Simulationists, Black Boxes, and 4e

[PrecociousApprentice]

Black box, white box, I'm the game with results.

See, to me it boiled down to the fact that the two editions went about design from opposite ends of the game. 3e said that it wanted to have an amazing number of options for world/monster/PC design. Then everything that is 3e as we know it was an emergent property of this design. The 15 minute adventure day, the 14 class 15th level PC, CoDzilla....

4e decided to design from the other end of things. There was a relatively specific output desired. Limiting the level of emergent properties was a design goal. They decided that at every level, they wanted a specific play feeling to be had. They wanted a certain amount of complexity, and didn't want further complexity to emerge. The the mechanics were designed to accomplish this. Some sacred cows were shoe horned into the system if they fit well enough, some were reconceptualized, and some were slaughtered. But it was output that mattered in the process.

In science this has become relatively standard practice for modeling systems. Create as simple a model as possible that accurately models a system, and don't worry about as much of the how and why as you do the what. Models =/= reality, but they have value in getting results.

Exception based design is just this process at a "fantasy system" level. "Reality" is the game fluff, the "model" is the game. Create a game that accurately models what your imagination wants, limit emergent properties (read excess complexities and game breaking combos) to a bare minimum. The result is that as a GM, your job is not to think "what can I create with these rules?" It is to think "I want this fluff in my campaign, what is the minimum number of game mechanics that I need to get this fluff."

As an interesting asside, Ebberon is the consumate front end design product in regards to world building, and is why it is the perfect 3e world. The design premise was "Given the full 3e ruleset, what is the logical world that would encompass all the emergent properties of that ruleset. Very interesting. The implied PoL setting in 4e takes the opposite approach. The premise is "Given that we want this sort of feeling in our game, what are the elements of worldbuilding that would have to be included to achieve that?" The design paradigms are opposite, and the feeling will be relatively opposite.

Which ruleset will appeal to which players? They both are toolboxes, but the toolboxes are designed to build different things. One is for creating emergent properties, and one is for output oriented design. Neither is superior, but I would say that those that prefer "D&Disms" are likely to like 3e, and those that have a specific design output in mind will prefer 4e. Either can be simulationist or not.

 

[Blackeagle]

A big advantage of process-response models is that they are robust in the sense that you can extend them to cover situations they aren't explicitly designed for. PnP RPGs need to work in a messy, uncontrolled environment, which means that rule-sets will be used outside of their design space.

True, but you have to live with the extra complexity all the time, not just when you're doing something outside the design space. One big advantage that RPGs have over scientific modeling is that we've got a DM. So rather than carry all that complexity around when you don't need it most of the time, why not just let the DM adjudicate what happens when you get outside the design space?

The second problem DnD has with black-boxes is that DnD is messy: making a good black-box is functionally impossible.

I have to laugh at this a bit. Black box modeling is used all the time in real world applications. Compared to the real world, D&D is incredibly simple and not the least bit messy.

The hard part is, and always has been, the specials (Ex/SA/SU). If your black-box covers those, its all good, but by that point you have an adequate model for a process-response design, which is more robust. This is why I am utterly unimpressed with the "4e monster design is easy" claim. We have seen no evidence, at all, that they have a system for designing good specials.

The thing is, 3e doesn't have a good system for designing specials either. Instead, they've got a big list of specials that you can choose from, (monster feats, special abilities from templates, abilities from class level) and a really complex, interdependent set of rules for applying them to monsters. The specials themselves are pretty much black boxes. 4e seems have pretty much the same sort of list (except they're all just powers) but rather than messing around with templates and class levels, you just slap the power you want on the monster (I'm guessing we'll see some guidelines for matching powers with monster levels).

 

[Alkiera]

Good analogy. Another way to look at it though is from from the Monster CR/Level perspective.

<snip>

As for the "Simulationists" (which I count myself among), whether 4E works for them or not depends on whether they're "big picture" or "fine detail." If you compared them to model train builders, a fine-detail Simulationist wants the train to work just like a real train works - only smaller. A big-picture Simulationist wants something that looks like a train and gets train-like results within the larger model city he's building. It doesn't matter whether or not there's an accurate transmission inside the train.

That's different than "process" vs. "black box" though because the level of detail the Simulationist needs to "have fun" is rather arbitrary. For instance, you could complain that 3E wasn't simulationist enough becuase it didn't properly model the digestive system of the Purple Worm. How fine grain do you need to get? Naturually the answer to that question is subjective.

I like what you said here.

Another way to look at it is the presence/absence of 'fiddly bits'. The 3.x approach to D&D had lots of fiddly bits; the skill point system, massively-multiclassed characters, all the different bonus types and templates etc. The conflict resolution system was relatively simple, but character building was begining to approach GURPS or HERO in difficulty, especially if you started above first level, and especially compared to earlier versions of D&D. For 4e, they seem to have stripped a lot of that out. Lots of the complicated bits are replaced by 1/2 character level. Skills are still around, but are now binary; you're either trained, or you're not; no needing X ranks in skill Y, and skill Z having a +2 if you have 5 ranks in skill R. Hitpoints are now a fixed number per level, based on class role, and vary between classes less than in 3.x. And that's just the PC side; on the monster side, there's a handful of tables (as best I can tell) indicating generic stats of monster type X(brute, soldier, artillary) by level. No longer do you build a monster by grabbing the 'standard array' or 'elite array', adding racial mods, then throwing levels of a class or pseudo-class on top, with all the PC skill/hp/BAB/saves fiddly bits. In 4e, you grab the row off the enemy type table with all the basic stats, give it a few attacks or other powers, a little racial flavor so it blends with others of its race, and you're done. You can tweak it some, sure, but there are guidelines for that, too.

The people who want all the fiddly bits are often simulationsists, but not always. Unfortunately for them, 3.x was really the only version of D&D like that, afaik. Unless you were a spellcaster, the majority of your character sheet in 2e was your inventory, the rest being about the size of a 4e monster statblock, or smaller.

What WotC seems to have done is rather like what Intel did with the Core CPU line. Instead of continuing from the rather different Penium4, they stepped back to the P3 design and worked from there to get the Core and Core2 processors. It seems like WotC decided that 3.x went too far from D&D's center, so 4th edition went back to a simpler era and moved forward from there.

 

[Lizard]

Another area where the 4e rules moved from a process-response model to a black box model is the grappling rules. In 3e, if you try to grapple someone, you provoke an AoO, you make a roll to grab, you make a roll to hold, you move into the target's space. Then next turn you're presented with a menu of actions you can choose from with detailed rules for each. It's definitely a process-response model with a nice little chain of cause and effect. In 4e you make one roll to grab an opponent. If you succeed you can hold on. Much more of a black box model, we don't care about all the details of the move, just whether you managed to latch on or not.

This simplifies the question of "grappled or not", but, judging from the admittedly stripped down rules in KOTS, leaves open other questions. A grappled person is "immobilized", which means he can't move. Fair enough. However, it seems he can take any other possible actions, including, say, attacking his grappler with a pike, casting a spell, removing an item from his backpack, etc. To my mind, this leads to all sorts of arguments at the table about what you can/can't do, and it falls into "Well, how are you holding him?" "I don't think you can grip him like that." "He can/can't reach his belt pouch.", etc. (As a rule, a grappled PC will probably cling to the letter of the rules and insist he can take any action but move; likewise, a PC grappling a monster will insist the monster can do nothing but try to escape. So it goes.)

Now, this can be handled 'black box' as well:"Here's what you can do when grappled. The End."

IME, given the choice between stopping the game to look up a rule, and stopping the game to argue about areas the rules are vague on, the first is preferable. 4e does look like it's going to require a lot more "Because I'm the DM, THAT'S why!" than 3e did.

 

[Kraydak]

True, but you have to live with the extra complexity all the time, not just when you're doing something outside the design space. One big advantage that RPGs have over scientific modeling is that we've got a DM. So rather than carry all that complexity around when you don't need it most of the time, why not just let the DM adjudicate what happens when you get outside the design space?

Which brings me to a bigger point which I will go into further down.

I have to laugh at this a bit. Black box modeling is used all the time in real world applications. Compared to the real world, D&D is incredibly simple and not the least bit messy.

A RL black-box model is tested/valid/used for a limited range of inputs. If you go outside that range, you need a better model. If you don't have one and can't make one, you do without. A DnD black-box model will be used outside of its design range. If that happens, well, see below.

The thing is, 3e doesn't have a good system for designing specials either. Instead, they've got a big list of specials that you can choose from, (monster feats, special abilities from templates, abilities from class level) and a really complex, interdependent set of rules for applying them to monsters. The specials themselves are pretty much black boxes. 4e seems have pretty much the same sort of list (except they're all just powers) but rather than messing around with templates and class levels, you just slap the power you want on the monster (I'm guessing we'll see some guidelines for matching powers with monster levels).

Of course 3e doesn't have a good system for designing specials. I don't believe it is possible to design a good system for specials (except, possibly, a set of tables a la Astral Construct, for an adequately broad definition of good). *Without* such a system, a black-box design for the non-specials isn't particularly useful: a process-response system for BaB/AC/HP etc... is easy and more robust.

The stuff I kept for the bottom:
Reading the 4e excerpts, it feels to me like 4e was designed, not for the Userslayers or UsersMs, but rather for the designers themselves. If as a designer you simplify your system to the point that DMs are forced to adjucate things on the fly, you have exported design decisions to the DM, making your life easier, and theirs harder.

Sure, designing monsters only for PC vs. NPC combat is easier than coming up with a full design, but it causes problems when players charm/diplomacize them. Writing modules, it is easier if you only need to place 4 magic items. It is the DM running the module who needs to figure out how to keep having the level+4 magic for 2 modules in a row being heavy blades with a party without a heavy blade user from causing party power problems from several levels to come. A job made more difficult because the level+4 magic item stands a good chance of being important to the plot or a central NPC.

If the Userslayers are non-agressive, there won't be any problem. But if they try to push boundaries, 4e dumps all the design work on the DM's shoulders. The designers aren't the ones who have to tell players that the NPCs somehow have awesome stats w/o lootable magic, but the PCs can't. The designers aren't the ones telling a player that his cool ability that does damage on a miss didn't hurt the minion, even though it does more damage than the other party member's hit that just dropped one. I could go on, but why bother?

Simplicity is only valuable if it doesn't create corner cases that drag games to screeching halts and overload the DM. If your play-testing groups have great DMs/non-aggressive players who are on board with the design decisions (other devs) you won't notice the potential problems. If you don't have such a situation though, you want a system with enough "why/how" explained that the DM can adjucate what happens when the players push the boundaries without excessive effort.

 

[Lacyon]

Are you referring to the experience total? Because the other numbers (attack bonuses, damage, etc.) do have a model for creating them. As someone stated earlier, that model is 1/2 level + appropriate ability score modifier (+ racial adjustments and monster role adjustments, probably).

No, this is not the model used for monsters and NPCs. For example, if you use the Customizing Monsters rules from the excerpt, attacks and defenses go up by +1/level, and damage goes up by 1/2 levels. However, even these models apparently don't apply when you adjust by more than 5 or so levels.

There almost certainly is a model for creating them, but it's a purely abstract game model - a [level X] [potency Y] [role Z] monster needs to have raw numbers pretty much in line with every other [level X] [potency Y] [role Z] monster to maintain game balance, not because of how many levels it's increased from its base.

The thing of it is, 3E's model of tying everything to the hit die/level is just as abstract, because hit dice and levels themselves are abstractions.

 

[ExploderWizard]

Black box, white box, I'm the game with results.

Thats all well and good, but the issue is defining "results". Exception based design can be good for modeling some mechanics but it can also lead to some in game effects which may not be desirable.

For example: Different laws of the physical universe for NPC's vs. PC's. On paper this can be great for getting the results you want. The "fluff" that comes with it may not be so great.

It it already a given that the PC's are special. Without any mechanical differences,PC's can sometimes act like they are "better" than everyone else in the world (Yes we all know they are but the actual character should not be so aware of this). If the rules of the universe explicitly state that yes, the PC's ARE in every way superior beings, then the result may be that NPC's are treated like dirt with actual bit of metagame justification.

 

[Rechan]

I"m pretty sure there is a system for monster creation. From the customizing monster article:

Boosting a monster’s level is easy. Just increase its attack rolls, defenses, and AC by 1 for every level you add. For every two levels, increase the damage it deals with its attacks by 1. The monster also gains extra hit points at each level, based on its role (see the “Monster Statistics by Role” table on page 184 of the Dungeon Master's Guide).

Also, it appears there's a system when it comes to minions/normal/elite/solo monsters:

Minions = 1 Hp.
Normal = Normal HP for monster role + level.
Elite = Twice HP of normal.
Solo = x4 or x5 normal; I'm not sure.

Elites gain 1 action point, solo 2.

It wouldn't surprise me if there's a table in the DMG that tells you how much damage a creature should do based on its level and role, either.

Then there's the CR of encounter = Total XP of encounter, in relation to Xp party can handle. XP of party can handle = level + number of PCs.

Either way, as you can see, there is a formula; it's just much smaller compared to 3.x

 

[Lacyon]

This simplifies the question of "grappled or not", but, judging from the admittedly stripped down rules in KOTS, leaves open other questions. A grappled person is "immobilized", which means he can't move. Fair enough. However, it seems he can take any other possible actions, including, say, attacking his grappler with a pike, casting a spell, removing an item from his backpack, etc. To my mind, this leads to all sorts of arguments at the table about what you can/can't do, and it falls into "Well, how are you holding him?" "I don't think you can grip him like that." "He can/can't reach his belt pouch.", etc. (As a rule, a grappled PC will probably cling to the letter of the rules and insist he can take any action but move; likewise, a PC grappling a monster will insist the monster can do nothing but try to escape. So it goes.)

You should know exactly what Immobilized means because the term is well-defined, AFAIK. You're non-teleportation movement speeds are reduced to zero (I think you also grant combat advantage to attackers, but I can't look it up right now). A player who wishes to achieve more than that needs to inflict a condition other than Immobilized - this may require an exploit or other power, or there may be a general rule allowing it. A monster that is capable of inflicting a stronger condition than that will either spell it out directly or direct your attention to a well-defined term.

There will only be arguments if one or more of the participants is unwilling to accept the rules.

Now, this can be handled 'black box' as well:"Here's what you can do when grappled. The End."

Your statements above have nothing to do with whether or not the process is 'black box'. In either case, you resolve the issue by looking up the terms.

If you invent your own terms, or modify the existing ones, you will need to make these new or modified terms clear to your players.

IME, given the choice between stopping the game to look up a rule, and stopping the game to argue about areas the rules are vague on, the first is preferable. 4e does look like it's going to require a lot more "Because I'm the DM, THAT'S why!" than 3e did.

I'm not sure why you think this. All the terms I've seen used appear to be well-defined.

 

[Blackeagle]

*Without* such a system, a black-box design for the non-specials isn't particularly useful: a process-response system for BaB/AC/HP etc... is easy and more robust.

I hardly think the 3e monster design system can be described as "easy". Building a high level monster from scratch is a hell of a lot of work.

I don't really think it's all that robust either, at least not in the way that matters most: CR versus actual power in combat. Building or modifying a high level monster is pretty much a crapshoot, you may end up with something that will TPK your party with ease, or something that's little more than a speedbump. Now, a lot of this has to do with 3e in general (particularly it's underlying math and some of the assumptions about what capabilities a party will have) rather than just the monster rules, but I don't think the monster rules help.

Reading the 4e excerpts, it feels to me like 4e was designed, not for the Userslayers or UsersMs, but rather for the designers themselves. If as a designer you simplify your system to the point that DMs are forced to adjucate things on the fly, you have exported design decisions to the DM, making your life easier, and theirs harder.

That assumes that adjudicating something on the fly is more difficult than applying the rules. At a certain point the rules become so complicated that it's easier to fudge it. Grappling is the classic example of this, a player or monster decides to start wrestling and the game grinds to a halt while the DM flips through the rulebook and players start trying to figure out how their current buffs affect their grapple check.

The designers aren't the ones who have to tell players that the NPCs somehow have awesome stats w/o lootable magic, but the PCs can't. The designers aren't the ones telling a player that his cool ability that does damage on a miss didn't hurt the minion, even though it does more damage than the other party member's hit that just dropped one.

Unless your players don't read the rulebook, then it seems to me that the designers are the ones telling the players these things.

Simplicity is only valuable if it doesn't create corner cases that drag games to screeching halts and overload the DM.

So, it comes down to the tradeoff between the complexity of a process-response model and the limitations of a black box model. Like I said, whether you like 4e is going to depend a lot on how you view this tradeoff.