Idle Musings - D&D design scope

[Crazy Jerome]

That's an interesting way to break things down. Do you have a theory regarding how these might be applied to 5e? Or how modules can supply either based on the same core 'chassis'?

No theories, and precious little I would dignify with label of even hypothesis, but a lot of intuitions. I seldom live up to it, but I try to honor the Sherlock Holmes injunction against forming theories with insufficient data.

As an example, here is one of these intuitions: Process-driven mechanics are insufficient for D&D by themselves, always cause trouble when pushed too hard, but nonetheless must be present in some degree for the mechanics to be acceptable for the vast majority of fans.

If true, that has several implications for modules, including:

• It needs some carefully selected, carefully limited, process-driven mechanics in key spots, both core and key modules, in order for most fans to find the game acceptable.
• Some of these mechanics will be in places where if the whole module is all process-driven, it will destroy the modularity of the game. Thus other mechanics will need to be carefully mixed in.
• There will be at least a few modules for which many people will not be able to understand why other modules even exists, because as far as they are concerned, the "process-driven" module will be the only acceptable alternative. It will take discipline to rebuff their demands that the rest be excised, and such modules then be embedded into the rules for their convenience.

When I have a bit more time, I'll try to make some of this a bit more concrete by talking about how it applies to critical hits.

 

[chaochou]

I don't know if they would fit what you are talking about by narrativst rather than sim, but Barony (a.k.a Conrad's Fantasy/Rogue Swords of the Empire), Crimson Cutlass, and Avengers of Justice ("Good Guys Finish Last" and "Villains Finish First") all by Better Games might fit the bill.

<snip>

Whoa! I would xp (but can't, apparently) purely for crazy obscure rpgs!

ETA: I can't XP Crazy Jerome either - if someone could that would be appreciated!

 

[jrowland]

To push your intuition further away from forming a hypothesis, let me speculate wildly (huh? Did I just type that?):

Lets take your elements and imagine them in Venn-diagram space. Are they all separate diagrams? or is there overlap? *MY* intuition says there is overlap, but likely very thin in some cases (ie covering a narrow game mechanic, like an attack roll).

*MY* intuition is also telling me that the size of these diagrams can be small or large depending on the observers subjective interpretation, but the overlaps (or lack thereof) remain.

So....In designing a "one system to rule them all", so-called "Core" would places where these diagrams intersect. The modules would be ways to "dial" these diagrams smaller or larger to suit ones vision for the game.

To make matters worse, I am the type of DM who might run an identical encounter using different decision assumptions (process, fiction, etc) based on my mood, the story, the players at the table, the characters being played, the process of the adventure to date, etc. I am very fluid in that regard, so want a system where I go from the playtest version of play in a fiction-decision style for a number of encounters then switch to a process-decision for the BBEG, then a rules-driven for deus-ex machina that throws the campaign arc in disarray (it is possible...and easier for players to swallow a deus-ex machina when it it is rules driven).

I think you almost have to the "core" outlines for 3-5 game systems (in rough terms) for these styles, find the common ground, then create mechanics that are "dial-able" such that you can go from one style to the next.

Some people want a fiction-decision style social interaction and a rules-decision combat. Some want a fiction-decision combst and a rules-driven social interaction.

YIKES!

I don't think WotC R&D has looked at it this deeply, nor is committed to providing such a scaleable/adjustable ruleset. Likely some will hit the cutting room floor.

 

[Crazy Jerome]

So despite the airy nature of the preceding, I do have some concrete thoughts on why these things matter. Neverthless, this will manage to ramble. You've been warned!

Critical hits in D&D provide a good example.

To begin with, from the "proper abstraction" perspective, critical hits do not belong in most D&D rules, in any form. Strictly speaking, the attack roll is a scaled check with a binary outcome--you bypass the Armor, dodging capabilities, magic, etc. of the target, or you don't. If you so "hit" then you roll for (meaningful) damage. If you roll high for damage, that would be a critical hit.

Note that I am here talking about incoherent abstraction, and this applies equally whether the rules for attacking and damage are process-driven, result-driven, etc.

Perhaps it can be seen most readily from the process-driven perspective, though. Roughly, you roll your attack to see if your swing connects, and then you roll your damage to see how well (based on the weapon, your STR, luck, etc.) A critical defined as "double damage on a 20" or "max damage on a 20" or any other such variant is suddenly violating the fundamental abstraction to make how well you connect matter on a 20. Thus the "process" becomes "how well you hit usually doesn't matter, but occasionally it does," and this is perfectly capable of producing results that people find a bit counter-intuitive. You crit with a 2d6 attack and roll snake eyes, then next round barely hit with a normal attack for 10 damage, followed by an almost crit for 7 damage.

Most people accept this, because they like having critical hits, want critical hits tied in the fiction to thing closest to the swing (the attack roll), and are thus willing to overlook the hole in the abstraction to get what they want. Max damage on a crit is one of those bits of tap dancing to try to finesse the issue.

If you stop here, you can probably get away with it. But the pressure to not stop here is immense. Already we hear it with Next. "Max damage on a crit is boring." Is max damage on a non-crit boring? Then why is getting an extra dose of it without even needing to roll the damage dice suddenly boring? Abiltiies to crit more often, up the damage, etc. will all work mechanically, but every last one of them will expose the hole in the abstraction. In fact, some of them will only exist to try to deal with the hole, the prime exhibit being the 3E confirmation roll.

Don't find that compelling? Let's try this thought experiment (not a proposal). Go back to early D&D, when all weapons did 1d6 damage as a base. Then apply the modern understanding that it is generally satisfying to have some success about 2/3 rds of the time. OK, for all weapon attacks, don't use a d20. Roll your d6. On a 1 or 2, you miss. On a 3-5 you do that damage. On a 6, you do that damage, and this is considered a "critical blow." (Presumably, you can apply your STR mod, magic plusses, etc. after the resolution above, to damage only.)

This is a cleaner abstraction, given the stated design goal, but the process-simulation has been cut out. You swing a sword in the fiction. In the mechanics, you roll damage, ignoring low damage, narrating it as a miss. (I'm aware it is less clear-cut than that, but work with me.)

I can state with reasonable confidence that most D&D players would not find this cleaner abstraction pleasing. So roughly I think we've established both the necessity and some of the pitfalls of the process-driven attack roll.

 

[Crazy Jerome]

Next, I'll try to sketch some alternative critical hit rules that might fit the abstraction differently.

One way to keep it all process driven is to change what the attack rolls means, to no longer make it binary. A crude example--for every point you beat the AC, do an extra point of damage. Now a 20 on the attack roll is already skewed towards the upper end of the scale, by definition, and the closer you get, the more it is skewed. (There are obviously several better ways of handling this in play, as it in effect turns AC partly into a soak mechanism.)

Result-driven has a lot of promise. The attack roll is almost all process driven. But the damage roll sits in kind of a gray area betwen process and results. There is some straight randomness (follow the process to find out whether that halberd did a base of 1 damage or 10 damage), but also some built in results that people will generally approve (the 18 STR fighter usually hits harder with it than the 8 STR wizard who somehow picked up halberd proficiency).

In the original abstraction, attack is binary, damage is not. Embrace that, and make the critical all part of the damage expression, not something handled by the d20 at all. The easy way is to have open-ended damage rolls. Roll at or near max damage, roll again and add. A critical hit is now when that activates, and particularly gruesome criticals are when it activates more than once. (There are, of course, problems with this in the math, presented that simply.) The original nature of the abstraction is preserved, and not incidently you can now provide other process-driven changes to the attack roll without this having side effects on criticals. The roll is still outside the control of the players, so it is still simulation.

For the rules-decision approach, anything overtly gamist and tactical will probably work somewhat, whether more metagame or in-game. You get five crits per day. You must choose to activate before you roll the attack. If it hits, you crit. The rules-decision is all about appreciating the likely threat of the opponent, your chance to connect, and whether the extra damage will make a difference soon. For a less metagame approach, crits opportunities are unlocked by getting opponents into bad situations. You'll see these in facing rules, flanking rules, etc. if somehow tied to crits.

For the fiction-driven approach, it can also be metagame or in-game. You get action points that are only useful when something has set your character off (via another set of mechanics). Spend one after an attack connects, but didn't crit, and you turn it into a crit. Or more similar to what I believe @LostSoul does in his 4E hack (if I understand it correctly), you could have crits happen only when you've manipulated an opponent into a bad situation in the fiction. If in the fiction, he turns his back on you, you've got a chance at a crit.

Those are probably all weak examples, and I'm sure if anyone that makes it this far will note the very fine distinctions being glossed over. There could be, for example, very little difference between manipulating tactical facing rules versus pushing character manipulation into someone turning their back. The differences would depend on other rules that I've largely ignored.

 

[jrowland]

...
This is a cleaner abstraction, given the stated design goal, but the process-simulation has been cut out. You swing a sword in the fiction. In the mechanics, you roll damage, ignoring low damage, narrating it as a miss. (I'm aware it is less clear-cut than that, but work with me.)

I can state with reasonable confidence that most D&D players would not find this cleaner abstraction pleasing. So roughly I think we've established both the necessity and some of the pitfalls of the process-driven attack roll.

I agree most D&D players would not find THAT cleaner abstraction pleasing, but I think you could "fix" the current abstraction of the process-driven attack to avoid the pitfall.

It depends on what a to hit roll means, I suppose. If a to hit roll reveals one of three states: Miss, Hit, or Crit then the abstraction hole is 'evident if the damage for these states overlap: Miss might do more damage than a hit that in turn might do more damage than a crit. As long as the damage space for each "to hit" state is distinct and fits the fiction of the three states, there is no hole. Some "hits" might do more damage than other "hits", and if you want to narrate that you have to wait for the damage roll. But it likely won't matter: A slash for 15 damage and a slash for 10 damage don't really need further fiction distinction, but if you want to describe one as "cutting a crimson ribbon across the chest, exposing the fat beneath" and the other as "cutting a crimson across the chest", then at least the damage can guide the fiction without worry of falling into the abstraction hole

However, if a to hit roll means a sliding scale of success where a hit on a 15 is better than a hit on a 12, then essentially you have a sliding scale of states. This is better for the fiction, since you can describe a miss on a 9 as a hit with no damage but a miss on a 3 as hitting air. Same for hits. However, the damage doesn't fit this fiction. A hit on 12 can do more damage than a hit on 15. I can see no way to resolve this when each point in "to hit" represents a marginal increase in the effectiveness of the strike in the fiction.

 

[Crazy Jerome]

@jrowland , Right. Now think about what it means to the abstraction to get a situational, process-driven bonus to your attack roll. Let's say that you are fighting on ice, and everyone has to make balance checks. Fail, and you end up prone. If prone, melee attacks get a +2 to hit.

Depending upon how you set up the mechanics and flavor them, sometimes that means that a "bonus to hit means that you hit more often" whereas other variations mean that a "bonus to hit means that you do more damage" (on average).

I'm not saying that it can't be resolved somehow (or even several somehows, for different groups). I am saying that as soon as we mix mechanic styles, every single modifier or situation has to be examined in this light. And there are things you probably should not do, given a certain set of assumptions. Such as, if you think that people like the process-driven nature of the attack roll, don't remove it and embed it into the damage roll, as in my crazy d6 example. And we have worked backed to one of my intuitions on this topic, that in some of the discussions, people are not clear on the assumption, not so examining, and sometimes proposing ideas that will work about as well as my d6 attack/damage example.

 

[chaochou]

Don't find that compelling? Let's try this thought experiment (not a proposal)....

I can state with reasonable confidence that most D&D players would not find this cleaner abstraction pleasing. So roughly I think we've established both the necessity and some of the pitfalls of the process-driven attack roll.

I'm running with this thought experiment to see where it goes...

So, my first thought is - why roll 'to hit' at all? In an attritional system where losing hp means nothing other than losing hp, where the only dividing line is between hp>0 and hp=<0, why not just roll damage?

Then I think about armour.

Two men-at-arms are facing me. Both have 50hp. I have an axe which does 1d8 damage. One man is wearing full plate and shield, the other is wearing a loincloth. That armour has absolutely no bearing on the blow I can deliver to either man with an axe. I can do max 8 damage to either which is nothing. Maybe a nick, maybe a loss of fatigue or a clumsy parry, depending on how it gets narrated.

What 'armour' determines is how often I nick or fatigue the platemail man relative to loincloth man. In other words, armour - especially over the course of an adventuring lifetime when die rolls tend to the mean - is simply a hit point multiplier.

Then I wonder why different weapons do different damage. Going back to our two men at arms, my best hit with an axe can only nick or fatigue either of them. Can I not also fatigue them with a dagger thrust? If so, why is the dagger only doing d4 damage? What is the damage roll really expressing, if anything?

Right now, I don't have any hard and fast conclusions, but my instinct is that because HP and AC are, in design terms, empty space - that the other building blocks of D&D combat (to hit and damage) must be as well.

To-hit rolls and damage rolls might provide the illusion of process simulation, but I'm far from convinced that they're not just more blank canvas to paint whatever you like over.

A comparison with Runequest is forming in my mind, but I'm not typing it up

 

[jrowland]

I am thinking, that there is a "hierarchy" here though. Not in the sense of one style is better than another, but rather in concept-formation.

In a results-driven game, the process is less important, its the results that matter. So, conceptually, as long as the process satisfies the process-driven players, the results (whatever they are) will be "enough" for the results driven player.

So there isn't really a tension/trade-off going on between the styles. Only the outputs of one have to be satisfying inputs of the other. In addition, you have to be able to "dial down" complexity of any single style. The results driven game probably doesn't need a complex process for their satisfaction to be complete.

So I think it would help to identify which of these styles is a "base" style that informs the rest and build the hierarchy from there.

Fiction driven is the tough one. Does it come first or last? Or is it special in that it is the "bridge" making these styles connect? I can see starting with fiction then determining a process which begets a result, but I can also see a process that begets a result that begets a fiction.

My heart tells me fiction first. My brain screams "THATS CRAZY! start with process!" and unfortunately I think my instinct that the hierarchy is really a circle is probably true.

 

[jrowland]

...

Then I think about armour.

...

In 1E, Gygax had weapon vs armor tables as well as weapon speeds. I think there was an original attempt to balance weapons in this regard. Sure, daggers do 1d4, but you hit roughly twice as often as the axe that does d8. As editions spawned, the damage die remained but the other stuff fell away.

In D&D abstraction, as we understand in the more modern versions, we should probably have the damage dice be the same. That is not likely to happen.

4E did that in the powers and class features. For the most part, class features and power choice determined damage, not really the weapon. That is, a dagger-wielding rogue did roughly the same damage as the two-handed axe wielding barbarian over the course of their careers.