Skills should not be superpowers in D&D

[Maxus]

Your stats are are raw and untrained capacity, nothing more.

Why?

Better question: Why can they ONLY be that? I see no huge moral imperative either way. I dig the current system, and I could dig K's version where some physical ability and knowledge is just assumed.

[K]

DnD wants to say that you can be a moron who knows reams of facts and great recall on one thing but terrible recall and ignorance on other things, a near autistic who is incredibly charming, a weakling who can scale mountains, and a clumsy oaf who can perform the most difficult acts of balance. That's what skill means in DnD.

Actually it does'nt really mean this at all. The only real thing we have to base what a stat "means" on is int where we know that 3 is animal intellegence. Even this is crap because animals show very different levels of intellegence.

Actually, it does mean this.

A person with a Str of 1 can lift a max of ten pounds over his head by the RAW carrying capacity rules. That same person can have 24 ranks in climb, a +3 for Skill Focus, and a +2 for Athletics, and -5 for Str, for a final raw mod of 24. With a climbing kit for +2, he can never fail a climbing roll so badly that he falls since the highest DC is 30 and he has a +26. At worst, he makes no progress for a round.

The game is telling you that a man who is medium encumbered by clothing (5 lbs, by the RAW) can be a master climber who can never fail to climb nearly impossible surfaces like slippery overhangs with no footholes.

We also know from the MM that creatures with Int less than 3 may not understand any language and also are not playable as characters, but creatures bigger than 3 know at least one language. That person might not have the ability to understand "language," but might have sage-level knowledge of architecture or heraldry or arcana and be capable of recalling vast amounts of knowledge on the subject.

And that's stupid.

[RobbyPants]

The game is telling you that a man who is medium encumbered by clothing (5 lbs, by the RAW)

I'm getting sidetracked here, but I thought the one set you wore didn't count against encumbrance. You only tracked the weight if you happened to be carrying extra clothes with you.

Still, your point stands. The guy wouldn't be able to carry the ropes, clips, pitons, and tools a climber would normally have without being encumbered.

[quanta]

Do you really think that someone in really good physical condition couldn't climb half as high as Chris Sharma does? Really? After a week or two of practice?

There are seriously boulder and cliff faces that may take even someone in good condition years to get half as high as Chris Sharma does. You're talking out your fucking ass and have no clue how difficult what he's doing is. I'm talking about a guy who climbs sheer wave-pounded cliff faces where if you fall, you fall into a giant school of stinging jellyfish.

In a week or two of practice, you'd be lucky to finish the first few moves on some of the climbs he does. Now granted, this is a particular subdiscipline of climbing where you may be able to climb the stuff he does as long as you're willing to use a lot of gear to aid in climbing (use gear to create holds and such as opposed to strictly as a safety measure). But at that point, you're at least an order of magnitude slower than him.

[Just another user]

A problem with D&D 3.x skills is that maximizing is too easy and come with practically no opportunity cost. Unlike many other games there is almost never a reason to not raise a skill to the higher rank possible, maximizing is essentially always the right thing to do, (with some rare exceptions like getting skills synergies, ranks in trained only skills and feast/PrC prerequisites.) and if in a game a choice is always the right one that game has a design problem IMHO

4e just took this and turned it from a 'bug' into a feature.

[Josh_Kablack]

Thank you J.a.U. for exemplifying this argument for me.

Here I thought I was just missing something, but no, really it's all actually incomprehensible. and you are outright gibbering at each other behind a mask of grammatical error, triple-negatives with exclusion clauses and buzzwords covering preconceptions


This one really ain't hard folks:

1. Admit that the last couple editions of D&D treat skills really inconsistently

2. Pick what you want skills to be in your own game / houserules

3. Then write rules that let skills be whatever you picked.

[Lago PARANOIA]

I'd be very interested as to how you could back up this claim, Lago. Most of the math in my research indicates otherwise.

Because +2 circumstance bonuses on a d20 RNG models results very poorly if the difference is significant. Unless you stack up so many of them to make the expression confusing. We know that Bevel Lemesk (guy who engineered the first Death Star) has no chance of winning a military engagement against Grand Admiral Thrawn even if the latter is really sick from the flu and is distracted by the death of his wife and kids. But this has a very real chance of happening unless you have a decent mechanic that prevents amateurs from sneaking up on experts. You know, like a skill system.

You can get pretty good granularity with only Untrained, Trained, and Mastered, plus an attribute. If the progression is 0/4/8 and you use D&D attributes, you have all the integers in the interval [-4, 12].

I don't know about you, but I don't want the difference between skills to have the same RNG difference across the board. People do get offended if a peasant has a chance of beating Conan in an arm-wrestling contest or Alexander the Great in military strategy so having such a large gap between levels of training is a big deal. But a lot of people (including me) think that a peasant should have a chance of convincing a king contrariwise to what his cunning succubus vizier is saying or painting some abstract art that even Vince Van Gogh is impressed by.

Reporting may also involve more than just Profession (reporter) checks, including Sense Motive, Gather Info, Sneak, and so forth. This gives even more granularity.

There's something to be said for resolving a situation in exactly one check. Yes, from the initial popularity of skill challenges there is a demand for being able to go in depth but I think it'd get old fast to have to do it more than occasionally.

It's totally okay for Legendary Blacksmith to be a prestige class (even if PCs aren't excited by it). In fact, I think there are at least 3 Legendary Blacksmith-style prestige classes in 3.5E!

I would argue that there isn't a need for "one +1 bonus short of Legendary Blacksmith". Who needs that guy?

Your idea is a tremendous waste of space. We do need to have a way of representing the Legendary Blacksmith but he does not need an entire page dedicated to himself. A blurb in the skill system is enough. Secondly, the whole one +1 bonus short of Legendary Blacksmith is a strawman. The difference in skill between 'Reknowned' and 'Legendary' does not always have to be +1 for every skill. I don't think that anyone has a problem with saying that the difference in skill between Maxwell or LaPlace and the top researcher at Unremarkable U. is higher than the difference between Manning and Random Quarterback. Or alternatively you could have diminishing returns from decreasing some skills and some skills spiral out of control faster than others (it's less expensive to get a 8 die pool for Profession: Lawyer than it is for Stealth).

[echoVanguard]

Because +2 circumstance bonuses on a d20 RNG models results very poorly if the difference is significant. Unless you stack up so many of them to make the expression confusing. We know that Bevel Lemesk (guy who engineered the first Death Star) has no chance of winning a military engagement against Grand Admiral Thrawn even if the latter is really sick from the flu and is distracted by the death of his wife and kids. But this has a very real chance of happening unless you have a decent mechanic that prevents amateurs from sneaking up on experts.

This argument seems a little contrived, and is very much a case of advising someone to throw the baby out with the bathwater. Analyzing both your argument and the math involved, the actual complaint you're making isn't that +2 circumstance bonuses "just don't work", but rather that they shouldn't be allowed to stack, which is much simpler and has a fairly good case behind it.

The rest of the stuff Lago said

I agree with most of the rest of your post.

echo

[Lago PARANOIA]

Analyzing both your argument and the math involved, the actual complaint you're making isn't that +2 circumstance bonuses "just don't work"

The +2 circumstance modifier doesn't work for several competing reasons and fixing one requires exacerbating the other.

Problem: The +2 circumstance modifier doesn't do enough to lock out certain kinds of people from skill checks.
Fix: Add more circumstances until it does.
New Problem: This creates too many variables to track without futzing things up.

You can even do it in the opposite direction if you'd like.

Now if the range of results is sufficiently small then a +2 circumstance bonus could probably (pathetically) model the difference in abilities. If you're giving a bunch of cloned kindergarteners an oral pop quiz on the Three Bears then it'd be meaningful enough on its own, but it's not sufficient even to model a 'realistic' skill system. So out of the window it goes.

[hogarth]

Your idea is a tremendous waste of space. We do need to have a way of representing the Legendary Blacksmith but he does not need an entire page dedicated to himself. A blurb in the skill system is enough.

I'm impressed that you can take an idea that I haven't described in any detail and compare it to an idea that you haven't described in any detail and claim that my idea is longer than yours.

Lago, the idea that skills going from Renowned (non-superpower) to Legendary (superpower) is the same distance as going from Competent (non-superpower) to Renowned (non-superpower) sounds terrible in the context of D&D, where skill points are cheap and fungible (i.e. one point = exactly one step higher for all skills at all levels). If you're not talking in the context of the D&D skill system, then it might be an interesting idea but you should probably start another thread.

[Lago PARANOIA]

I'm impressed that you can take an idea that I haven't described in any detail and compare it to an idea that you haven't described in any detail and claim that my idea is longer than yours. Smile

You wanted a prestige class for a single skill.

Lago, the idea that skills going from Renowned (non-superpower) to Legendary (superpower) is the same distance as going from Competent (non-superpower) to Renowned (non-superpower) sounds terrible in the context of D&D,

You mean in the context of 3rd Edition D&D, because the other D&Ds don't use that exchange rate. Since you didn't specify an edition (and in fact we're talking about an is-ought system that doesn't exist) why does a fix have to conform to this standard?

More importantly, though, I don't give a shit whether D&D can't model this range of skills. That means D&D has a shittily designed skill system, not that we should abandon the paradigm. Which is why the system I proposed doesn't look anything like the systems 3E nor the other editions of D&D used. Because D&D's skill system is shittily designed.

Before you start anywhere the first question you should ask is 'can my skill system model Legendary, Reknowned, Competent, Mediocre, and Untrained users of this skill check?' because those are the five basic levels of competence in source material. 4E couldn't even meet that pathetic standard and should've been thrown out immediately. 3E D&D can just barely meet it if you jiggle the class and level system, look at source book B, E, and K, sprinkle some magical items into the mix, and voila! So it's better but still needs to be reworked from scratch.

Now with that said, if you want to have a rank 'beyond' Legendary that can only be able to be achieved with superpowers than go for it. Maybe Legendary was a poor choice of words to use. But the basic idea is that even though a skill system will normally only be used for mundane human achievement you still need a level of skill between Einstein and a top biomedical engineer that regularly gets his articles published in Nature.

[hogarth]

You mean in the context of 3rd Edition D&D, because the other D&Ds don't use that exchange rate.

Well, 2E and 3E have a "1 point = +1 system". Pre-2E doesn't really have "skills" in the sense I'm talking about. You're right that I was excluding 4E.

But the basic idea is that even though a skill system will normally only be used for mundane human achievement you still need a level of skill between Einstein and a top biomedical engineer that regularly gets his articles published in Nature.

I agree 100% with this. But my point all along has been that it's okay to say that Einstein has the equivalent of 5 (3E) skill ranks and a high Int, and that's as good as you get within the skill system; there shouldn't be a point at which you invest one more skill rank and you miraculously become Reed Richards, pulling force field belts and laser pistols and ultimate nullifiers out of your ass.

[Lago PARANOIA]

there shouldn't be a point at which you invest one more skill rank and you miraculously become Reed Richards, pulling force field belts and laser pistols and ultimate nullifiers out of your ass.

What system should be used instead and why is it an improvement to skills?

[echoVanguard]

Problem: The +2 circumstance modifier doesn't do enough to lock out certain kinds of people from skill checks.
Fix: Add more circumstances until it does.
New Problem: This creates too many variables to track without futzing things up.

You can even do it in the opposite direction if you'd like.

Now if the range of results is sufficiently small then a +2 circumstance bonus could probably (pathetically) model the difference in abilities. If you're giving a bunch of cloned kindergarteners an oral pop quiz on the Three Bears then it'd be meaningful enough on its own, but it's not sufficient even to model a 'realistic' skill system. So out of the window it goes.

You seem to be operating under a fundamental misunderstanding as to the nature of a circumstance bonus. It is not intended to model the difference in abilities at all - it is intended to model the effect that circumstances have upon a result, which is key in determining how these sorts of rolls are intended to work. To use your previous example, Bevel Lemesk and Grand Admiral Thrawn have wildly divergent levels in the skill we're using to determine the outcome of this military engagement (we'll call it "Tactics"). For example, we can say that Bevel has 4 ranks in Tactics, while Thrawn has 9.

If Thrawn's command base has very poor instruments, meaning that he is getting little to no information about the state of the battle and is experiencing problems directing his units, that could be modeled by him taking a -2 circumstance penalty to his Tactics roll. Meanwhile, Bevel has designed and deployed a new hyperspace spy probe, which lets him see and direct his units in real-time and gives him a +2 circumstance bonus to his roll. In this instance, Thrawn is not only outclassed at a technology level, but also suffering from equipment-based penalties due to his own situation, which means he might very well be unable to use his tactical knowledge to make up the difference in the realities of the situation. It's obvious that Thrawn is the better tactician by an order of magnitude, regardless of anything else, but the mitigating circumstances might actually make that nigh-irrelevant to the actual outcome of the event.

echo

[Lago PARANOIA]

K and I were arguing whether circumstance bonuses could replace skill ranks, not enrich the outcome of results.

[K]

K and I were arguing whether circumstance bonuses could replace skill ranks, not enrich the outcome of results.

Actually, I was arguing that stats could replace skill ranks and simple trained/untrained could generate modifiers to those rolls based on circumstance.

[echoVanguard]

K and I were arguing whether circumstance bonuses could replace skill ranks, not enrich the outcome of results.

Actually, I was arguing that stats could replace skill ranks and simple trained/untrained could generate modifiers to those rolls based on circumstance.

I don't think that is a good idea. Circumstance modifiers, as the name implies, are circumstantial, meaning they apply only to that particular instance of the action attempt. Trained/Untrained modifiers can work but model actual skill training poorly unless you have additional levels of granularity beyond the binary. There's some argument for stats being broader than mere raw aptitude, but then you run into problems with blending. Short version: you almost certainly want stats + skill, and you want skill ranks to actually mean something and have significant levels of both granularity and comparative impact.

echo

[hogarth]

there shouldn't be a point at which you invest one more skill rank and you miraculously become Reed Richards, pulling force field belts and laser pistols and ultimate nullifiers out of your ass.

What system should be used instead and why is it an improvement to skills?

See post #1.

[Just another user]

A system that handle it well is -IMHO- JAGS.

I don't remember the finer details because it is some time from the last tie I've read it, but it can summed up like that, every skill have two values, a numeric rank, that work as usual, and a grade of mastery divided in four grades (something like apprentice, trained, expert and master - not the exact names-) that are, essentially, advantages, in the GURPS sense, every grade come with their bonus, and 'tricks' and things that only their grade can do.

rank and grade are not strictly related so you could have a 'trained' with a skill of 18 and a 'master' with a skill of 14, the apprentice is potentially better (maybe he is specially gifted) but the master can still do things that the apprentice can't. an interesting detail is that higher grade of mastery don't give bonus to a a skill but they let you ignore penalties, so a master would still be better in difficult situations but it would not automatically succeed every standard check.

So to become Reed Richards and pulling out gadgets you would need to buy the appropriate grade of mastery rather than buy that one skill point which probably would need a stronger in-story justification to do.

[Swordslinger]

Before you start anywhere the first question you should ask is 'can my skill system model Legendary, Reknowned, Competent, Mediocre, and Untrained users of this skill check?' because those are the five basic levels of competence in source material. 4E couldn't even meet that pathetic standard and should've been thrown out immediately. 3E D&D can just barely meet it if you jiggle the class and level system, look at source book B, E, and K, sprinkle some magical items into the mix, and voila! So it's better but still needs to be reworked from scratch.

The D&D skill system (any edition) is generally just horrible.

Dump the idea of skill ranks altogether and go with individual abilities. Nobody cares about +2 ranks in nature, but if you get the ability to follow tracks or feed yourself in the wild, that matters. And I like the idea of abilities that just work, as opposed to things you're constantly making skill checks for. I like the idea that a skilled climber can traverse basic obstacles without even making a check.

Investing in a skill should produce a level of certainty about the outcome. Right now, we have too many skill checks.

[Just another user]

There is "take 10" for that.

[tussock]

Take-10 and take-20 are those things where they admit their skill checks don't usually produce desirable outcomes and so you shouldn't use them. Then you're not allowed to take-10 in combat until you are, so you have to use all those undesirable outcomes anyway (except for the "non-combat" skills, which don't use the skill check system at all).