Very high skill checks

[WiserOdin032402]

Really one of the main issues is cutting fighting-types on skills because they can fight and then trying to pretend like this isn't a game where the primary method of interaction is combat.

[Omegonthesane]

I don't think there is a problem with every 12th level person being better than any 1st level person at EVERYTHING. If you're a 12th level wizard and you can run a marathon better than a 1st level Warrior, that's fine by me.

But we still have the 1st level Wizard setting things on fire with his mind better than the 12th level Warrior, right?

-Username17

Not necessarily, the warrior could have a magic sword that shoots fireballs by 12th level.

Though admittedly your works convinced me long ago that "Warrior" should have Pokévolved to some flavour of Magic Knight by level 12 due to the difficulty of justifying why you should keep up without an explicit "this person is superhuman".

[Username17]

Settings things on fire with his mind is the Wizard's shtick. Stabbing things in the face with the warrior's.

And yet, there's never been an edition where a 12th level Wizard wasn't better at stabbing things in the face than a 1st level Warrior. The 12th level Wizard isn't better at stabbing level appropriate enemies, and indeed in editions where the level treadmill is particularly steep the Wizard may not consider "stabbing things" to be a worthwhile option - but the bottom line is that the 12th level Wizard has a BAB of +6/+1 and gets more hit points from Con Bonus alone than the Fighter has total hit points. When DDMW says he's OK with the 12th level Wizard being better than the 1st level Fighter at literally everything, he means literally everything - and yet there are still things the 1st level Wizard can do better than the 12th level Fighter in that framework.

Just something to think about the next time someone says that it's OK for a character class to have no phlebtonium out to level 20.

high level martials should be good at stealth relative to low-level opponents; high level rogues should be good at stealth relative to high-level opponents

This is a fundamentally bad way to think about stealth. Characters should be hiding from creatures that are more powerful than they are. Children should hide from giants. Halflings should hide from Ring Wraiths. And so on.

If you find that your skill system is not allowing entry level characters to hide from dragons and ogres your system is bad.

-Username17

[Dean]

A stealth system where you could reliably beat massively higher level opposition kind of circumvents the meaning of level. It also means that every combat in the entire campaign probably starts with an ambush. Still, it's a fair point about genre emulation for stealth.

Maybe you could keep the stealth game relatively flat across levels, making your chances of hiding from a guard or the Ancient Naga not much different. That would mean stealth was always something of a coin toss but still a better idea if you know your combat odds are much less than 50%. This would also allow for some monsters to be unusually bad at perception and making those the closet trolls seems sensible. So War Trolls and Giant Scorpions and the like can still annihilate level appropriate opposition in melee but now the alternate strategies of kiting or hiding are highly suggested by the game engine.

This also means every enemy in the game has about a 50/50 chance of being able to sneak up on you, which could be annoying but that seems like something you could fix with just a litttttle Oberoni, in this case ignoring those rules when enemies would use them on players if that would be annoying.

[jt]

It's okay for characters of the same level to have wildly divergent abilities so long as the selection of challenges they can face using those abilities is in the same ballpark. Getting to the kingdom of the merpeople can be done by casting Water Breathing, being so buff that you can just hold your breathe and swim down, or convincing a whale to let you ride in its mouth, and the people who can do these things may be hopeless at copying each other's approaches, but they're all believably contributing to the same party.

Stealth specifically is a mess because it's both a problem-solving ability and a class of challenge. It's perfectly okay for the rogue's jacked stealth stat to let him sneak in and steal the merking's crown, when the alternative is that the bard convinces him to lend it to her, or the wizard trades it for a fancier crown made of illusion-stuff. But when the merking finds out and unleashes his pet kraken to chase after you, it's not okay for "avoiding the kraken" to be something that only the rogue is good at.

Maybe solving things with stealth is actually a separate skill and/or system from hiding from huge monsters? Is it actually two skills: Heist and Hide?

[MGuy]

If you make a fighty dude able to participate reasonably in level appropriate adventures no one gives a shit about anything else. Really that's the whole ball game. The real issue is that 'fight' has a lot less conceptual space and therefore problem solving applicability than 'uses magic'. We've had this discussion 100000000000000 times and that's the take home thing from all of them. Just don't let the thing that your swordy dudes 'do' only be 'swording' or 'fighting'. If your super strong brick house dude can ably participate in the adventure no one (outside of this board maybe) will give a shit if the wizard can beat up people from 10 levels ago.

Everyone nominally should be able to participate in all parts of the game. They may not be able to overcome 'every' challenge just from their ability list. The Bard might not be able to get their self to Atlantis and may need to depend on another person or contraption to get themselves there but as long as they can then do something that helps once they've been bused down there 'that's' what matters.

There are plenty of people who want to play 'just' a fighty character. I think it's the most picked class in 5e and that's probably true in most editions/games. There's a lot of media that just has fighty dudes punching all their problems away after all. Tony Stark has the genius to create whatever he wants and still spends most of his time just doing the blasty blasty and punching Thanos. So you probably should make a game that generates these kind of characters. In catering to these people you have to look at what they actually care about. People who want to play the fighty guy only 'feel' bad if they have nothing to do through some significant portion of the game. Some of them even 'only' care that they can fight good because they are not interested in other parts of the game. Have alternate ways that these people can do shit that matters on their character sheet and people will be fine with it.

------------
As for stealth this isn't a big complicated conundrum. If you're giving random abilities to people different ways to participate in 'avoid the kraken' is trivially easy. The wizard uses the same illusions he did before to avoid it. The bard perhaps charms it into believing that the merfolk it is looking for isn't here or puts it to sleep. This is not the hard part of dealing with stealth. Dealing with big creatures not being as capable of finding small creatures also isn't a big problem 'if' you already have a functional stealth system together. This is the same thing as the crusader kings trait thread Lago put up. If you actually make a stealth system that can work, adjusting the outputs so small people can more easily hide from big people would be similarly easy to generate. You won't be doing it in the d20 system we have because it's fucked but 'if' you have a functioning system that has reasonable outputs then it's a lot easier.

I don't see the reason to fret over size differentials making things harder or easier to find before you've determined how capable people are at finding things in general.

[maglag]

If you make a fighty dude able to participate reasonably in level appropriate adventures no one gives a shit about anything else. Really that's the whole ball game. The real issue is that 'fight' has a lot less conceptual space and therefore problem solving applicability than 'uses magic'. We've had this discussion 100000000000000 times and that's the take home thing from all of them. Just don't let the thing that your swordy dudes 'do' only be 'swording' or 'fighting'.

Youmu Konpaku from Touhou only does swording but she's so good at it that besides combat she:
-Single-handedly tends to a massive country/world-sized garden plus doing all sorts of maid work and still has enough free time to go adventuring.
-Cut through time and space for traveling.
-Can cut sound and light to make others unable to hear or see whitout actually hurting them.
-Can even cut conceptual stuff like slicing somebody's fear to make them braver.

So even "swording" can do a lot of stuff if you push it enough.

[Foxwarrior]

-Can even cut conceptual stuff like slicing somebody's fear to make them braver.

So even "swording" can do a lot of stuff if you push it enough.

Being a wizard is just "saying words and thinking about things" that's been pushed really, really far.

[jt]

If you make a fighty dude able to participate reasonably in level appropriate adventures no one gives a shit about anything else. Really that's the whole ball game. The real issue is that 'fight' has a lot less conceptual space and therefore problem solving applicability than 'uses magic'. We've had this discussion 100000000000000 times and that's the take home thing from all of them. Just don't let the thing that your swordy dudes 'do' only be 'swording' or 'fighting'. If your super strong brick house dude can ably participate in the adventure no one (outside of this board maybe) will give a shit if the wizard can beat up people from 10 levels ago.

I dunno dude, I thought this was all pretty well hashed out too but people seem to want to dive into those debates again instead of answering the initial questions I posed.

Are there core plot-solving/obstacle-bypassing abilities that a high level character should have that aren't covered by something here? More clever uses for high-level skills that seem like they should be included? Ways to shore up the skills that are still weak on the above list?

[merxa]

seems like skills become a second source of magic once a sufficiently high heal check animates corpses, or resurrects the dead, if a high diplomacy check turns into telepathy, detect thoughts, dominate.

which is fine, the world does become very explicitly magical, like everyone above a certain level is a nen user that can manifest supernatural powers.

again, i think tying these thresholds to ranks instead of high DC checks makes more sense for tiering, preventing (some) abuses, and consistency -- gaining a climb speed at 5 ranks in athletics makes more sense to me then saying you can suddenly have a climb speed whenever you can roll a DC 25 check.

In 3x having a climb speed gives you a bonus to climb checks and lets you take 10 in combat, so there might be a few situations where someone needs to make a check even if they have a climb speed.

I'm not sure if spontaneously being able to fly because you have an athletics of 40 is especially flavorful, maybe gaining airwalk is more on target -- and maybe making gaining airwalk at 10 ranks or whatever would be better if it was a option out of a list of options, like PF2 skill feats but the choices matter.

[Kaelik]

The persistent idea that "high skill checks let you do magic level stuff" has always been extremely stupid.

Just give people class features to do high level shit. For fucks sake.

We have trillions of homebrewed classes with high level abilites that are functional in play and no one has ever created a skills give magic system that wasn't at least two of:

1) A buff to Wizards not Fighters.
2) Composed of all the terrible failings of the 3.5 skill system that make it horrible to use (buy a +30 item, fail half the time at things you should do easily or succeed half the time at things that should be impossible, ect.)
3) Too little to fix any problem.



Obviously this is a criticism of D&D specific solutions, you can do a lot with Shadowrun skills, but you don't need to because while mages are really good so is a fleet of fucking drones.

[nockermensch]

If you make a fighty dude able to participate reasonably in level appropriate adventures no one gives a shit about anything else. Really that's the whole ball game. The real issue is that 'fight' has a lot less conceptual space and therefore problem solving applicability than 'uses magic'. We've had this discussion 100000000000000 times and that's the take home thing from all of them. Just don't let the thing that your swordy dudes 'do' only be 'swording' or 'fighting'.

Youmu Konpaku from Touhou only does swording but she's so good at it that besides combat she:
-Single-handedly tends to a massive country/world-sized garden plus doing all sorts of maid work and still has enough free time to go adventuring.
-Cut through time and space for traveling.
-Can cut sound and light to make others unable to hear or see whitout actually hurting them.
-Can even cut conceptual stuff like slicing somebody's fear to make them braver.

So even "swording" can do a lot of stuff if you push it enough.

Youmu "just swords things" because in ZUN writing, every* named character has certain assumed abilities like "flying 24/7" or "seeing magic".

I mean, in Youmu's original game, Yuyuko sends her to the land of living to "collect every bit of Spring" (as in, the season), and Youmu does that. By swording, probably. Then she notices at once that the protagonists have the remaining bits and danmaku ensues.

This shows that for a "sword guy" to contribute on adventures he needs:
[*] to go where the action is
[*] to see what the adventure is about
[*] to affect what needs to be affected.

Zun addresses this by Author Fiat and large amounts of beer. A game designer needs to write these abilities.


* isn't it sad, Rinnosuke, Akyuu and Kosuzu?

[Dean]

As for stealth this isn't a big complicated conundrum. If you're giving random abilities to people different ways to participate in 'avoid the kraken' is trivially easy....

I don't see the reason to fret over size differentials making things harder or easier to find before you've determined how capable people are at finding things in general.

You should read what's being discussed again. The discussion isn't that the small should be able to hide from the large. It's that the weak should be able to hide from the powerful, irrespective of size. The fellowship hides from the Ringwraiths and frightened teens hide from Jason even though they are all the same relative size. It's the statement that stealth should, in some way, work irrespective of (or even inversely to) level which is normally the benchmark we use for competence. Because stealth is most often used by someone much weaker than the person they are sneaking by.

It's a fair statement about the use of stealth in the material we're trying to replicate but a hard problem to resolve. Because if the powerful can be stealthed by frightened peasants and camping teens it probably means everyone can stealth everyone, and now you have to figure out why every combat isn't an ambush and every fight doesn't end with the losers vanishing into mist the moment the tide turns against them to come at you later.

I would actually say, as a guideline to both you and jt, that any time you handwaive any element of creating a successful stealth game as "trivially easy" you are almost certainly wrong. As it is truly wildly deeply difficult.

[Chamomile]

The more powerful skills are, the more you've written yourself into a weird hybrid level/point-buy system that you don't actually want. Has anyone bothered to make the case for why we want skills to be comparable with class features?

[Pedantic]

The more powerful skills are, the more you've written yourself into a weird hybrid level/point-buy system that you don't actually want. Has anyone bothered to make the case for why we want skills to be comparable with class features?

It's the transitive property of mundanity. "Magic" is specific techniques that break the rules, "normal" things are achieved by skill checks, we want normals to compete with mages, so their high level abilities must be coded as skill checks so they can remain "normal."

All of this would be so much easier if D&D had a legacy of more than 1 resource management system. There's been other attempts, but the thematic line gets drawn as "spells that cost resources" and "actions that are normal and thus cost no resources."

[MGuy]

As for stealth this isn't a big complicated conundrum. If you're giving random abilities to people different ways to participate in 'avoid the kraken' is trivially easy....

I don't see the reason to fret over size differentials making things harder or easier to find before you've determined how capable people are at finding things in general.

You should read what's being discussed again. The discussion isn't that the small should be able to hide from the large. It's that the weak should be able to hide from the powerful, irrespective of size. The fellowship hides from the Ringwraiths and frightened teens hide from Jason even though they are all the same relative size. It's the statement that stealth should, in some way, work irrespective of (or even inversely to) level which is normally the benchmark we use for competence. Because stealth is most often used by someone much weaker than the person they are sneaking by.

It's a fair statement about the use of stealth in the material we're trying to replicate but a hard problem to resolve. Because if the powerful can be stealthed by frightened peasants and camping teens it probably means everyone can stealth everyone, and now you have to figure out why every combat isn't an ambush and every fight doesn't end with the losers vanishing into mist the moment the tide turns against them to come at you later.

I would actually say, as a guideline to both you and jt, that any time you handwaive any element of creating a successful stealth game as "trivially easy" you are almost certainly wrong. As it is truly wildly deeply difficult.

I'd advise you to get the point so you might be less wrong here. You don't even have to have more powerful people have scaling perception. You're making an assumption on a system you haven't made. The point I was making is that you need a foundation before you can attend to these things. If you have an argument against that please explain to me how you expect to approach these problems without first determining how capable people are of seeing each other and what general shape your stealth system is going to be. I didn't hand wave the problem. I said without a basic system you can't approach them. Also your assumption that all encounters become ambushes is really weird.

[jt]

The more powerful skills are, the more you've written yourself into a weird hybrid level/point-buy system that you don't actually want.

That is, in fact, what I actually want.

I'm doing an exercise where I design a system that somebody could believe is an edition of D&D. I wanted to do separate accounting of combat-relevant abilities and non-combat abilities so people can't buy things from one context at the expense of another and make characters that have to sit on their thumbs for chunks of the session. 3E already has three(!) power-accounting systems (class levels, feats, skills), so I picked class levels and skills. There are reasons I picked that pair, but any two of them would've worked.

Why I picked that pair:
[*] Since classes traditionally grant skill points, I can have classes with better concepts for non-combat utility grant those abilities as class features at the expense of granting fewer skill points. (This also can be made to work for class + feat.)
[*] Of the three, classes have the most precedence as combat-oriented, and skills have the most precedence as not, while feats have always been a weird mixed bag.
[*] I had a better design for letting casters buy non-combat utility spells with skill points than I had for letting them buy them with feats.

But I'd like to stress that I think any of the level/feat/skill pairs would work to create a combat/non accounting split. And if someone thinks a different one would be good I'd be way more interested in a thread exploring the ramifications about that than a thread arguing about which one to use.

I would actually say, as a guideline to both you and jt, that any time you handwaive any element of creating a successful stealth game as "trivially easy" you are almost certainly wrong. As it is truly wildly deeply difficult.

I never said that? The only post I made directly centered on stealth was talking about how it was weird and difficult and maybe the skill needs to be split up to get traction on the problem.

(But my goal for this project is, "Good if possible, but at least better than 3E," and I do know how to make at least a small improvement on 3W's stealth system. So stealth is trivial )

[maglag]

-Can even cut conceptual stuff like slicing somebody's fear to make them braver.

So even "swording" can do a lot of stuff if you push it enough.

Being a wizard is just "saying words and thinking about things" that's been pushed really, really far.

Saying words, thinking about things, special implements and literal handwaving.

If you make a fighty dude able to participate reasonably in level appropriate adventures no one gives a shit about anything else. Really that's the whole ball game. The real issue is that 'fight' has a lot less conceptual space and therefore problem solving applicability than 'uses magic'. We've had this discussion 100000000000000 times and that's the take home thing from all of them. Just don't let the thing that your swordy dudes 'do' only be 'swording' or 'fighting'.

Youmu Konpaku from Touhou only does swording but she's so good at it that besides combat she:
-Single-handedly tends to a massive country/world-sized garden plus doing all sorts of maid work and still has enough free time to go adventuring.
-Cut through time and space for traveling.
-Can cut sound and light to make others unable to hear or see whitout actually hurting them.
-Can even cut conceptual stuff like slicing somebody's fear to make them braver.

So even "swording" can do a lot of stuff if you push it enough.

Youmu "just swords things" because in ZUN writing, every* named character has certain assumed abilities like "flying 24/7" or "seeing magic".

I mean, in Youmu's original game, Yuyuko sends her to the land of living to "collect every bit of Spring" (as in, the season), and Youmu does that. By swording, probably. Then she notices at once that the protagonists have the remaining bits and danmaku ensues.

This shows that for a "sword guy" to contribute on adventures he needs:
[*] to go where the action is
[*] to see what the adventure is about
[*] to affect what needs to be affected.

Zun addresses this by Author Fiat and large amounts of beer. A game designer needs to write these abilities.

You don't need to see magic to find out where the adventure is. Like in Perfect Cherry Blossom the Winter is not ending and so the first half of the game is literally going around at random beating up random people until they run into Alice that after being beaten up coughs "just follow the fucking trail of petals, bitch".

Essence of Scarlet Devil there's a giant red mist bloating out the sun spreading from a central point, not exactly easy to miss either.

Imperishable Night big red unnatural moon keeps repeating, again go fly at random beating random people until somebody coughs up the right direction.

Phantasmagoria of Flower is flowers blooming everywhere, and only the boss characters has any hint what is actually behind it until the couple final stages when plot exposition is dumped on them. Most characters just go around beating random people until running into the true final boss. Most characters don't even realize they're in an adventure until they run into the true final boss!

Mountain of Faith a new goddess arrived and is literally screaming "I'M THE ONE IN CHARGE NOW, ANYBODY WHO HAS A PROBLEM WITH THAT COME UP HERE AND FIGHT ME!" and the game is going up said mountain beating up her minions until reaching the top and beating the big bad.

Subterranean Animism evil spirits are pouring out from below and then the heroines get the plot dumped on them anyway by previous bosses. Marisa even goes meta and treats it like a video game although a RPG instead of bullet hell.

Undefined Fantastic Object is "GIANT HUGE SHIP FLYING AROUND GO CHECK IT OUT!"

And so on. There's always some giant massive clue that it's impossible to miss to draw the protagonists to adventure. It's baked in the system, the plot is a big shining beacon drawing the players from far and wide.

If the final destination isn't obvious from stage 1, then just go around at random beating up random people trying to find a clue or hope they run into the true culprit by sheer luck. Which is a perfectly viable strategy in a tabletop RPG, if you beat up everybody, eventually you'll beat up the true culprit!

Then every problem can be conveniently solved with enough swording/damnaku/grazing since it's always some sort of gothic girl behind it.

Also for all we know Youmu can fly by swording just like she can do sword danmaku.

[owlassociate]

The fellowship hides from the Ringwraiths and frightened teens hide from Jason even though they are all the same relative size. It's the statement that stealth should, in some way, work irrespective of (or even inversely to) level which is normally the benchmark we use for competence. Because stealth is most often used by someone much weaker than the person they are sneaking by.

It's a fair statement about the use of stealth in the material we're trying to replicate but a hard problem to resolve. Because if the powerful can be stealthed by frightened peasants and camping teens it probably means everyone can stealth everyone, and now you have to figure out why every combat isn't an ambush and every fight doesn't end with the losers vanishing into mist the moment the tide turns against them to come at you later.

I think there's a few things one could do to model something like this. For one, decouple skills from level. Skills have ranks (e.g. untrained, trained, expert) and each rank provides flat bonus. This allows you to keep the numbers pretty tight and avoids the problem of 20th level farmers and sages.

Next, you don't have to make everyone extremely stealthy. The example with the hobbits (who are naturally very stealthy) hiding from ringwraiths (who are nearly blind) demonstrates this perfectly. As long as things like stealth and perception do not directly scale with level, situations like that will just work out. Same goes for something like children (which are small and thus have an easier time hiding) hiding from ogres (which are very stupid and thus bad at finding things).

Then, there's the example of teens hiding from Jason, where I would ask if the teens really need to able to hide from Jason all that reliably. The movie kind of falls apart when all the characters just disappear and Jason never finds his victims. This is the kind of the base situation where no one in the hide and seek minigame has an advantage over the other (as long as Jason is the seeker).

[Username17]

The more powerful skills are, the more you've written yourself into a weird hybrid level/point-buy system that you don't actually want. Has anyone bothered to make the case for why we want skills to be comparable with class features?

I'm going to take a step back and talk about Chekhov's Gun. In good storytelling, if you show a gun in Act I, that gun should be fired before the end of the final act. The corollary is that if you fire a gun in the final act, you should probably show it in an earlier act. That's called foreshadowing and it's good storytelling.

In a cooperative storytelling game, characters often get new abilities. That is, they "fire guns" and it's better storytelling if those abilities are in some way established before hand. That's ultimately the reason we have "levels" and "character classes" in the first place - so that the abilities of tomorrow will in some way be contiguous with and set up by the abilities of today and yesterday.

So it is not an unreasonable suggestion that whatever mundane abilities appear on a character sheet at the beginning of a character's life have some upgrade path wherein they would set up and establish the abilities that would be relevant in later, higher powered stories.

Now there are several problems with the things jt is talking about. The first is that, obviously, d20 skills don't work that way and in fact don't really work at all at higher levels. When you ask what Athletics should do at higher levels the answer must always start with "Be in a skill system that isn't such a fucking embarrassment at higher levels." But beyond that, we have the more basic problem that not everyone who is good at swimming at level 1 grows up to be Aquaman. Sometimes a character just happens to need to swim somewhere and it's not a setup for a character to eventually learn hydromancy and talk to fish.

These are general problems with the concept. In specific of course you have the issue that jt's list is chocolate covered bullshit. Like, Sense Motive having scaling DCs based on sense motiving level appropriate threats is obviously not a meaningful upgrade is the sense that teleporting or raising the dead is. I mean, fucking obviously, right? It's just doing the same thing you were doing in the first episode at level 1, not doing anything remotely cool.

Anyway, the natural end to this line of thought is the following:

• Define the challenge space at different levels of play.
• Have a better skill system than any edition of D&D has had.
• Define a set series of phlebtonium abilities that allow stronger characters to meaningfully interact with the challenges that are defined as being beyond starting characters.
• Tie those abilities to skills such that characters who have different starting skills would have the ability to activate different phlebtonium abilities and thus each character has a different set of power growth options.

But you'll note that this is basically World of Darkness if hypothetically the mechanics were any good, and not an edition of Dungeons & Dragons at all. If you're playing D&D you already have a system for doling out character advancement and growth options and it's called "going up in level and getting more class features."

-Username17

[Chamomile]

The more powerful skills are, the more you've written yourself into a weird hybrid level/point-buy system that you don't actually want.

That is, in fact, what I actually want.

Then you need to re-evaluate the life choices that brought you to this moment.

The advantages of a skill-based system is that you can create a completely customized character who can "multiclass" into other archetypes without any complicated rules support. Your heavily armored knight can pick up some fire magic on the side as easily as dropping some points into the fire magic skill. The disadvantage is that it's possible to create a character who is hyper-specialized to the point of being unable to participate in certain mini-games, or who's so spread out between so many different schticks that they suck at everything (three points in swords, fire magic, and stealth instead of six points in one and three in another means they're mediocre at everything). Also, if the power scaling is such that you have 40 points in the Armor and Swords skills, splashing 5 points into the Fire Magic skill is probably a waste.

The advantages of a class-and-level system is that, provided the classes are built properly, it is impossible to show up to an encounter without the tools needed to solve it. Everything you need is built directly into your class, and you can communicate what you are and what you can do very clearly just by saying that you're a 7th-level Cock Hurler. The disadvantage is that there's minimal customization. There are probably some bells and whistles you can add via feats or skills and it probably means something that you are an elf or that you come from Kentucky, but ultimately your class is always going to account for well over half of your character's total abilities. Multiclassing is usually a train wreck, so if you want your armored cavalier with some fire magic on the side, you probably need to write a Fire Knight class.

These two are inherently opposed to one another. The more control players have over their build, the more they'll be able to create bad characters who can't compete in all the areas they need to. A totally open skill system is the system - no starting caps on skill level or anything - is the system most prone to the creation of unusable characters. A completely controlled class system - no skills or feats or anything - is the system most resilient to the creation of unusable characters.

Your goal is to create a system where players are less prone to making characters with poorly rounded characters, specifically, characters who don't have non-combat abilities. Your solution is to lean more heavily on a system that makes it easier to make poorly rounded characters. To the extent that this can work, it is only because 3.5's class balance is such a trainwreck to begin with that even bad ideas can be an improvement if you work hard enough at them. It would be both easier and more effective to just build those non-combat abilities directly into people's classes, though. If you're really committed to doing this with skills, you need to be making a far more drastic overhaul of the skill system such that, for example, people can't spend all the skill points they need to be good at social encounters on Gymnastics and Athletics, and they're forced to try and figure out how those could possibly be relevant - something they could do with any skill system that doesn't specifically set out to thwart them.

[angelfromanotherpin]

I've run games where skills scaled to give the equivalent of spell effects from four levels ago, and it created some pleasingly mythic atmosphere, but very few of the players could really engage with the material. I basically had to remember it for most of them most of the time.

My current homebrew works from the premise that any degree of mastery in a skill is 1st-level appropriate. Actual level-appropriate skill effects are class features and/or feats. It has worked much better.

[nockermensch]

Also for all we know Youmu can fly by swording just like she can do sword danmaku.

While we can stay here discussing Touhou all week, it's not the thread theme and the quoted bit is the key: By the end of the day Youmu behaves as a magic user, just as any character in the series*. But like a youxia character, she has martial arts as the components/visual effects for spells.

* LOL, Renko

[maglag]

Also for all we know Youmu can fly by swording just like she can do sword danmaku.

While we can stay here discussing Touhou all week, it's not the thread theme and the quoted bit is the key: By the end of the day Youmu behaves as a magic user, just as any character in the series*. But like a youxia character, she has martial arts as the components/visual effects for spells.

Then we're back at the base problem that only magic users are allowed to do anything fantastic, which screws everybody else from the start.

Hesitation-Cutting Sword "Slash of Departure from Hesitation"
• User: Youmu Konpaku
• Notes: Confirmed in the Netherworld, just a drive-by slasher
• Reference Level: I wanna reference this more than I expected to
A Spell Card where she just smacks you with her sword.
"This isn't danmaku, is it?" Touche.

And Touhou also has "spells" like:

Oni Sign "Complete Massacre on Mt.Ooe"
• User: Suika Ibuki
• Notes: Usually seen at parties, a real spicy drunk
• Reference Level: Won't
She catches and throws you, usually when she's drunk at parties.
Again: not danmaku. Hurts a ton, too.

Which by a local magician's own words aren't really spells in anything besides name. Youmu's "magic" is her swording people really hard. Suika's "magic" consists of being just really freaking strong then grappling the heck out of her opponents.

And before you go "But Suika can change size and turn into mist", allow me to point out to her friend Yuugi that can't do neither while her "danmaku" consists of rapid-throwing rocks at you and screaming/stomping with such power as to create shockwaves. Is Yuugi a spellcaster because of that? Is the Hulk also a spellcaster when he changes size, makes shockwaves and grapples/flings stuff?

[Iduno]

d20+mod is really good because its linear and simple. It works well at low levels and mid levels with limited modifiers. It falls to shit at higher levels. It's also swingy and requires a secondary roll to really provide a good idea of if a success was marginal/average/critical. When you try and push that onto a single die roll it just enhances the negative ends of the curve.

Level-based games have an issue with running off the RNG, or not getting level-appropriate things (D&D 5th). It's a known issue, and people use levels when it's an appropriate trade off for the players not needing to be smart/sober/motivated enough to improve their characters on their own.

I think there's a few things one could do to model something like this. For one, decouple skills from level.

That could also work.

Are there core plot-solving/obstacle-bypassing abilities that a high level character should have that aren't covered by something here? More clever uses for high-level skills that seem like they should be included? Ways to shore up the skills that are still weak on the above list?

I think it's been said that "do most of the stuff a wizard does, with the fluff that they're doing X skill so hard it's the same as magic" is the answer to that one. Stealthing so hard you're invisible jumping so hard you can fly wouldn't be unreasonable if someone is focused on that. The game is partly based on myths, after all.