Exception-Based Design?

[aramis erak]

In essence this seems so broad as to encompass the vast majority of RPG design, which may be your point.

Not quite, but adjacent...
Most games do not try to avoid exceptions.
FATE Accelerated is about as close as any I've seen. Everything is done using stats and assets.
Cortex Prime comes pretty close, as does 2d20 Dune...

but my actual point is that the mechanically interesting parts of a design are usually exception based.

This is why I flagged the board game Cosmic Encounter, because it is probably the clearest example ever of a pure exception based design. Simple core rules, and everything unique about a given alien is simply a description of how it can break the rules. Later sets diluted that in a sense by adding additional subsystems, but in all cases every player gets to use all of them, they just provide added stuff to create exceptions in. This illustrates the complexity increase path for a system like that.

Agreed.

The thing about D&D 4E that appealed to me was that essentially, there were only a dozen total exception mechanics; they were reflavored extensively using a narrative limitation.

The thing about D&D 5E that annoys me is there are at least 2 dozen exception types, and by 8th level characters, I can't keep track of who can do what when as a GM...

2d20 Dune and STA have several key exception mechanics reused widely:
if (condition == True):
— reroll 1d20
— reroll (score) d20s
— gain extra momentum (spendable bonus successes)
— reduce the momentum/threat cost for ___.
— ignore some existing trait.
— Give an extra trait (general trait, advantage, complication, or value)
— allow using a different ability than standard.

STA also adds:
— reroll some d6's in a damage/progress roll
— add extra qualities to a given action
— add extra Effects to a given action

The core mechanics include the Trait...
If trait applies:
— making it harder: D+1
— Making it easier: D-1
— making it impossible: D=∞
— making the impossible possible (D set by GM)
`
About a dozen for STA.

One of the key elements is that the major metacurrency, Determination, normally cannot be spent. (That's its default condition.)

Several mechanics are supersets or subsets thereof:
STA Values are traits that specifically do making the impossible possible: you have to invoke a Value to spend Determination. Official answers noted that they're still traits, and can provide the usual trait benefits if not used for Determination spends.

In both, Advantages are traits that only apply when...
— Making it easier: D-1
— making the impossible possible (D set by GM)

...and Complications are traits that only apply when...
— making it harder: D+1
— making it impossible: D=∞

Not as straightforward as Fate, especially FAE, but still on the low number of exception types.

 

[Thomas Shey]

I never got any sense that there was an underlying cost system or something like that with 4e – only vague guidelines like "A 1st level weapon-based encounter power should do 2W with a minor effect or 1W with a significant one". What it did have was keywords and conditions galore, and often pressed these keywords into service in ways that felt a bit off, such as a power that instilled fear in its targets which was translated as a push.

I've seen indications at the time to the contrary, that there were, indeed, templates for creating powers and the like with semi-detailed guidelines for what to do. This came up in the context of this being slowly abandoned over time, leading to the more hit-or-miss late in the day powers and feats.

 

[Neonchameleon]

To spring-board off my "RPG Exhaustion" and "Character Creation App Dependency" threads, I'm thinking about something that adds to my frustration. The exception-based design present in games like D&D 4e, 13th Age, and Pathfinder 2e. Consider the following 1st-level fighter abilities - which are not limited usage (like a daily or encounter power). This is something the player (and GM) must keep up with constantly, along with likely 4+ other abilities ... all for a beginning character.

As a 4e DM part of the point is that I don't have to keep up with this. I can leave it all under the control of the player. And "along with likely 4+ other abilities ... all for a beginning character." How is this any different from a starting wizard? A starting wizard in 5e knows three cantrips and six first level spells of which they can prepare Int+level (likely 4) and cast two with one shot recharging on a short rest and both on a long. (And remember how complex wizards get by level 3 with two entire levels of spells, ten spells in the spellbook, and the ability to prepare six).

So the only difference you are pointing out here is that the fighter isn't a simple "I smash it" class - instead it's a tactical master of the battlefield. If that's not what you want 4e eventually included the Slayer and Knight.

The difference between standardised rules based design and exception based design as a DM is that under standard rules based design as a DM I am supposed to be able to know what all those spells do when I run a monster - and often the monsters will only have the name of the spell (or, worse, the feat in 3.X for literally every monster) in their stat block. But I don't have to know what the exception based design exceptions do; it's up to the player not to both of us to know that.

 

[Retreater]

How is this any different from a starting wizard?

Because now it's for every class, and every player. Usually a group has 1-2 players willing to keep up with details like spells, not 4-5 (IME).
And in the example of 4E, it's hammered home "you NEVER waste your turn on a Basic Attack." That means that every character's action is a complex litany of Effects, Conditions, and Interrupts, every time.

 

[Mustrum_Ridcully]

I used it because that's what the 4e designers used to describe 4e at the time. Perhaps the usage has changed in the 15-ish years since.

The way I understand it, in a "non-exception"-based game, Tide of Iron as a single power doesn't exist. Instead, there is a combat rule for shoving people around, and it states modifiers and actions you need to take to do so. Everyone acn do it, and this is pretty much the only way you can do it.
But in D&D 4, there are countless of powers that let you shove people around the battlefield, and they all work different from each other and knowing one doesn't tell you what the others do. The only common thing is that shoving people around has defined game terms (slide, push, pull in D&D 4), but how far, how many, how difficult, what resources or actions to use, it's all up to a particular power.

Spells kinda always worked like that in D&D. Particularly 3E D&D tried to unify it all - this is how you bullrush, trip, disarm people. OF course, there were still some exceptions - this is how you can do it without provoking attacks of opportunity, without penalty, or whatever, but they were the same options for pretty much all characters (and often also monsters).

The strength I noticed in play from D&D 4 is that you don't really need to read up some general rules in the rulebook (those parts tended to be so simple you learned them), and only the player using their power needed to read and apply it. As a GM, I often really enjoyed seeing how my players could surprise me with some of their abilities because I didn't know them all.
Of course, the drawback is that person needs to read it correctly, and needed to make all the "exception" thingies he had. They couldn't just say: "Okay, I am going to shove the goblin into the firepit", they needed to see if they had a power that could do it (or ask the GM to use the stunt rules, which kinda defeated the point of having all these neat power packages and also, being already overloaded with available options, most people don't even think of.)

I don't know if the terminology or understanding of the terms changed or are not what I think they are, though

 

[AbdulAlhazred]

I would call what @Thomas Shey is describing an effects based system.

Even though Rob Heinsoo uses the label ‘exception based’ for the 4e structure, I would respectfully disagree. 4e is a pretty textbook example of effects based design, where the powers are built from a library of effects in different combination. That became one of the criticisms of 4e - that the abilities across classes were too similar; a risk with most effect-based systems. And recreating the AEDU structure in effect based systems like HERO or GURPS is very easy, which I would suggest backs up this point.

Well.... Up to a point! So, I'd use my system, HoML, as a point of reference. It is largely based on 4e, but over time we drifted from AEDU to using power points plus levels of effect, plus kickers. The result is kinda similar, you have a bunch of choices that are fairly well-defined, and a limited ability to pop off a big effect, but AEDU did one thing really well, and that was insure there was a lot of variety in these keystone moves. Because you can't just use the same one over and over you have to plan more and be aware of what you are expending in a way that isn't true in my game. Now, in practice, some classes might as well have been spamming. Like, you can basically get 5 flavors of Twin Strike as a ranger. Yeah, they will have SLIGHTLY different marginal effects, but it still feels like doing the same thing over and over. Of course the ranger HAS been panned for this... So, that's an issue with my game, that would also be an issue with Hero system AFAIK (I can't comment on GURPS, only played it a few times WAY back in the old days).

Naturally the potential cognitive load of effect based systems increases based on the number and range of effects. For players this only usually matters in character creation when the player has to pick from the full range of possible effects; during play they only have to remember the effects their character makes use of. The GM may need to know them all, however, and that can be tricky.

The thing is, I don't think there's really a system that meets this definition of exception based, at all. Any system that offers any variety whatsoever is going to allow for some sort of 'bennie' or additional thing to happen that is not simply 'roll a d8 instead of a d6' or whatever that would fall under the most strict version of the exception based definition. I'm not saying there is NO RPG that could be described that way, there's probably one somewhere, but it would be hard to find! I mean, maybe TFT comes close, at least the Melee part.

 

[AbdulAlhazred]

I never got any sense that there was an underlying cost system or something like that with 4e – only vague guidelines like "A 1st level weapon-based encounter power should do 2W with a minor effect or 1W with a significant one". What it did have was keywords and conditions galore, and often pressed these keywords into service in ways that felt a bit off, such as a power that instilled fear in its targets which was translated as a push.

I agree, there was EDITORIAL 'rule of thumb' on what WotC published (and they definitely gave those guidelines to the freelancers that did most of the actual design) but there wasn't a 'point system' or something behind the scenes. And there are some powers that break the rules in 4e, that just are not like other ones. Often one would appear like that and it would then become a sort of template for others, but it was quite possible to do different stuff. This is quite different from a game like Champions where basically all the powers end up being reflavoring of 4 or 5 effects, and the point system strongly encourages almost identical designs because there is high synergy between certain things. It is also an easy system to 'game', and thus the GM is required to play police, not a situation that would work within 4e's paradigm.

The upshot is that a formulaic power system was not really a 4e option. Frankly this problem has been the main issue with HoML, which makes it not really a publishable game. I'd actually have to rip out the whole power system and go back to bog standard 4e AEDU in order to make it workable for any group where the GM wasn't expected to be in charge (or else the players are totally in charge and running it like a neo-trad design). Actually @Manbearcat called it 'neo-trad' and I think he's kind of right, it mostly comes out that way, though MY game doesn't when I run it, but that's not strictly system. This is a big factor in game design, some things seem like they'd be great ideas, like a more Hero-like approach to powers in 4e, but THEY DON'T WORK given the game's actual intended play style. At least not without a strong bossy GM or something like that.

 

[Manbearcat]

I agree, there was EDITORIAL 'rule of thumb' on what WotC published (and they definitely gave those guidelines to the freelancers that did most of the actual design) but there wasn't a 'point system' or something behind the scenes. And there are some powers that break the rules in 4e, that just are not like other ones. Often one would appear like that and it would then become a sort of template for others, but it was quite possible to do different stuff. This is quite different from a game like Champions where basically all the powers end up being reflavoring of 4 or 5 effects, and the point system strongly encourages almost identical designs because there is high synergy between certain things. It is also an easy system to 'game', and thus the GM is required to play police, not a situation that would work within 4e's paradigm.

The upshot is that a formulaic power system was not really a 4e option. Frankly this problem has been the main issue with HoML, which makes it not really a publishable game. I'd actually have to rip out the whole power system and go back to bog standard 4e AEDU in order to make it workable for any group where the GM wasn't expected to be in charge (or else the players are totally in charge and running it like a neo-trad design). Actually @Manbearcat called it 'neo-trad' and I think he's kind of right, it mostly comes out that way, though MY game doesn't when I run it, but that's not strictly system. This is a big factor in game design, some things seem like they'd be great ideas, like a more Hero-like approach to powers in 4e, but THEY DON'T WORK given the game's actual intended play style. At least not without a strong bossy GM or something like that.

I think it here is referring to 4e D&D?

My sense of the game is as follows:

* Player-authored Quests + Goal & Stakes-driven noncombat conflict resolution w/ table-facing resolution governed by fail forward/change the situation-state upon resolution + the the deep theme & premise loading of the cosmology/PoL setting/and all the various build components (Background, Class, Theme, Paragon Path, Epic Destiny) + the default transparency and player control of the game engine = either Story Now design and priorities or NeoTrad design and priorities. The pivot points here are the difficulty dial and stakes. If encounters are difficult and stakes/consequences/fallout can dynamically change the course of play, then its a Story Now meets Gamist hybrid. If the difficulty and stakes are muted (with similarly muted dynamism of situation-state and gamestate), then its NeoTrad.

* Alternatively, a GM can "say no" (which you're specifcally told to "say yes") and control the Quest system, forgo Skill Challenges, and try to work against all of the other aspects of system and ethos that really make Traditional play rather difficult with 4e. Good luck (but why would you when there are plenty of better alternatives?)!

* Alternatively alternatively, the group can basically play and prep a Pawn Stance action-adventure, Gamist crawl. It won't work like Classic D&D crawling, but it absolutely works swimmingly in its own way.

---

But 4e trivially works best (and was clearly intentfully design) as a Story Now meets Gamist hybrid or a NeoTrad experience.

 

[Neonchameleon]

Because now it's for every class, and every player. Usually a group has 1-2 players willing to keep up with details like spells, not 4-5 (IME).
And in the example of 4E, it's hammered home "you NEVER waste your turn on a Basic Attack." That means that every character's action is a complex litany of Effects, Conditions, and Interrupts, every time.

So what you are saying is that the problem is that the Slayer and the Elementalist Sorcerer weren't in the PHB? This isn't a problem with the approach just the implementation.

 

[dbm]

I don't think there's really a system that meets this definition of exception based

I haven’t proposed a definition of exception based design?