K wrote:
We could have a whole separate thread about how the DnD skill system is bad, but the overall point is that the monster design is affected by all of your other systems and the designs that we know now are because the monster design was deformed to fit into all of these other, poorly working systems as an afterthought.
I certainly agree that a system that makes a big difference between a +1 and a +2 is often too granular. I personally prefer a system that does something like
Trained +5
Expert +10
Further, having some 'maximum rank' that is less than +23 may help keep everyone on the RNG.
But how would you do skills?
What do you think should allow 'training' to improve? Should stealth and perception be part of the trained skill system, or should they just revert to untrained and untrainable ability checks?
For DnD? Cap infinite progression, allow skills to mimic spell-effects at a certain arbitrary point (but preferably much sooner than epic) and make basic tasks achievable (albeit with a low chance of success) without needing any skill ranks at all.
For example, Move Silently should just straight up give you invisibility at a certain skill rank. For 3.5, that means thieves should be able to mimic invisibility as the spell at 3rd level. Possibly sooner. And unlike the spell, it's not subject to spell resistance or a saving throw (perhaps a spot check) and has infinite duration.
So that makes the rogue's shtick (aside from sneak attacks) be the super-sneaky dude without having to bone the people who don't have move silently or the invisibility spell. You actually want to send the rogue out on scouting missions and save the wizard a spell slot, since they have functionally equivalent abilities.
Diplomacy could also work the same way - the bard (from first fucking level) can mimic a charm person. And etc.
Capping infinite progression at a certain level is needed, as well, because of RNG issues. But mundane tasks shouldn't be DC 20 or some retarded shit like that, so people can at least try to do them. This does mean that the rogue is going to auto-succeed at sneaking past mostly everything and the bard is going to auto-charm mostly everything. Who gives a fuck.
From a general perspective, I'd like to simply have skills be a thing you do, with a bonus for a specific usage of the skill you're good at, and have four or five or w/e 'ranks'. Trained, Skilled, Journeyman, Master, etc with specific set bonuses. This entails keeping the RNG in line with the bonuses, but that should be a general fucking rule of game design anyway.
Then, once you have absorbed the lesson, that your so-called "friends" are nothing but meat sacks flopping around in the fashion of an outgassing corpse, pile all of your dice and pencils and graph-paper in the corner and SET THEM ON FIRE. Weep meaningless tears.
Aside from the MANY other things wrong with 3.x stealth in particular one of them is the simple way it makes you roll again, and again, and again.
You want to sneak past some guard? Fine, roll EVERY TURN at half speed to get past him.
The problem with that is that yeah, suddenly auto-success or auto-failure IS more attractive, because if we try a granular system that even generates a small chance of failure on an expert then sooner or later they fail because they have to roll easily 5 times in succession MINIMUM to sneak past most observers.
Not to mention multiple observers all get a chance to roll that natural 20 spot check and screw you each, so there is even more RNG screw you built in.
The basic problem is the mechanics try to model it as an attack roll style set up, without acknowledging that unlike attack rolls you cannot really afford even one miss.
There needs to be a model that even if it comes down to rolling, needs to require LESS rolling to generate ONE stealth success.
I've gone out of my way to attempt that in my own system with Stealth Persistence (only roll again if the roll gets harder) and limited observers (only roll against the best observer).
Thing is, I'm not sure that is good enough. And that's probably close to the minimum you could bring the concept to before you start considering binary autosuccess autofail mechanics.
Phonelobster's Self Proclaimed Greatest Hits Collection : (no really, they are awesome)
The 'roll against best observer's concept is one I use in Shadowrun, ad it works pretty well.
Then, once you have absorbed the lesson, that your so-called "friends" are nothing but meat sacks flopping around in the fashion of an outgassing corpse, pile all of your dice and pencils and graph-paper in the corner and SET THEM ON FIRE. Weep meaningless tears.
In our system, we generally roll once unless circumstances change. Observers may also force a re-roll (example, you were standing in plain sight but you managed to hide - they know you're there so 'look for you' and get a reactive roll to notice you even if you succeeded before).
We've tried to make it a static DC based on the observer, with a bonus for each additional observer. So if you're sneaking up on a group of 20 people, you're going to make 1 roll but it is harder than sneaking up on a group of 4 people.
PhoneLobster wrote:Aside from the MANY other things wrong with 3.x stealth in particular one of them is the simple way it makes you roll again, and again, and again.
You want to sneak past some guard? Fine, roll EVERY TURN at half speed to get past him.
The problem with that is that yeah, suddenly auto-success or auto-failure IS more attractive, because if we try a granular system that even generates a small chance of failure on an expert then sooner or later they fail because they have to roll easily 5 times in succession MINIMUM to sneak past most observers.
This is also a problem with mass number of stealthers or observers. When you have 8 orcs trying to spot you, even if the orcs have weak perception, One of them is bound to roll high. So you're either unspottable, or you almost always get seen.
It also makes sneaking in as a group almost impossible, because someone is going to botch their roll.
Given all the dice rolls involved, if it's at all possible for stealth to fail, it will fail, even if the chance is very low on any individual roll.
I think it'd make sense to have an opposed Stealth/Perception roll periodically, but with how much a check succeeds by giving a length of time (or distance perhaps, under some circumstances) before the next check.
Multiple opponents I'd probably lean toward DDM's method (one roll, gang-up bonus). If using a dice as variable as d20, multiple rolls gives the beholder (non-D&D usage of the word) too much of a bonus.
vagrant wrote:Cap infinite progression, allow skills to mimic spell-effects at a certain arbitrary point (but preferably much sooner than epic) and make basic tasks achievable (albeit with a low chance of success) without needing any skill ranks at all.
Well, the first question is what skills are supposed to achieve? And I don't mean what does {SKILL X} do on a success or fail, no I mean what design purpose is your skill system serving.
If you want skills to be one of the primary means by which characters distinguish themselves from other characters, then you want a lengthy skill list to allow for a lot of differentiation. Conversely, if you want skills to be one of the minigames via which characters can progress through plots, then you need to keep the skill list small enough that most PC groups will have reasonable odds of having multiple characters with at least reasonable chances of succeeding on most skill rolls.
If skills are just kind of an afterthought (like 2e AD&D's NWPs) then you want a game which can run without needing to use them. If skills are something certain character types are expected to bring to the table (Shadowrun, HERO), then skills need to be as viable as the alternatives other characters can bring to the table (Cyberware, Superpowers, etc)
....And what it boils down to is that pretty much no edition of D&D to date has had designers who even asked these questions, let alone answered them.
"But transportation issues are social-justice issues. The toll of bad transit policies and worse infrastructure—trains and buses that don’t run well and badly serve low-income neighborhoods, vehicular traffic that pollutes the environment and endangers the lives of cyclists and pedestrians—is borne disproportionately by black and brown communities."
The core benefit of skills was supposed to be opening up basic stuff to everyone. Mages were supposed to stealth and fighters would know spellcraft and thieves would sing and all kinds of cross-class hijinks was supposed to be a core conceit.
Clearly, that did not happen. It feels like someone had a role-protection bitch fit in the middle of design and decided to tank the whole idea, but no one had any alternative and so they had to ruin all the numbers so that no one could use the system for anything.
Let me take as an example the 3.x skill 'balance'.
On the one hand, not falling down is something that everyone should have a reasonable chance to succeed on. If that's the case, rather than making it is a skill, one can just make it a simple Dexterity check (and cap the maximum DCs to make it work). If the highest DC is 15, anyone with god-like balance can walk across the marble-strewn floor by taking a 10.
But if someone wants to be 'really good' at something like that, should they be able to train?
A number of 3.x skills might passably be replaced by simple ability checks if the DCs don't scale in a crazy way. But if someone wants to specialize, I can see why they'd want 'extra' skills with that.
For example, if you make Handle Animal a simple Charisma check (and the DCs don't go past 15 or so) you're going to have a Beast Master type character that wants to succeed on every check with a 2 or better. They want to get 'expert' in handling animals so that, not only can they do anything everyone else can do with the skill, they can do it better/faster etc.
Normal riding might not deserve ranks, but trick riding?
How do you balance the need for everyone to be able to handle everything versus the person that wants to be very specialized/very successful at that type of task?
deaddmwalking wrote:Let me take as an example the 3.x skill 'balance'.
On the one hand, not falling down is something that everyone should have a reasonable chance to succeed on. If that's the case, rather than making it is a skill, one can just make it a simple Dexterity check (and cap the maximum DCs to make it work). If the highest DC is 15, anyone with god-like balance can walk across the marble-strewn floor by taking a 10.
But if someone wants to be 'really good' at something like that, should they be able to train?
A number of 3.x skills might passably be replaced by simple ability checks if the DCs don't scale in a crazy way. But if someone wants to specialize, I can see why they'd want 'extra' skills with that.
For example, if you make Handle Animal a simple Charisma check (and the DCs don't go past 15 or so) you're going to have a Beast Master type character that wants to succeed on every check with a 2 or better. They want to get 'expert' in handling animals so that, not only can they do anything everyone else can do with the skill, they can do it better/faster etc.
Normal riding might not deserve ranks, but trick riding?
How do you balance the need for everyone to be able to handle everything versus the person that wants to be very specialized/very successful at that type of task?
So far TRD has considered 'Talents' which are basically sub-skill specializations built in to broader skills. You basically unlock the Talents you want as you rank up. Full details still being jiggered.
Josh_Kablack wrote:Well, the first question is what skills are supposed to achieve? And I don't mean what does {SKILL X} do on a success or fail, no I mean what design purpose is your skill system serving.
The point of a skill system, as far as I'm concerned, is to answer the question "What happens if I do X?" where X is something that someone will inevitably want to do in a non-combat situation. Conversely, actions that will never come up in play don't belong in a skill system (that sounds obvious, but you might be surprised).
Note that what out-of-combat actions are "inevitable" depends on the genre, of course. In an action movie-style genre, jumping over stuff, hiding, or chasing someone are the type of actions that have to be covered, and in a mystery genre you have to cover actions like looking for clues, intimidating people or trying to tell if someone is lying.
K is premature in saying that a skill system is meant to "open up basic stuff to everyone"; without a skill system, you can't even say whether an action is open or not open. E.g. you can't even say that a wizard is bad at sneaking if his success/failure is covered by "ask your GM".
vagrant wrote:Cap infinite progression, allow skills to mimic spell-effects at a certain arbitrary point (but preferably much sooner than epic) and make basic tasks achievable (albeit with a low chance of success) without needing any skill ranks at all.
I read through the thread linked, but...honestly, I don't see a problem with skills being superpowers. Maybe it's because I don't think mundane in DnD necessarily needs to correspond to realistic (as in actual reality). Having skills mimic spell-effects would also make skill-characters viable concepts - you could make a workable character whose concept actually is 'gets shit-ton of skills.'
Or you set different skill progression rates - so while wizards, fighters, and rogues can all train 'Sneak' or 'Move Silently', the wizard is capped to 'Trained' rank for 'Sneak', the fighter is capped to 'Journeyman' rank, and the rogue can hit 'Master'. I see no problem with specialists being far better at a skill than someone else - its just that mundane scenarios, like 'Sneak past the Baron's guard dogs' should be doable by everyone. But something like 'Have the rogue sneak into the castle and assassinate the guard captain before opening the gates' should be restricted, or at least easier to achieve with a skillful character than a different one.
My issue with the DnD skill system has always been 'Why bother training skills when a wizard can duplicate the effects and do it better at the cost of a single spell slot?' The above, I think, fixes that.
Then, once you have absorbed the lesson, that your so-called "friends" are nothing but meat sacks flopping around in the fashion of an outgassing corpse, pile all of your dice and pencils and graph-paper in the corner and SET THEM ON FIRE. Weep meaningless tears.
3e's skill system was just supposed to be a series of small evolutionary improvements over 2e.
In 2e, a non-weapon proficiency gave you a check value that was often the same as or worse than a straight attribute check. Depending on how your DM ran things, it was totally conceivable that it would straight up be better to be untrained than trained when attempting an action (see my gauntlet to shadzar about sneaking up on a sleeping Ogre, which he has still refused to pick up for obvious reasons). So in 3e your skills are expressed as a bonus to your ability check, and all actions are by definition better when trained than when untrained. They even set the basic bonus at first level to be the difference between taking a non-proficiency penalty and not (+/-4) just to drive the point home.
In 2e, you gained more non-weapon proficiency slots, and you could spend them on getting new proficiencies or on improving one you already had at the rate of +1 per slot. This left some people maximizing one proficiency and other people spreading out, and in any case no one really knew what they were supposed to do. So in 3e they give you enough ranks to boost all your skills, put caps so you have to spread your ranks out, and so on.
In 2e, some skill stuff worked on percentages, some things were ability checks, some things were modified attack rolls, and so on. It was a mess. So in 3e every skill operates as an ability check with a bonus on it.
And that's it. 3e's skill system is an afterthought, and all it's really supposed to do is not do a couple of things that really pissed people off about how non weapon proficiecies worked in 2nd edition. That's really as far as the rabbit hole goes. All the stuff about bonus inflation and shit are totally real problems, but they are emergent problems that seem to have caught the designers by surprise. They were too busy fighting the demons of 1989 to notice that they had created a system where a 10th level character could have a +25 to their skill and that people would subsequently design things with DCs of 35 or more which were well beyond impossible for most characters of almost any level.
vagrant wrote: My issue with the DnD skill system has always been 'Why bother training skills when a wizard can duplicate the effects and do it better at the cost of a single spell slot?'
As I said in the other thread, as soon as you accept that class features (like spellcasting) are supposed to make skills look like crap, this issue vanishes.
Why should I accept that? Is that part of this hate-on for mundane characters that some people seem to have?
Then, once you have absorbed the lesson, that your so-called "friends" are nothing but meat sacks flopping around in the fashion of an outgassing corpse, pile all of your dice and pencils and graph-paper in the corner and SET THEM ON FIRE. Weep meaningless tears.
I don't understand. Why should you accept that the issue goes away?
I'm not saying that someone couldn't come up with a kick-ass skill system that's well balanced and scales perfectly from Aunt May level to Galactus level. I'm just saying that there are a bunch of thorny issues that you have to solve first, as you point out yourself. If you're willing to accept that skills only work on modest tasks, all the problems about balance and worrying about breaking the random number generator just disappear.
Well yes, spellcasting is feature better than skills in 3.5. Theoretically, though, wouldn't it be better not to say 'Spells>Skills, everyone play a caster' and instead 'Here's a few ways to make skillful guy a character concept that doesn't feel small in the pants when the other dude is a wizard'?
The problem with the DnD skill system (in 3.5) is while it seems like 'Dude with skills' should be equivalent to 'Dude with spells' because they both have classes and levels and (in very stupid occassions) end in the same party, the end result is that spellcasting is flat-out better.
But that's a quirk of 3.5, and one that can be remedied. (Perhaps not in the core system itself, not that I've tried, but in general design.)
Then, once you have absorbed the lesson, that your so-called "friends" are nothing but meat sacks flopping around in the fashion of an outgassing corpse, pile all of your dice and pencils and graph-paper in the corner and SET THEM ON FIRE. Weep meaningless tears.
vagrant wrote:Well yes, spellcasting is feature better than skills in 3.5. Theoretically, though, wouldn't it be better not to say 'Spells>Skills, everyone play a caster' and instead 'Here's a few ways to make skillful guy a character concept that doesn't feel small in the pants when the other dude is a wizard'?
The way to make a skillful class look cool is through good class features. "You have the same skills that every other class has" is not a good class feature.
vagrant wrote:Well yes, spellcasting is feature better than skills in 3.5. Theoretically, though, wouldn't it be better not to say 'Spells>Skills, everyone play a caster' and instead 'Here's a few ways to make skillful guy a character concept that doesn't feel small in the pants when the other dude is a wizard'?
The way to make a skillful class look cool is through good class features. "You have the same skills that every other class has" is not a good class feature.
I don't understand why your "class feature" can't be a bonus to a skill though. Everyone can poke someone with a sharp stick, I doubt anyone has a problem with the Paladin having a "class feature" where he's better at it. And I equally doubt that anyone has a problem to the Paladin's poking with a sharp stick using the same "attack roll" that every other character uses. Why then, would you balk at precisely the same reasoning when it's that everyone can sneak across a room, and the Assassin has a "class feature" where is is better at it?
FrankTrollman wrote: Everyone can poke someone with a sharp stick, I doubt anyone has a problem with the Paladin having a "class feature" where he's better at it.
Just to clarify -- are you arguing that the 3.5E Paladin is a cool class because "Full BAB" is a good class feature?
FrankTrollman wrote: Why then, would you balk at precisely the same reasoning when it's that everyone can sneak across a room, and the Assassin has a "class feature" where is is better at it?
Why shouldn't an Assassin get a class feature like Greater Invisibility instead?
vagrant wrote:Well yes, spellcasting is feature better than skills in 3.5. Theoretically, though, wouldn't it be better not to say 'Spells>Skills, everyone play a caster' and instead 'Here's a few ways to make skillful guy a character concept that doesn't feel small in the pants when the other dude is a wizard'?
The way to make a skillful class look cool is through good class features. "You have the same skills that every other class has" is not a good class feature.
I don't understand why your "class feature" can't be a bonus to a skill though. Everyone can poke someone with a sharp stick, I doubt anyone has a problem with the Paladin having a "class feature" where he's better at it. And I equally doubt that anyone has a problem to the Paladin's poking with a sharp stick using the same "attack roll" that every other character uses. Why then, would you balk at precisely the same reasoning when it's that everyone can sneak across a room, and the Assassin has a "class feature" where is is better at it?
That seems like an oversimplification of Hogarth's point. The paladin, fighter, ranger, and barbarian all have the "class feature" of being better at poking with a sharp stick; but they look cool in different ways by both how they're better at poking and having abilities the others don't possess.
Come see Sprockets & Serials
How do you confuse a barbarian?
Put a greatsword a maul and a greataxe in a room and ask them to take their pick
hogarth wrote:
The way to make a skillful class look cool is through good class features. "You have the same skills that every other class has" is not a good class feature.
I really don't think skills should ever be the focus of a class. Skills should be minor background stuff that every character gets, but not something you devote a whole class to.
The rogue is one of the worst concepts in D&D, and one that continually gets perpetuated through all the editions, because they're stuck on having this thief who is a mediocre fighter and somehow solves problems with skills instead of force. 3E/4E partially helped this by turning the rogue into a flanking master of the cheap shot, but that's seldom what people really want to play.
What people really want is batman. You fight well, you can sneak around, pick locks and do investigations.
