's entirely built on exceptions. If anything, exceptions make a game much easier to learn. A DM or player need only learn the exceptions that apply to the monsters he runs or the character he builds.

Magic would be inelegant because all of its rules are based on exceptions except for one thing: All of the exceptions are spelled out in front of you when they apply. When I play a certain card, all of the relevant rules for applying that card are right there in front of me. There's nothing to look up (or at least that's how MTG should be). In D&D, whenever you cast a spell, you should go read it and then apply it. But that bogs the game down. MTG rules exceptions are terse, bite-sized modifications. D&D feats/spells are generally much larger. Your own example of Disarm I, II, and III has a MTG feel to it. And the finer grain allows for those rules to be terse, bite-sized morsels you can apply as needed quickly and painlessly.

I remember when our group played D&D 3.0 the first time we made a habit of reading the spells closely because "Yeah, we know what that does" was almost always wrong. The funniest one though was lightning bolt. I don't remember the details, but when we compared it to the 2e lightning bolt, we found out we had been playing the 2e lightning bolt incorrectly for the last 10 years. That just floored us.

Trying to get back on track, while your goal of 100% fun is nobel, I don't think it is really related rule elegance. People have fun playing complex games like Star Fleet Battles and I don't think I ever heard those rules called elegant.

(Oh, and what was the rule in Coloretto? I don't remember that game being very deep. Perhaps we missed something you found.)

 

[jmucchiello]

I'm going to strongly disagree with this. If the amount of other abilities is strictly tied to hit dice (and therefore hit points), creatures like the nymph, dryad, and the frail nivian of my homebrew would not exist--they would have to have more hit points and be as tough as a grizzly to have all their nature-oriented special abilities.

Or you just make them vulnerable to common weapons. Heck, to make a paper tiger, just have them take quadruple damage from normal weapons. This allows HD=CR for high magic, "low hit point" critters.

 

[Hussar]

Or you just make them vulnerable to common weapons. Heck, to make a paper tiger, just have them take quadruple damage from normal weapons. This allows HD=CR for high magic, "low hit point" critters.

Yes, but now you are adding complexity back in. You have made exceptions to the rules. Since the goal is elegance, adding in exceptions is a bad thing isn't it?

I remember when our group played D&D 3.0 the first time we made a habit of reading the spells closely because "Yeah, we know what that does" was almost always wrong. The funniest one though was lightning bolt. I don't remember the details, but when we compared it to the 2e lightning bolt, we found out we had been playing the 2e lightning bolt incorrectly for the last 10 years. That just floored us.

I've had a number of moments like this. Really does blow the mind.

 

[genshou]

Or you just make them vulnerable to common weapons. Heck, to make a paper tiger, just have them take quadruple damage from normal weapons. This allows HD=CR for high magic, "low hit point" critters.

Why not just stick with lower hit dice? A nivian is like a house of cards. She can't take much abuse of any kind before she tumbles.

 

[Prime_Evil]

I have recently been reading Eric Raymond's The Art of Unix Programming and it strikes me that many of the principles of good programming that he expresses could also be applied to RPG design. Based upon his work, I would like to propose the following 'laws' of RPG design. I'm sure that not everybody is going to agree with them, but they are a good starting point for a discussion of elegant RPG design. Let the flames begin!


Rule of Modularity
A rule system should be built out of simple parts joined together by common mechanics that are as intuitive as possible. The links between different subsystems should be as clean and standardized as possible – an approach that promotes flexibility and adaptability.

Ideally, final choices about how the rules work should be pushed as far toward the end user as possible. The rules should provide many options and support many different styles of play. GMs should be able to customize the core system based upon their needs and preferences.

The core game mechanics should be as small and tight as possible. Most of the rules that are not absolutely essential for making the game work should be moved to optional 'plug-ins' that GMs can use or ignore as they like without breaking the game as a whole.

Each new addition to the game system should do only one thing, but should do it well. When introducing several new features to the game, it is better to create a number of small extensions rather than one large one.

Additions to the core system should be self-contained. Game designers should try to avoid introducing dependencies between optional extensions.

By keeping the system as modular as possible, it is possible to modify one part of the rules without destroying the integrity of the game as whole.

Responsibility for maintaining game balance should be shared between the game designer and the individual GM. The notion that the game designer should build checks and balances into the core mechanics that make it impossible to break the system is ill-advised -- the same limitations that prevent you from doing really stupid things with the rules may also prevent you from doing really amazing things with the rules.


Rule of Clarity
Clarity is better than cleverness. It is better to have elegant and graceful rules than comprehensive rules that try to anticipate every eventuality. (Actually, it would be best to have rules that are both graceful and the cover almost every situation, but this seems to be an unattainable ideal. When in doubt, always choose playability over comprehensiveness).

Increasing the complexity of the rules to cover every possible situation that might arise is usually a bad trade off. Complex rules are more likely to harbor issues with game balance.

Furthermore, the more complex the rules are, the harder it becomes for future developers to build upon them.


Rule of Separation
As much as possible, try to separate crunch from fluff. Create generic mechanics and leave it to individual GMs to decide how to implement them in specific settings or genres. Don't tie any new game mechanics that you introduce too closely to a specific setting or genre.


Rule of Simplicity
Always design for simplicity; only add complexity where you absolutely must. However, note that simple rules are not necessarily the same as elegant rules -- complexity is only bad when it is unnecessary.

Actively resist rules bloat. Don't try and cram too many features into any given subsystem.

Always try to re-use existing rule mechanics rather than introduce new ones.

Encourage a design tradition that places a high value on coherence and simplicity.


Rule of Parsimony
Introduce new game mechanics only when it is clear that nothing else will do. Always weigh up whether the benefit of introducing a new rule mechanic will outweigh the complexity that the new mechanic will add to the system as a whole.


Rule of Transparency
Try to ensure that anybody can look at the game mechanics that you design and immediately understand what they do and how they do it.

Try to avoid the use of complex mathematics as much as possible. Follow simple mathematical progressions that are easy to remember. GMs should be able to calculate the probability of success or failure of any dice roll in their head.

The game mechanics that you design should be able to demonstrate their own appropriateness each time they are used.


Rule of Robustness
If your rules are simple and transparent, they will be robust.

Avoid introducing lots of additional mechanics that handle special cases. If the basic mechanics are robust, it should be possible for GMs to handle special cases on the fly.


Rule of Least Surprise
When designing a new rule, always do the least surprising thing. Avoid gratuitous novelty and excessive cleverness.

Don't require the GM or players to learn any new game mechanics if you can possibly avoid it. If you do need to introduce new game mechanics, make them as intuitive as possible.

Novelty is a barrier to entry; it puts a learning burden on the gamer, so try to minimize it.

Find similarities between the rules that you want to introduce and existing parts of the system that players already know about. Make the structure of your new subsytem mimic the structure of existing subsystems as closely as possible .

Also, pay attention to the history of RPG design. Over the past three decades, RPGs have developed conventions about how rule systems should work. You can use these traditions to your advantage in order to tame the learning curve.

The Rule of Least Surprise should not be interpreted as a call for conservatism in game design. As in other sorts of design, laws and guidelines are not a substitute for good taste and judgment. Always consider your tradeoffs carefully — and evaluate them from the end user's point of view.

Nonetheless, the bias implied by the Rule of Least Surprise is a good one to hold consciously, mainly because game designers have a tendency to be too clever for their own good.


Rule of Silence
Well-behaved rules do their jobs unobtrusively, with a minimum of fuss and bother. Bad rules require constant checking against rulebooks.

Don't have any more charts and tables than are absolutely necessary.


Rule of Economy
Gaming time is precious. As a general principle, it's better to waste the developer's time than the time of the gaming group.

Well-designed rules should treat the gamer's attention and concentration as a precious and limited resource, only to be claimed when necessary.


Rule of Extensibility
Never assume you have the final answer. Distrust all claims for “one true way”.

Nobody is smart enough to optimize for everything, nor to anticipate all the uses to which a rule system might be put.

Designing rigid, closed rule systems is an unhealthy form of arrogance.

Always leave room for future growth and development; otherwise, you may find that you are locked into unwise design choices because you cannot change them while maintaining backward compatibility.

Make it easy for future designers to plug new functions into the game system without having to rebuild the core game mechanics from scratch.

Embrace the concept of an open system that is extensible and contains lots of hooks for customization.