Rule Sets

[jburneko]

Hello All,

So, I've been thinking about rule sets and complexity.  Most people talk about systems being either rules-light or rules-heavy which ultimately isn't very useful because we can go around in circles all day about how light is light and how heavy is heavy.  So, instead I've created something that I think is far more useful in indentifying likes and dislikes about rule systems.  I propose that their are three general 'types' of rules.

Global Rules: These rules are 'in play' all the time.  That is they can effect any given player at any time.  The simplest example of a global rule is a system's base resolution mechanic.  Other global rule examples: include weapon damage, movement rules and situational modifiers.

Modular Rules: These rules are local to a given player and only apply to that given player.  An example of a local rule would be a spell.  Only spellcaster players need concern themselves with rules of a given spell.  Other modular rules include, skills, advantages and flaws.

These two are the big distinction.  Global rules can be thought of as 'Rules ALL Players should know.' and Modular Rules can be thought of as 'Rules that only Players who are affected by them need know.'

At first I thought this distinction was sufficient but then I realized that there was a third category.  This is more of a meta-category since I think ALL rules can be broken into either Global or Modular rules however some Global or Modular rules may also be Combinatoric Rules.

Combinatoric Rules are rules that can interact, combine or otherwise modify other rules.  Most rules in D&D3E are Combinatoric rules.  For example you may know what your attack bonus is to hit something but that may be affected by Spells, Feats and Skills or a combination of them.

I think Combinatoric Rules are worth their own distinction because I think they have a HUGE impact on rule complexity.  The more Combinatoric rules their are, the more ways their are to combine individual rules to sort of produce, sub-rules.

The reason that I think this vocabulary is useful is because it allows me to say exactly what I like and dislike about a rule set.  For example after I thought of this I can say with perfect clearity that I like a rule set that contains a small number of Global Rules, a large set of Modular Rules but little to no combinatoric rules.

This all arose because I was comparing D&D3E to 7th Sea in my mind.  Now across all the sourcebooks their must be just as many rules in 7th Sea as their are in D&D3E.  But I like 7th Sea and not D&D3E even though their individual rule sets are compariable in size.  Here's why:

7th Sea
Global Rules: Small.
Modular Rules: Huge.
Combinatoric Rules: Small.

D&D3E
Global Rules: Large.
Modular Rules: Medium.
Combinatoric Rules: Large.

Does this make sense to you?  It makes a lot of sense to me.  In 7th Sea I can quickly memorize the global rules and then carefully consider any modular rules individually.  If I Riposte, I need only worry about the Riposte rules because little to nothing effects the way Riposte works.  But in D&D3E I have to know a LARGE global rule set and any modular rule that applies to me has to be considered in light of a lot of other modular or global rules.  I can't just look at a spell and concern myself only with the details of that spell.  I have to compare the rules of that spell with a lot of other rule based factors that apply at the given moment.

Comments?

Jesse

[Mithras]

Hi,

I do agree that rules-lite/heavy is a complicated argument. I call one of my own games 'rules-lite' - what do I mean??

I suppose I mean that everything is resolved with ONE ROLL. That I can MEMORIZE all of the rules. That the task resolution and chargen fits on 4 SIDES OF A4.

These are the attributes I consider when I hear the phrase 'rules-lite'.

For me, 'rules-heavy' has MULTIPLE ROLLS to resolve combat and other tasks, the chargen, combat and task systems CANNOT BE MEMORIZED (although I swear my old DM had memorized the DMG and PHB!!) and these sections DO NOT FIT ON LESS THAN A DOZEN SHEETS OF A4.

But, to get back to your post - I suppose your definitions do hold water.

[Don Lag]

I think that discussing the subject might be interesting enough for those interested in RPG Metrics (if you can call it that), and probably useful for game designers if any conclusions or common vocabulary can be reached.

Personally, I'm not sure I agree with the distinction you're making: Global/Modular.

Specially because I think it's a rather blurry distinction. For example, would the rules for the "heal light wounds" spell REALLY be modular in most D&D groups? as far as my memory reaches everyone knew what to do in order to get that spell working (come in close to the priest, make sure he isn't busy, roll a d8 as high as you can).

It's both in this sense, and in a more abstract one that I believe that defining modular rule as one that "applies only to a given player" isn't very useful. The more abstract argument I can think of is that the mere existence of a rule means that I, as a player, must abide it. A given rule for Wizards could affect decisions I make in choosing and building a Warrior character.

Complexity of rules, in this sense, affect the players, not the characters, and in such a sense are always "global" IMHO.

I would propose rather an approach similar to computational complexity analysis. How many in-game decisions to you need to make in order to resolve an action, for example. How many rolls, how many decisions and how many calculations.

Stuff I'd look at in determining the complexity, or "weight" of a given set of rules:

What type of decisions/processes are required?
Division is more complex and slower than multiplication, which is more complex and slower than substraction, which is more complex and slower than addition, which is more complex (but sometimes faster) than referring a table. Choosing a higest die or counting successes seem to be universally accepted as the fastest mechanisms available under most situations.
Also, operating on "big" numbers (over 20 for most people) seems to be slower than small numbers (less than 10). Comparing big numbers doesn't seem much of a problem though.
ETC.

How many different "rule forms" are present?
If a single rule form is used consistently throughout the game, not only does it tend to avoid breaking the game into  "modes" (travel mode, conversation mode, combat mode, spellcasting mode,...), but also allows a quick and firm grasp on rule concepts and requires much less disperse memorization.
Rolling high d20's for some type of situation while roling low d10's for another in the same game seems more complicated than always rolling many d6 counting the 1's for every given situation.
Do some situations have gradual description depending on rolls/stats/whatever while having a binary form for others (Hit/Miss & Degree of success in a same game for different situations).

How many static and dynamic quantities are usually referenced?
Checking a value is rather fast, if it's on you're character sheet and you've memorized it. Crossreferencing a table isn't as fast, and scanning for applicable modifiers on a lista is rather slow. It seems that referenced values are always gradually memorizd though, decreasing there impact on play.
In AD&D for example, you usually need to check on weapon's THAC0 (static), target AC (static), circumstance modifiers (dynamic).

How much bookkeeping and on what scale?
Bookkeeping takes importance in as much as tracking a value requires operations being made (normally adding/subtracting, referencing and writing).
Palladium systems require lots of bookkeeping per melee of combat (mainly tracking action consumption). D&D requires keeping track of initiative orders each round.
etc.


This is just my idea of what one should concentrate on analysing in a game, to be anything more than proto-guidelines there should be some degree of quantification: how much more complex is adding two d100 rolls than referencing a d20 table?

[Ron Edwards]

Hi folks,

My contribution on this topic begins with search time and handling time, as mentioned in my System Does Matter essay - but only glossed over in the big GNS essay.

I'd certainly like a solid discussion of Jesse's points (and I do like the breakdown, albeit agreeing with Don Lag that some fuzziness often applies), search & handling times, and any other related matters.

Best,
Ron

[Don Lag]

An afterthought regarding global/modular rules.

Although I don't regard the classification as a useful one for analysing (analyzing?) game complexity, I think it does work as a first approximation to designing game delivery or game implementation (if you really want to push the software parallel).

What I mean is there is no relation between a game's "globality" or "modularity" and it's complexity. Most wargames are absolutely global rules-wise and nevertheless as complicated as AD&D-with-every-imaginable-rulebook which is HIGHLY modular.

I would also contest the definition of modularity, suggesting rather "rules that can be freely removed, interchanged or aggregated without disrupting other (modular) rules" than the definition originally proposed by jburenko.

Ron: where can one find the mentioned essay? I remember some discussion a long time ago about handling time and search time. I'm not very convinced that they are the most useful tipifications regarding a complexity analysis, but I'd rather read your essay than start making assumptions based only on the chosen terminology. (nevermind I just found it)

[ This Message was edited by: Don Lag on 2001-10-30 17:09 ]

[Don Lag]

Ok, read the essay. Now a few thoughts...

Searching Time seems to definately regard implementation issues. In a perfectly designed game, finding a certain rule definition should be quick and negligible. Of course a perfect design isn't possible, but I think there's something to gain from working with ideal cases at first then moving on to real cases which are more difficult to analyze.

Handling Time seems to be basically what I'm calling Complexity. And envolves exactly how much "time" needs to be assigned to actually playing out the rules.

Notice that "time" shouldn't be measured in actual seconds, minutes, hours. Rather (and maintaining the software/algorithm motivation), "time" should be measured regarding the number and complexity of operations needed to be performed.

i.e. Calculating damage doesn't take 2 minutes, but rather a dice roll a dice addition a stat substraction and a pencil recording.

Of course I'm not sure what level of detail in describing these operations is actually useful, but at least conceptually I believe it's better to consider "time" in such terms.

A final idea (after reading the essay):
One seems to be leaded to think Gamists, Narrativists and Simulationists have different complexity needs or tolerances. Would it useful to establish different "time" values to certain operations based on gaming style? adding dice has a different "narrative time" than it's "gamist time" Just an idea, you'd actually have to convince me if you buy into it :smile:

[jburneko]

Okay, I think I'm finally ready to respond to all this.  Don's posts have been taking the Analysis of Algorithms approach which largely has to do with actual execution of the rules as individual operators.   This relates heavility to the stuff Ron talks about near the end of System Does Matter.

*I* however was only talking in terms of player knowledge and MENTAL cross referencing.  You can think of my classifications as an attempt to answer this question: What is the minimum number of rules that an individual player needs to know in order to maximize his playing potential.  Now, the cynic might answer none, he can always just say what he wants to do and the GM can apply the rules as necessary.  Yes, this is true but that's not what I'm talking about.  I'm talking about the ideal situation in which tany one player knows perfectly all rules that are available to them.

To get back to Cure Light Wounds.  Cure Light Wounds is a Modular Rule because only those with access to the spell NEED to know the rule to use it effectively.  Sure, the Rogue may know that the Cleric over there can heal him and over time the Rogue player may actually learn the rules for Cure Light Wounds verbatim.  But the Rogue does not HAVE to know the rules for Cure Light Wounds to maximize his gameplay potential.

Now Cure Light Wounds or more appropriately Cause Light Wounds is also a combinatoric rule because to use it on an unwilling opponent I have to make a Touch Attack.  Now a Touch Attack has a very specific rule definition.  So in order to fully understand Cure Light Wounds I have to ALSO know and cross reference the Touch Attack Rule.

Touch Attack however is a Global Rule because any and all players may make a Touch Attack.

Now all this talk of minimal rule knowledge with maximal playing potential may sound very gamist but I'm not talking about actual in game goals or purpose of play.  I'm solely talking about what the ideal player needs to know to get the most out of a system.

Sorcerer for example consists mainly of Global Rules.  ALL characters have telltales, ALL charcters know ALL the rituals, the base resolution mechanic including the roll over victories rule apply to ALL characters.  It has a few modular rules, namely, the demon abilities.  And more interestingly very few combinatoric rules.  Binding is Binding is Binding and Shapeshift is Shapeshift is Shapeshift.  

Is this clearer?

Jesse

[Ron Edwards]

Jesse, you're making tons of sense to me. Keep rollin' with it.

Best,
Ron

[Valamir]

Yeah I think there is something here Jessie.  But I also think theres a couple of other factors such as Optimal Knowledge vs. Minimum Knowledge and Learning Curves.

To use your example a Rogue technically does not NEED to know how a cure light wounds works at a minimal level of knowledge necessary to get started.  But he definitely needs to know it to play optimally.  If he's used to games where Cures can happen at a distance, he'll be in for a rude awakening when he's down to his last couple HPs and the Cleric is 10 spaces away with a move of 9.

The warrior may not NEED to know how a wizards spells work minimally but to play optimally, he'd better know that that than enemy wizard's magic missiles hit automatically, and not to be within 30' of his own wizards Fireball in order to play optimally.

Thus, I would say that Global vs. Modular is perhaps not the best distinction.  Core vs. Specialized might be more useful terminology.  Cures and Spells aren't part of the Core rules a rogue needs to know, but are part of the advanced rules he will definitely want to learn eventually.

A game's learning curve then becomes a measure (necessarily an abstract one) that relates how much effort is required to go from a knowledge of the Core Rules to a full understanding of all of the Specialized Rules.

[Mike Holmes]

On 2001-10-30 17:32, Don Lag wrote:
A final idea (after reading the essay):
One seems to be leaded to think Gamists, Narrativists and Simulationists have different complexity needs or tolerances. Would it useful to establish different "time" values to certain operations based on gaming style? adding dice has a different "narrative time" than it's "gamist time" Just an idea, you'd actually have to convince me if you buy into it :smile:

Hmmm. If Ron's initial essay gave that impression (and I'm not sure that it does), he has since totally switched poles on that one.

What can be said is that it may be traditionally be true that Gamist and Simulationist games have more rules, and Narrativist games have fewer. But this is not necessary in any way. The trend that caused both of these phenomenons may not have created better games. More importantly, making rules lite Sim or rules heavy Narr, may just be what the doctor ordered. Or, in other words, we haven't seen any proof that these tendencies do make for better games. There are many here who are working on games that contradict this trend, and may end up being better games than those that fit the trend.

Ron would say that a game that is designed to cater to a particular audience should not have a preponderance of rules that do not cater to that audience. So lots of Sim ruled does not make for a good Narr game. But lots of Narr rules might.

Mike

[Mike Holmes]

I like Jesse's model. I can see his "modular" argument. It implies rules that the game could function without. If I drop Cure Light Wounds from the game, My priest can still select other spells. Remove the ability of Priests to cast spells, entirely, and now you've got some major redesigning to do. If I drop a particular feat from a class, that class may well still function (though possibly imbalanced). Chuck HP and you have to redesign combat.

I do disagree that combinatory rules are a bad thing, however. They are only bad if the global rules are so voluminous that the combinations cannot be intuited. So, if D&D had one rule for hitting, and no special global touch attack rule, then you wouldn't have to worry about the problems of this combination. Reduce your global rules to just a few, and the combinations become easy to figure out.

Universalis has three global rules. Very complicated rules, but once you understand them, you can do anything with them. If they didn't combine it wouldn't be much of a game.

Mike

[Don Lag]

*I* however was only talking in terms of player knowledge and MENTAL cross referencing.
(...)
What is the minimum number of rules that an individual player needs to know in order to maximize his playing potential.

My personal opinion is that this doesn't matter as much as rule consistency and simplicity; in the interest of good game design anyway. I'll just drop the subject on complexity in fear that it mnight actually constitute thread hijacking :smile:

And I still have a hard time understanding how a global/modular descrition dependeing on actual gaming habits doesn't lead one to conclude: this rule is global for group of players over here, modular for this other group over here, etc.

In my view, what rules "should be known" by each player is something that can't be prescribed by the author, but something that grows of actual gameplay and is unique for every group of roleplayers. Therefore, not a very good way to actually compare different systems. Perhaps different gamning habits?

I CAN see however, objective relationships happening between rules regarding mutual references. The more mutual references rules have among each other, the more coupled the rules are, and therefore less modular. The less mutual references rules have, the less coupling there is and greater modularity is acheived.

I'm not totally sure you don't mean this exact same thing, maybe just in a wording that I interpret differently. Mike Holmes seems to think we ARE making the same point:

I like Jesse's model. I can see his "modular" argument. It implies rules that the game could function without.

Anyway, I think it's a rather interesting subject.

[ This Message was edited by: Don Lag on 2001-10-31 21:40 ]