What am I forgetting about 4e D&D?

[kieza]

There is kind of a limited degree of this among the original 4 sources. Martial, Divine, Arcane, and Primal trend to feel more Striker, Leader, Controller, and Defender respectively
Martial: Ranger and Rogue are two of the highest output strikers, Fighter is the most aggressive defender, and Warlord can easily be the most aggressive leader
Divine: Paladin has Lay On Hands for healing and a lot of healing powers, Invoker honestly can almost be played as a full leader if it had heals, and Unity Avenger can do a solid amount of party support
Arcane: Warlock has a lot of conditions it throws out, while sorcerer does a lot of effective AoE (technically a part of controller), while Bard and Swordmage also have a lot of debuff/enemy interference options
Primal: Barbarian and Druid are both very bulky for striker/controller, and while Shaman itself isn't very defendery, it does have the spirit companion that can be a very efficient source of damage ablation.
Psionic has the power point mechanic

Shadow and elemental never really got enough material to establish an identity.

There are also a few other patterns, but they are less consistent. Most primal classes, for example, have some manner of end of encounter buff as a major component for their dailies (at least some of them). Barbarians have Rages, wardens have Guardian Forms, Druids have...no clean term, but they have wild shape enhancers. Martial classes have the combat style feats, while divine classes (other than runepriest, which was a latecomer) have domain feats and channel divinity.

I agree, but I think that this "power source identity" is hampered by the fact that each class has a unique selection of powers. So yes, they have broad similarities, but the distinct power selection means that they don't have strong ties to each other. Or to put it another way, the writers may have declared that classes are related by their power source, but they don't behave similar enough to each other to make that really convincing.

I instead designed a pool of powers for each power source, which are available to each class within that source.

Pros:

• More commonalities between classes of a power source.
• Didn't have to design a "wizard fireball" and a different "sorcerer fireball." Which also means I didn't have to design as many unique powers.

Mixed Results:

• Classes are less strongly tied to their intended role. Fighters' class features nudge them towards being tanks and Wizards' nudge them towards battlefield-control spells, but they don't have to lean into those playstyles. I like this, but the downside is that there are probably more trap options than in 4e. (A Fighter is allowed to take only ranged-weapon powers, which nominally work with their class features, but would likely result in getting swarmed and then bogged down in melee.)
• It leaves more room for system mastery. Following on from the previous option, there are tons of interesting, atypical combinations that can be made. It's even possible to have a ranged-weapon fighter and have them be fairly effective as a tank. (Suppressing Fire is a power...) And I like this! I like the ability to make a wider variety of character archetypes within a single class, and I like that understanding the system lets you build weirder and marginally more effective characters. But it also means that, as a first-time player, you might see an interesting concept and have no idea how to make it work. Which I tried to combat by writing longer and more detailed "how to build a character of this class" sections, but that's of limited help with really out-there combinations.

Cons:

• I do have to spend longer thinking about whether a power could be broken. To give one example, I have an arcane power that's essentially an improved Total Defense action: same action, same effect, plus it grants a moderate amount of temp HP that lasts until the next turn. (Flavor: the caster raises a ward that absorbs a small amount of harm before shattering.) Perfectly fine, on a Sorcerer, Warlock or Wizard: they use one of their power slots to get that, and now they have a spell that's slightly better than Total Defense (and if they use it every round, they can hardly do anything else). But on an Artificer...who can hand out gadgets that let allies copy their spells as a minor action...I'm concerned about the balance of letting them make a gadget containing that spell and hand it to the party tank, who could then attack normally AND get the benefit of Total Defense AND get temp HP, every single turn.

[RandomPeasant]

To an extent, I think this can be well-addressed by making off-skills have higher DCs, such that using an off-skill is sometimes-to-often worse than using an untrained skill, all other things being equal.

Not really. The basic problem with 4e's Skill Challenges is that because the challenge ends after a fixed number of failures, any action anyone takes that is not the action with the highest chance of success is worse than not taking an action. Changing the DCs can allow you to control which of the PCs is the one who rolls over and over, and you can add various epicycles that stop people from playing optimally like "you must roll every round" or "you can only roll each skill once", but it's the basic framework that is the problem and there's no change you can make to that framework that is as effective as dropping it and switching to "the Skill Challenge ends after X rounds, regardless of how many failures were rolled in those rounds".

I think the lesson here is that you should be very careful about actions having a potential effect that makes the party worse off than if the action had never been taken. I don't think you should never do it, because there are obvious flavor and mechanical reasons you might want to do that, but having it as the default should be a non-starter. I can imagine a situation where "each action has a chance of ending the challenge" makes sense, but that strikes me as a) fairly specific and b) something that requires careful design to get right. "Protect the Rogue while she carefully disarms the traps" is an interesting encounter, but it needs to be designed in a way that gives the rest of the party actions that protect the Rogue and are worth taking.

Some Skill Challenges contain social situations, social situations are not Skill Challenges.

This is just bikeshedding. The premise of Skill Challenges, as they relate to social situations, is that some social situations that would be modeled as an unstructured series of skill checks (and possibly other actions) in 3e will instead be modeled as a Skill Challenge. And the reality of Skill Challenges is that in those situations the Paladin rolling Sense Motive with a low chance of success goes from "probably neutral" to "probably negative".

I agree, but I think that this "power source identity" is hampered by the fact that each class has a unique selection of powers. So yes, they have broad similarities, but the distinct power selection means that they don't have strong ties to each other.

I just gets you back to the same problem. Either classes are different (and having unique power lists makes sense) or classes are similar (and it doesn't). I think trying to do variations on AEDU is limiting yourself for too little gain. Just make Martial like Tome of Battle, Arcane like Vancian Spellcasting, Psionic like 3e Psionics, and Divine and Primal like ... whatever you think Divine and Primal should be.

But on an Artificer...who can hand out gadgets that let allies copy their spells as a minor action...I'm concerned about the balance of letting them make a gadget containing that spell and hand it to the party tank, who could then attack normally AND get the benefit of Total Defense AND get temp HP, every single turn.

The problem here is that "you can give powers to your allies" is not the same system as "you can't do that". You're hitting the other side of the problem. If classes have different power systems, they need to have different power lists.

[kieza]

I just gets you back to the same problem. Either classes are different (and having unique power lists makes sense) or classes are similar (and it doesn't). I think trying to do variations on AEDU is limiting yourself for too little gain. Just make Martial like Tome of Battle, Arcane like Vancian Spellcasting, Psionic like 3e Psionics, and Divine and Primal like ... whatever you think Divine and Primal should be.



The problem here is that "you can give powers to your allies" is not the same system as "you can't do that". You're hitting the other side of the problem. If classes have different power systems, they need to have different power lists.

I don't think I agree with that.
A) There can be multiple axes along which classes are similar, or different. My classes are similar to others within the same power source in what powers they can learn, but they are different in how they interact with those powers, and they are especially different from classes of a different power source. The class features that do this align roughly with the four roles of 4e: Artificers make their spells benefit allies more (by letting allies actually use them). Sorcerers make their powers better at defending allies. Warlocks make their powers do more damage. Wizards make their spells' lingering effects linger longer.
B) I also think that there can be a spectrum of class differences, ranging from "There is only one class" or "pure point-buy" to "Every class works completely differently and you have to learn complex new rules for each one." 4e, in my opinion, got a little too close to "only one class," although I don't actually think it was as bad as some detractors would say. 3.5e was closer to having a new subsystem for each class. (Magic, psionics, ToB, some of the oddities like Truenaming and...Incarna? It's been a while...but they were all very distinct from each other and from your pure martials.) I'm aiming to take just a couple of steps back: there are differences in what the powers of different power sources do and how classes interact with those powers, but it all occurs within a unified rule system and you don't have to learn a new system to play a different class. Within those variations, there are groupings and similarities between thematically similar classes.

• Groups of classes that are thematically similar have access to the same powers. Artificers, sorcerers, warlocks, and wizards manipulate the same fundamental forces to achieve broadly the same effects.
• Within those groups, class features alter how the classes use those powers. As mentioned, artificers can give copies of their spells to allies (with some advantages, and some restrictions, that I didn't describe). Sorcerers are immune to their own area spells and can add a mark-like effect to them. Warlocks can augment spells to deal additional damage. Wizards can make their spells more efficient by extending their durations without spending as many actions as other classes would have to.
• Aside from their power selection, groups differ in the availability of their powers. Martial classes all get a feature letting them use more of their greater powers without resting; Arcane classes all get a feature letting them change what powers they know rapidly. Sorcerers can retrain a power after every victorious encounter, Warlocks can exchange health for temporary new powers known, Wizards have several spellbooks and can study them to "refresh their memory" and change which pre-defined selection of spells they have ready. Artificers, in a slight departure, can alter the enchantments, and therefore the item powers, on their and their allies' magical items.

[sandmote]

The difference is that 4E gives much more leeway to replace one skill with another by refluffing. Something like "I roll stealth to quietly cross the river" or "I roll history to remember how rivers have been crossed in the past" would not fly in 3E/PF but are common responses in 4E. In all systems players want to use their highest skills but not all systems let them.

I mean, it depends on the situation. "I tumble up the crevasse" is an explicit use the tumble skill in 3.5e, and has a lower DC than climbing up a perfectly smooth, flat, vertical surface, even when the climber can brace against two opposite walls (a DC of 50 for the former and a DC of 60 [70-10] for the latter).

"I try to avoid shifting the rocks sticking up out of the river as I cross over them like I'd avoid squeaking floorboards," I'd probably allow to work, especially without a separate balance skill; I'd probably bump up the DC needed to succeed though. And for skills that make no sense with what the PC is doing, or which you don't want to work in a particular situation, the DMG1's example Skill Challenge on pages 76-77 does give an example of a particular skill giving an automatic failure if a PC attempts to use it.

With a History check, I'd probably ask the player what physical actions the PC takes as part of the check, or list a DC of say, 40+, and ask if they're sure they want to do that. Alternatively, if another PC has just failed a check, I might allow a History check to recall it isn't done that way and have a success here negate the earlier failure (or count as a success itself, to keep things moving and avoid lowering the stakes). The general point being that there's ways to push the PCs away from nonsense without an outright "no" and multiple ways a check can be applied.

This is just bikeshedding. The premise of Skill Challenges, as they relate to social situations, is that some social situations that would be modeled as an unstructured series of skill checks (and possibly other actions) in 3e will instead be modeled as a Skill Challenge. And the reality of Skill Challenges is that in those situations the Paladin rolling Sense Motive with a low chance of success goes from "probably neutral" to "probably negative".

I contest that there is significant difference between saying "social situations are like this" and saying "some social situations are like this." The former is appropriate when one option is forced on users of the system, and the latter when the DM is able to choose how to resolve the situation using this or other mechanisms. The Skill Challenge structure is there in addition to the 3e system of skill checks, not instead of it.

In 4e, social situations are modeled as Skill Challenges. The question is not "who is rolling Insight". It is "is anyone other than whoever has the best bonus at some skill rolling anything at all".

Responding to this situation under the understanding that it pops up when a social situation is actually modelled as a Skill Challenge, it can occur in 3.Xe just as well. In practice, it applies in any situation with a limited amount of time (read: checks) that can be made before failure. A suspicious nobleman who's decided to order the party thrown out if they can't explain what they're doing in his home; if the party looks suspiciously quiet long enough, spouts things he considers nonsense, and/or makes an aggressive move, they're going to fail.

[MwaO]

This is just bikeshedding. The premise of Skill Challenges, as they relate to social situations, is that some social situations that would be modeled as an unstructured series of skill checks (and possibly other actions) in 3e will instead be modeled as a Skill Challenge. And the reality of Skill Challenges is that in those situations the Paladin rolling Sense Motive with a low chance of success goes from "probably neutral" to "probably negative".

4e DMG literally tells you not to do what you are suggesting because it does not contain enough choices of skills to give an entire party a wide range of skills to choose. Yes, 4e plays wildly differently than expected when you ignore the advice given to you by the writers of the book in the section where they describe the option.

If the Paladin with a low chance of success at Insight is rolling Insight in a Skill Challenge, that means one of two things has happened:
It is not a good skill challenge as DMG tells you, because the Skill Challenge does not contain a wide choice of skills. You, the DM, know their skills and should be picking some choices that they could be expected to roll. If you choose not to do that, you are ignoring what the DMG tells you to do.

Your player does not understand the point of a skill challenge is similar to a combat scenario — yes, that Paladin can throw a hammer for a Ranged Basic Attack and ignore their encounter powers too, but just because they can do something bad, does not mean they should be doing the bad thing unless it is the only possible option.

Responding to this situation under the understanding that it pops up when a social situation is actually modelled as a Skill Challenge, it can occur in 3.Xe just as well. In practice, it applies in any situation with a limited amount of time (read: checks) that can be made before failure. A suspicious nobleman who's decided to order the party thrown out if they can't explain what they're doing in his home; if the party looks suspiciously quiet long enough, spouts things he considers nonsense, and/or makes an aggressive move, they're going to fail.

This is not a Skill Challenge. It is a series of checks that is probably actually at most a couple of checks made by one PC because they're the party face.

[RandomPeasant]

I also think that there can be a spectrum of class differences, ranging from "There is only one class" or "pure point-buy" to "Every class works completely differently and you have to learn complex new rules for each one."

Sure. You can have a bunch of classes that use the same (or basically the same) subsystem. That's how 3e Psionics and Tome of Battle work. But I don't really see that as addressing the fundamental point. Yes, if you have classes that are sufficiently similar, they can share powers. But if they aren't, they can't. That's why your Artificer is causing you problems. "Who does this target" is really a very big change! You can't just slap that on to powers and expect it to work easily and not constrain the design space.

I contest that there is significant difference between saying "social situations are like this" and saying "some social situations are like this."

I just fundamentally don't see what "sometimes this is not true" means as a rebuttal to "sometimes this is true". Like, yes, I understand that you still can do non-Skill Challenge checks. But if your defense of a system is that you can just not use it, perhaps the system is not working very well. Are there some significant number of situations where the game benefits from actions falling into exclusively "optimal" and "harmful" buckets? I would say probably not. So it would seem to me that a system that divides actions like that is simply not very good.

In practice, it applies in any situation with a limited amount of time (read: checks) that can be made before failure.

That's not what Skill Challenges do. You can make as many checks as you want. The challenge ends after a number of failures. What you are describing is a good system. It is a system where people are encouraged to participate and take actions, but still to make sure their actions are as effective as possible. Skill Challenges are not that system. At least, not the version in the DMG, and not any of the versions I saw before I stopped caring.

If the Paladin with a low chance of success at Insight is rolling Insight in a Skill Challenge, that means one of two things has happened:

Stop. The problem is not "a low chance of success". The problem is anything other than the highest chance of success. If you want to argue that a Paladin rolling Insight at +3 when the Sorcerer rolls Diplomacy at +12 against the same DC should be a wrong choice, fine. I think that's bad too, but it's not the point. The point is that the Paladin rolling Insight at +11 when the Sorcerer rolls Diplomacy at +12 (again, against the same DC, but you get exactly analogous problems if you vary the DCs), that's wrong too.

that Paladin can throw a hammer for a Ranged Basic Attack and ignore their encounter powers too, but just because they can do something bad, does not mean they should be doing the bad thing unless it is the only possible option.

Ah, but the Paladin throwing the hammer is not a bad thing. It is, as the 3e skill check was, worst-case neutral. If the Paladin throws the hammer and hits, he does some damage and the party makes progress. If the Paladin throws the hammer and hits, he's no worse off than if he'd done nothing. With a Skill Challenge this is not true. If you roll Insight and fail, your party is now one failure closer to the end of the challenge. Unless you are forced to act, it is better to do nothing than to make a skill check that is not the skill check with the best chance of generating a success.

[MwaO]

Stop. The problem is not "a low chance of success". The problem is anything other than the highest chance of success. If you want to argue that a Paladin rolling Insight at +3 when the Sorcerer rolls Diplomacy at +12 against the same DC should be a wrong choice, fine. I think that's bad too, but it's not the point. The point is that the Paladin rolling Insight at +11 when the Sorcerer rolls Diplomacy at +12 (again, against the same DC, but you get exactly analogous problems if you vary the DCs), that's wrong too.

Everyone gets to choose where they put their stats & race & background and unless something is weird going on, everyone is capable of getting the same number on their skills. You are describing a situation where the party has unequally spent resources and wow, the person who spent more resources on being good at a particular skill is better than someone who didn't.

Also, you are supposed to vary DCs on any given Skill Challenge to include Easy, Moderate, and Hard checks, and because you are the DM you have looked at people's character sheets and made choices on that basis. Merely rolling your best skill, if the DM decides that say Intimidate is a Moderate DC and Diplomacy is a Hard DC — aggressive NPC as an example — isn't necessarily the best choice.

It might be helpful for you to reread the section on skill challenges before I answer you again?

Ah, but the Paladin throwing the hammer is not a bad thing.

They are actively increasing the chance of party failure significantly by doing a bad action. Unless that is their only choice because say they are immobilized or something along those lines. If you play in a game where your DM softballs you, sure, there may be no consequences for it. But it is most certainly a significantly bad action compared to almost any choice of any Paladin attack power, particularly when in the example given, your PC has not used up their encounter attack powers yet.

[RandomPeasant]

Everyone gets to choose where they put their stats & race & background and unless something is weird going on, everyone is capable of getting the same number on their skills.

I will give you that everyone having the exact same chance of success in a Skill Challenge resolves the problem with Skill Challenges. I find "make everyone mathematically indistinguishable" to be a less compelling solution than "create a system where players who are less likely to succeed are not discouraged from participating", but I cannot claim my personal preferences are in any sense universal.

You are describing a situation where the party has unequally spent resources and wow, the person who spent more resources on being good at a particular skill is better than someone who didn't.

The problem is not that the +12 bonus is better than the +11 bonus. The problem is that if someone has the +12 bonus, not being there is better than the +11 bonus.

Merely rolling your best skill, if the DM decides that say Intimidate is a Moderate DC and Diplomacy is a Hard DC — aggressive NPC as an example — isn't necessarily the best choice.

Can you explain to me how rolling +12 against DC 22 rather than +7 against DC 17 changes the math in any way I should care about at all?

They are actively increasing the chance of party failure significantly by doing a bad action.

Not relative to taking no action. It is true that if the Paladin takes a bad action the party might lose a fight they'd otherwise win. But they would still have lost that fight if the Paladin had simply declined to show up. With skill challenges, this is not true. If I roll two successes and two failures, and you roll three successes and one failure, we get five successes. But you alone (supposing your performance was statistically typical) would've rolled nine successes. My presence made our results worse. Do you see the problem, or do I need to explain things a different way?

[kieza]

Sure. You can have a bunch of classes that use the same (or basically the same) subsystem. That's how 3e Psionics and Tome of Battle work. But I don't really see that as addressing the fundamental point. Yes, if you have classes that are sufficiently similar, they can share powers. But if they aren't, they can't. That's why your Artificer is causing you problems. "Who does this target" is really a very big change! You can't just slap that on to powers and expect it to work easily and not constrain the design space.

Well, that's the thing, I don't expect it to work easily. That's why I listed it in the Cons section. That one spell was the most egregious example, because it was unique in being a self-only spell that was balanced against the Total Defense action. In the end, I pared the spell down to just grant the temporary hit points, and take a less valuable action than Total Defense.

I think that something 4e did wrong was to tie classes too tightly to their role. It led to a lot of cases where, as mentioned, a class had weird powers because the designers felt it needed powers with particular mechanical effects, so they wrote built the mechanics and figured out the fluff later. And, relatedly, it meant that a lot of powers either weren't available to classes that should reasonably have had them, or had near identical versions for different classes. (Rogues can tumble but not Rangers, rangers can dual wield well but not rogues, etc.)

I didn't like the proliferation of systems in 3.5e, but something I think it got right was that there was overlap between the abilities of similar/related classes: Wizards and Sorcerers had the exact same spell list, lots of divine classes could turn undead, etc. It had more verisimilitude than 4e's paradigm of "oh, he cast magic missile? Must be a wizard, because only wizards can do that." Having that, to me, is worth some tradeoffs, like having looser balance and closing off some corners of the design space. (And if I'm doing a good job, the corners that get closed off will be the remote ones, that are metaphorically full of cobwebs, because the concepts they contain are less popular/useful/irreplaceable.)

[Waddacku]

It's actually frequently better to use powers, items, and rituals because those can give free successes. The PCs are supposed to be busting out their resources in both SCs and combat.

Another thing to keep in mind is that a skill challenges is not a status quo until resolved. The situation changes with the actions taken, whether successful or not. This in turn affects the available further options, because actions have to make sense. Rolling History to know how to climb a wall doesn't replace the actual Athletics check to climb it, but it might get you some benefit for the climb. You might, of course, have a power or other button available to you that will just let you flat out get up that wall without a check at all. And, just to be completely clear, this is all predicated on getting up that wall being relevant to accomplishing the greater goal somehow.

I believe earlier in the thread (or another one) there was some discussion of a hypothetical river crossing skill challenge. The issue I have with the example is that it doesn't sufficiently specify a goal, which you for one need in order to decide what actions are relevant, but more importantly to be able to figure out what success and failure (both per roll and for the entire SC) actually entails. If it's just a simple obstacle for the party to get across, handling them individually or using group checks is probably more appropriate. What stood out to me here is that if you phrase it as "getting across the river with the wagon" things get a little spicier right away, because that makes much clearer stakes: three failures means you have to give up on getting the wagon across. That could mean it got destroyed by the river, or that you had to leave it behind, or that you lose time having to find a different route, or that you can't go to wherever you were going, etc.. Which it is depends completely on the goals of the characters and the actions taken to achieve them. Of course, other things can go right or wrong, too. Swimmers getting swept away and whatnot. The possibilities are endless, but what you do know is if you get your X successes you get across with the wagon and if you get three failures you don't, one way or another. Having those numbers lets players know how close they are to each respective outcome, allowing them to make informed decisions about risk-taking and resource expenditures, based on what's important to them and their characters.

And yeah, absolutely I think SCs should be mostly improvised, because of both how open they are making them hard to write in advance and because they generally ought to be initiated by the player decision to pursue a particular goal in the scene.
I've read decently compelling things suggesting that was also the intention of the designers but it doesn't come out very clearly in the rule books, and AFAIK most published SCs were bad, including the example one in the DMG explicitly not actually following the laid out model. I've read one in a module (The Slaying Stone, which I've seen praised a lot but doesn't look very good) where if the party fails to navigate through a forest they just... have to go back to where they started and can't try again. Bizarre.



Sorry for the rambling, I don't have the energy to proofread or edit before posting. I'll try to explain myself better if there's anything unclear or impossible to follow.

[dgnslyr]

Not relative to taking no action. It is true that if the Paladin takes a bad action the party might lose a fight they'd otherwise win. But they would still have lost that fight if the Paladin had simply declined to show up. With skill challenges, this is not true. If I roll two successes and two failures, and you roll three successes and one failure, we get five successes. But you alone (supposing your performance was statistically typical) would've rolled nine successes. My presence made our results worse. Do you see the problem, or do I need to explain things a different way?

On the other hand, I feel it's necessary to point out that character options are a medium for self-expression, and players choose character options based on what they want their character to do. A player that chooses Insight proficiency on their paladin wants their paladin to show insight and be insightful, and a player that doesn't choose Insight proficiency probably would not be interested in rolling Insight in the first place; they hypothetical non-proficient Paladin rolling Insight fundamentally should not happen because if a player wanted a character that solves problems with their Insight, they'd have taken Insight proficiency in the first place. So it's not a failure of the system to punish a player for rolling a non-proficient skill, but a failure of the player for not having a cohesive idea for how their character is supposed to act.

That said, I'd consider it a failure of the GM if the list of permissible skill checks is excessively narrow, and a player doesn't have even a single proficient skill that can contribute at all. I personally prefer it when the Skill Challenge is left open-ended, and it's up to each of the players to narrate how their character contributes with the skillset they have, but this also assumes that the point of a skill challenge is for players to express how their characters succeed, and not if they succeed.

[MwaO]

I will give you that everyone having the exact same chance of success in a Skill Challenge resolves the problem with Skill Challenges. I find "make everyone mathematically indistinguishable" to be a less compelling solution than "create a system where players who are less likely to succeed are not discouraged from participating", but I cannot claim my personal preferences are in any sense universal.

Everyone in theory can have the exact same chance of success with their top skills as anyone else can. People usually do not because they do not want to do that for a variety of reasons.

Someone might not want their primary and secondary stats to be the same. They might have a racial +2 or background +1 or 2 or an item or feat bonus. These are all choices that the player makes and just as there are some players who focus in on one thing in combat and some players who go for a variety of combat options, there are some players who focus in on one or two skills and some players who go for a variety of skills.

And in practice, both groups of players have mechanical benefits from doing so and it is not clear there is a 'right choice'

[RandomPeasant]

Rolling History to know how to climb a wall doesn't replace the actual Athletics check to climb it, but it might get you some benefit for the climb.

In what sense does it not? Skill Challenge successes are fungible. If I get one from History, that's one less I need from Athletics.

A player that chooses Insight proficiency on their paladin wants their paladin to show insight and be insightful, and a player that doesn't choose Insight proficiency probably would not be interested in rolling Insight in the first place; they hypothetical non-proficient Paladin rolling Insight fundamentally should not happen because if a player wanted a character that solves problems with their Insight, they'd have taken Insight proficiency in the first place.

I think this at once proves entirely too much and is not really responsive.

Yes, players use their skill investments to signal how they want their characters to approach problems. But that doesn't mean that players should never be in a situation where none of the skills they've invested in are relevant, or that the appropriate response to such a situation should be to do nothing. Sometimes a player feels that the action their character would take right now is to roll Insight, and while I don't necessarily think the player should have a particularly high (or even non-zero) chance of success in doing so, the idea that it would be harmful as a general rule seems to go too far in the other direction.

But more than that, the issue is not simply "rolling Insight might be bad". It's that, mathematically speaking, your chance of success falls every time someone rolls a check at worse relative odds than the best check. Maybe the player did invest in Insight, and it just happens that their bonus is smaller than the Rogue's Intimidate bonus. In this situation the player has invested in the strategy, and has signaled that they want their character to solve problems this way and it is still a bad idea for them to use the skill to solve the problem at hand. It's hard to see that as anything but a failure, but it's an inevitable result of the "count failed checks" approach to Skill Challenges.

this also assumes that the point of a skill challenge is for players to express how their characters succeed, and not if they succeed.

That strikes me as a false choice. Also, it's really a lot of rolling to do if we're all meant to accept that the outcome is a matter of course. It seems to me that, even accepting this is the design goal, "each player rolls one check" is still a better solution than all the accounting 4e's system asks you to do.

Everyone in theory can have the exact same chance of success with their top skills as anyone else can. People usually do not because they do not want to do that for a variety of reasons.

So you agree that, in practice, Skill Challenges produce situations where most of the players participating at all is actively harmful?

[MwaO]

So you agree that, in practice, Skill Challenges produce situations where most of the players participating at all is actively harmful?

No. If you make a Skill Challenge, the explicit intent is to engage the entire party and that means you think about the why say a Str/Dex PC is going to be engaged in it. And the Con/Cha PC. And the Int/Wis PC. Or any other set of stats.

If you can't answer those questions for your intended group(s), you don't have a viable Skill Challenge and that either means you have "this should be a series of checks which are not a Skill Challenge" or "think about how to include everyone and repeat until this is true."

And yes, this means if you have a DM who doesn't understand Skill Challenges, yes, you can have a lousy time with them. But you likely have an equally bad time with them as a DM when they're running regular non-combat and then they get bounced from DM'ing duties.

[RedWarlock]

A bad session ran by a DM who understands neither the extent of the rules nor the spirit behind them shouldn't be used by detractors as an indictment of the system. That is in no way arguing in good faith.

[RandomPeasant]

No. If you make a Skill Challenge, the explicit intent is to engage the entire party and that means you think about the why say a Str/Dex PC is going to be engaged in it. And the Con/Cha PC. And the Int/Wis PC. Or any other set of stats.

And Skill Challenges fail at that intent. Thus far, the only solution you've proposed to fix them is "make everyone's odds of success the same", and that solution is so obviously terrible that you won't even defend it! And without that solution it is simply mathematically false that everyone is encouraged to participate. If the Skill Challenge ends after a fixed number of failures (and the rules say you do) and you accumulate less successes after a fixed number of failures with a lower chance of success (and math says you do), then you end up with less successes at the end of the Skill Challenge if anyone but the player with the best skill does anything.

If you see a solution for that, feel free to propose it. But I find "a real DM would just do the mathematically impossible thing" to be less than compelling as an argument.

If you can't answer those questions for your intended group(s), you don't have a viable Skill Challenge and that either means you have "this should be a series of checks which are not a Skill Challenge" or "think about how to include everyone and repeat until this is true."

Here's how you include everyone: don't use the Skill Challenge rules. They structurally do not include everyone, and if you want to include everyone you should use different rules that do not have that flaw. Rules like "a Skill Challenge consists of a fixed number of rounds in which everyone rolls one skill". That includes everyone. This isn't a DMing problem. It's a rules problem, and nothing you can do as a DM will be better than using rules that are good instead of rules that are bad.

A bad session ran by a DM who understands neither the extent of the rules nor the spirit behind them shouldn't be used by detractors as an indictment of the system. That is in no way arguing in good faith.

I don't care about the DM at all. I would just like the "Skill Challenges are fine" people to explain how "we get nine successes if Greg isn't there and give successes if Greg is there" isn't a reason for Greg not to participate. And no one is willing to engage with that point. It's all "a good DM would give Greg something to do", but when you get to the question of what Greg is supposed to do it's just crickets.

[Beoric]

A bad session ran by a DM who understands neither the extent of the rules nor the spirit behind them shouldn't be used by detractors as an indictment of the system. That is in no way arguing in good faith.

Perhaps. But the sheer number of people who appear to be unable to run a good SC, as evidenced by the fact that we are still arguing about it 14 years later, may be either an indictment of the system or an indictment of WotC's ability to communicate the mechanic.

[MwaO]

They structurally do not include everyone, and if you want to include everyone you should use different rules that do not have that flaw. Rules like "a Skill Challenge consists of a fixed number of rounds in which everyone rolls one skill". That includes everyone. This isn't a DMing problem. It's a rules problem, and nothing you can do as a DM will be better than using rules that are good instead of rules that are bad.

You keep saying things such as this which point out you haven't even read DMG's version of Skill Challenges let alone DMG 2/Rules Compendium's versions. So I'm done.

[MwaO]

Perhaps. But the sheer number of people who appear to be unable to run a good SC, as evidenced by the fact that we are still arguing about it 14 years later, may be either an indictment of the system or an indictment of WotC's ability to communicate the mechanic.

Or...1st year adventures had Skill Challenges written before the writers had finished Skill Challenges in DMG, and those skill challenges gave people ideas that were to put it bluntly, wrong. Such that 14 years later, some people think those were the rules rather than no, they were not — they were the playtest rules that as soon as they got playtested and R&D realized what was wrong, they were changed before release...but yet, as always, some people don't read the DMG.

There are still people who think the DM doesn't have the final say in 4e even though the House Rules section of 4e explicitly gives the DM that power.

[Beoric]

Or...1st year adventures had Skill Challenges written before the writers had finished Skill Challenges in DMG, and those skill challenges gave people ideas that were to put it bluntly, wrong. Such that 14 years later, some people think those were the rules rather than no, they were not — they were the playtest rules that as soon as they got playtested and R&D realized what was wrong, they were changed before release...but yet, as always, some people don't read the DMG.

This is part of what I was referring to when I said "indictment of WotC's ability to communicate the mechanic". And as I recall, in general SCs continued to suck in later adventures, and later advice also continued to suck.

I think a common problem was to use SCs for situations where it was really reaching to include a broad array of skills. I seem to recall more than one social challenge where one method of persuading the aristocrat a policy or tactic by allowing someone to use Athletics to show their physical prowess.

The fact is that circumstances that requiring broad variety of very disparate skills based on a broad variety of physical and mental abilities are rare, and will not always occur in an adventure. But (I infer from the fact that I can't think of a single WotC 4e adventure without one) the insistence of including at least one SC in every published adventure, regardless of whether it made any sense, combined with dereliction on the part of any editors, designers or playtesters in reviewing these SCs, meant WotC pushed out a massive amount of bad examples.

I think another problem with SCs is that they support a playstyle that is not universal, and are antithetical to a competing playstyle that is still fairly common. That is, SCs support a playstyle where mechanics drive narration, as opposed to narration driving mechanical choices.

SCs reward choosing approaches to problem solving that chooses a character's best skill, as opposed to choosing the most sensible in-world approach. So players approach SCs by trying to use their characters' best skill, and try to narrate an approach that justifies that skill use (assuming they even bother with the second part). Then, if sufficient successes or failures are realized, the DM narrates the result.

A corollary if this is that skill choice determines the obstacles, rather than the other way around. So a player in a travelling SC may use athletics to "climb a cliff, swim a river or push through dense scrubland"; this is in contrast to a situation where the adventure puts a cliff, river or scrub in front of the party, and the party decides how to deal with it.

Note by implication this playstyle is generally ok with a "closed" encounter that has a limited number of ways to resolve it (X successes before Y failures using listed skills, with further limits on the number of times a skill can be used).

But there are still people who play a narratively driven style, where it is more important to choose a clever approach than to pick your best modifiers. Sure, it is great to play to your skills, but you are better off if you do a smart thing even you aren't good at it, than if you do a foolish thing that you are good at. So a low charisma fighter has a better chance of convincing the king to send reinforcements by making an honest and direct assessment of the situation in rough language, than by doing windsprints to show how tough he is. And a lot of those people prefer "open" encounters where any action can be tried, even things the DM hasn't thought of, and for obstacles to inform their choices, rather than for their choices to inform a justifying narrative.

Both playstyles are valid (although you can probably tell that I prefer the second), but having a mechanic that only supports one playstyle, and then forcing tables to engage with it in every single adventure, was a poor choice if they wanted the product to have broad appeal.

EDIT: It would have been interesting to create an adventure that included both playstyles, say the option of engaging in either a small wilderness pointcrawl or a wilderness travel SC, and see which DMs and players preferred.

There are still people who think the DM doesn't have the final say in 4e even though the House Rules section of 4e explicitly gives the DM that power.

I don't think the House Rules section helps, because it is framed as a way of creating campaign-specific variant rules, as opposed to providing latitude to the DM to make ad hoc rulings at variance with published rules. "House rules are variants on the basic rules designed specifically for a particular DM’s campaign," and the focus of the section is on designing permenant rules.

Any suggestion that this section can be leveraged to allow the DM to make ad hoc rulings was more than overshadowed by commentary admonishing DMs not to vary the functioning of a power or feat just because it made no sense in the circumstances. Not to mention the clearly stated and poorly reasoned justification for proning oozes, which more clearly than anything else drove home the implication that rules should not be varied just because they make no sense.

[MwaO]

SCs reward choosing approaches to problem solving that chooses a character's best skill, as opposed to choosing the most sensible in-world approach. So players approach SCs by trying to use their characters' best skill, and try to narrate an approach that justifies that skill use (assuming they even bother with the second part). Then, if sufficient successes or failures are realized, the DM narrates the result.

If the DM establishes that they should try that, sure, that's the approach to take. As Skill Challenges itself describes in Rules Compendium, that's one approach to try for any given skill challenge. Or the DM might tell them which skills to use, how hard they are to use, etc...you can have scenes or narration of what happened.

Up to the DM.

I don't think the House Rules section helps, because it is framed as a way of creating campaign-specific variant rules, as opposed to providing latitude to the DM to make ad hoc rulings at variance with published rules. "House rules are variants on the basic rules designed specifically for a particular DM’s campaign," and the focus of the section is on designing permenant rules.

The very large point of the house rules is you can change whatever you want, but it works better if you get buy-in from the players — you want to make adhoc changes, you get the buy-in from the players to do that at the start of the game — this was quite common in LFR where DMs would say they wanted to make a ruling and if there was a problem, resolve after the game.

[Beoric]

If the DM establishes that they should try that, sure, that's the approach to take. As Skill Challenges itself describes in Rules Compendium, that's one approach to try for any given skill challenge. Or the DM might tell them which skills to use, how hard they are to use, etc...you can have scenes or narration of what happened.

Up to the DM.

Ok, but whether the players are picking the skills, or the DM is picking the skills, it is still "skills first, narration second". This is a long way from, "There are three paths heading more or less north. Ones you know goes through the highlands, the second goes through the woods, and the third follows the river along the bottom of the valley. What do you do? [After the players choose the valley road] You come to a large boggy area that looks like it extends for several miles. It is very difficult terrain and will significantly reduce your speed, what do you do? [Players ask if they can discern a decent path though the bog, and you let the ranger make a Nature check]."

In the first case mechanics drives player choices and how the situation is narrated. In the second, the situation as narrated drives player choices, and mechanics respond to the choices the players make.

It is worth noting that in my example, every player gets to participate, even if only one player makes a check, because it is a group decision.

[MwaO]

Ok, but whether the players are picking the skills, or the DM is picking the skills, it is still "skills first, narration second". This is a long way from, "There are three paths heading more or less north. Ones you know goes through the highlands, the second goes through the woods, and the third follows the river along the bottom of the valley. What do you do? [After the players choose the valley road] You come to a large boggy area that looks like it extends for several miles. It is very difficult terrain and will significantly reduce your speed, what do you do? [Players ask if they can discern a decent path though the bog, and you let the ranger make a Nature check]."

In the first case mechanics drives player choices and how the situation is narrated. In the second, the situation as narrated drives player choices, and mechanics respond to the choices the players make.

It is worth noting that in my example, every player gets to participate, even if only one player makes a check, because it is a group decision.

It is also worth noting that DMG literally describes obstacles such as these as not a Skill Challenge. So it resolves exactly the way you wanted to do it. Skill Challenges are an add-on to the traditional way skills work in D&D. You still can do the traditional way and you should be doing it most of the time. And then when you feel you can meaningfully engage the entire party with a skill challenge, then you run the skill challenge.

[RandomPeasant]

There are still people who think the DM doesn't have the final say in 4e even though the House Rules section of 4e explicitly gives the DM that power.

And this is just the Oberoni Fallacy. Yes, the DM could replace Skill Challenges with something that wasn't bad. They would then not be using Skill Challenges.

This is part of what I was referring to when I said "indictment of WotC's ability to communicate the mechanic". And as I recall, in general SCs continued to suck in later adventures, and later advice also continued to suck.

Skill Challenges went through a lot of revisions, and had a great many words dedicated to them. I'm quite prepared to accept that somewhere in there there's some good (or even replacement-level) advice. But the problem, particularly from the "why did 4e fail" perspective is that Skill Challenges as they are described in the DMG have incentives that run directly counter to their design goals. And that's inherent to the "count failures" paradigm. Within that paradigm, the only way to fix it is "everyone rolls at equal odds", and that's an obviously-terrible solution. You can patch around it by forcing people to roll, but that is (to my mind) obviously worse than embracing the "count rounds" paradigm where participation is encouraged naturally.

I think a common problem was to use SCs for situations where it was really reaching to include a broad array of skills. I seem to recall more than one social challenge where one method of persuading the aristocrat a policy or tactic by allowing someone to use Athletics to show their physical prowess.

This is a problem because of the "count failures" paradigm. If a failed roll was always (or even usually) neutral, you could simply accept that in some situations the Paladin would make a crappy Intuition roll or the Wizard would toss of a bad Athletics check and that would be fine. As it is fine in 3e when there's a combat encounter with some undead and the Rogue's sneak attack isn't very useful. What becomes an issue is having those actions be negative expected value. Then you get players desperately justifying the thing that doesn't hurt their team even when it doesn't make sense.

I think another problem with SCs is that they support a playstyle that is not universal, and are antithetical to a competing playstyle that is still fairly common. That is, SCs support a playstyle where mechanics drive narration, as opposed to narration driving mechanical choices.

People have made this distinction a number of times, and I just don't buy that it's as strong as suggested. Mechanics and narration exist together, and should work together. I also don't think your example follows very well. Skill Challenges are supposed to have variable DCs, and even beyond that it's not really implausible to imagine a situation where a character is really good at some tangentially-related action but pretty mediocre at other actions, such that they're better off doing the tangential thing even though it's not as impactful. You might imagine a skilled negotiator with no wilderness survival skills being better off trying to get a good deal on supplies rather than charting a course or clearing jungle, for instance.

The issue, again, is that Skill Challenges push you very hard towards taking the mechanically best action, whatever that action is. And that in turn means that when there is a small delta between "how the flavor and mechanics do interact" and "how the flavor and mechanics should" interact, people are pushed very sharply into that gap. People are very loss-averse, so penalizing sub-optimal actions pushes behavior in entirely predictable, and largely negative, directions.

A corollary if this is that skill choice determines the obstacles, rather than the other way around. So a player in a travelling SC may use athletics to "climb a cliff, swim a river or push through dense scrubland"; this is in contrast to a situation where the adventure puts a cliff, river or scrub in front of the party, and the party decides how to deal with it.

I think here you get into a very thorny question about where challenges come from in the first place. One of the things I'm signaling to the DM when I invest in Athletics is that I'm interested in having challenges that involve climbing, swimming, jumping, or other physical activity. If I get more challenges like that, on some level that's the game working correctly. Certainly, when I'm confronted with a situation that is genuinely ambiguous, I'm going to try to solve it that way (e.g. climb the tower instead of bribing the guards). But it's also true that when you drop that veil it pisses people off for entirely understandable reasons.

Any suggestion that this section can be leveraged to allow the DM to make ad hoc rulings was more than overshadowed by commentary admonishing DMs not to vary the functioning of a power or feat just because it made no sense in the circumstances.

It's also literally just the Oberoni Fallacy. Yes, if you change the rules they do different things and may not have flaws the printed rules have. That's an admission that the printed rules are bad, not an argument against that premise.

In the first case mechanics drives player choices and how the situation is narrated. In the second, the situation as narrated drives player choices, and mechanics respond to the choices the players make.

But those player choices can't be entirely extricated from the mechanics either. Suppose your choices are something like "haunted graveyard", "frozen tundra", and "trapped tomb". If your party has kitted up with a bunch of anti-undead abilities, the haunted graveyard might be the obvious choice, on the expectation that it will have undead enemies that the party is well-equipped to counter. Conversely, if they've got a bunch of survival powers and fire magic, the frozen tundra probably looks like the right call. And if they have two different skill monkeys and someone with a repeatable summon for brute-force trapfinding, the trapped tomb could be a cakewalk. The players are making choice for their characters, but those choices are made in light of mechanics that represent both character capability and potential challenges. Trying to put either approach first is going to tend to separate them when the goal is for them to work in harmony.

It is also worth noting that DMG literally describes obstacles such as these as not a Skill Challenge. So it resolves exactly the way you wanted to do it. Skill Challenges are an add-on to the traditional way skills work in D&D. You still can do the traditional way and you should be doing it most of the time. And then when you feel you can meaningfully engage the entire party with a skill challenge, then you run the skill challenge.

I once again find "you can just not use it" to be a less than compelling defense of a mechanic. Skill Challenges were one of the big selling points of 4e. If we are mostly supposed to ignore them, that strikes me as a concession that they have failed to provide a useful structure for non-combat challenges and should be redesigned.

[Waddacku]

In what sense does it not? Skill Challenge successes are fungible. If I get one from History, that's one less I need from Athletics.

While the rules encourage the DM to say yes if possible to player ideas, they also emphasize the importance of those actions making sense. That is to say, the coherence of the fiction is prioritized.
Each check is also significant enough to change the situation in ways that affect further decision-making, such as introducing or removing NPCs, gaining or losing tangible resources for the PCs, or enabling new skill uses. That's of course not an exhaustive list. The rules also allow for skill checks that don't directly lead to the resolution of the skill challenge (i.e. successes) to provide smaller benefits to future rolls.
The X-before-3 structure serves as a pacing mechanism for this narration, it tells you when the scene is or isn't resolved yet. That means that each success corresponds roughly to 1/X of resolving the situation in the PCs' favor. This then comes down to the DM's call on whether the player's proposed action (and consequence) is too small or too big to be appropriate for one check.

So in that example, since I have supposed that getting someone (or this particular PC) to the top of the wall is appropriate for one success, I decide that merely pulling up theoretical knowledge on how to make that easier is too little for one (but still potentially worth some other benefit).
I would generally recommend against allowing substituting one skill or ability for another without exceptional reason, since it leads to exactly the kind of collapse into one person rolling their highest skill all the time that you are concerned about.
As for repeated usage of one skill, well, repeating the same action over and over when it doesn't make sense to do it more than once is nonsensical. That basically leaves us with cases where a success is a partial completion of some task (repairing a large piece of machinery, maybe) or an ongoing effort (holding a door against a skeleton horde, maybe), and those are still subject to the situation changing somehow with every check.

There are two more things I'd like to mention briefly that further impact options:
One is that while the SC sets the pacing for the scene, things happening that should terminate it in one way or the other do so. This doesn't mean you should allow a single roll of "I solve the problem", of course, but consider say, a chase in a tunnel. If the tunnel were to collapse between the parties, making passage impossible, the chase is over.
The other I already mentioned in the previous post, but PC resources apply just as much in skill challenges as anywhere else. Relevant powers and rituals can even give you multiple successes at once, instead of or in addition to a roll.
Basically players have massive leeway in thinking up ways to be contribute (or steer things according to their own agenda).
As a DM you also have the option of presenting issues that can't just be handed off to another PC to deal with. Sometimes, it's worse not to answer when spoken to, even if your Diplomacy is bad.

In conclusion, I have to disagree with you. Not all skill checks are worth the same, and successes in the challenge aren't interchangeable since they correspond to different changes in the state of the fiction. Different approaches lead to different end results (and states during the challenge, of course), with one major detail determined by the overall success or failure of the skill challenge.

[Beoric]

It is also worth noting that DMG literally describes obstacles such as these as not a Skill Challenge. So it resolves exactly the way you wanted to do it. Skill Challenges are an add-on to the traditional way skills work in D&D. You still can do the traditional way and you should be doing it most of the time. And then when you feel you can meaningfully engage the entire party with a skill challenge, then you run the skill challenge.

Well, yes, that is kind of my point. There are consequences to the play experience depending on whether you choose a skill challenge, or to run something "the traditional way". But if you write something the "traditional way" a group that prefers a mechanics-first playstyle can just opt to use Athletics to swim the river, and off they go; whereas a group that prefers a narrative-first, problem-solving approach can swim, or build a raft, or fell a tree across the river, or follow the river trying to find a ford or a bridge.

But if you instead make a skill challenge, only the mechanics-first playstyle is supported. The problem-solving playstyle is not supported because the obstacles don't exist until a skill has been chosen. A DM could try to make it into a series of concrete obstacles on the fly, but the mechanics don't support it.

So when you choose to put a skill challenge into a published adventure for activities that could be resolved in the traditional way, you are lessening the play experience of people who prefer that narrative-first approach to gaming. It limits your market, is all.

Hypothetically, you could limit skill challenges to situations that do not lend themselves to being resolved in the traditional way. But if you also require that the SCs employ a lot of different skills, I am having trouble imagining what those situations might be. I suspect they don't arise very often, certainly not at-least-once-in-every-adventure, which makes it kind of a corner-case mechanic.

[PoeticallyPsyco]

Only read the first page so far, so sorry if this has already been said.

Utility Powers. Good idea; giving every class a way to contribute outside of combat without sacrificing build resources that could have been spent on combat. However, you get way too few of them, and a decent number of them are still primarily useful in combat, bringing you right back to having to choose between sacrificing combat potential versus utility potential.

[Waddacku]

Well, yes, that is kind of my point. There are consequences to the play experience depending on whether you choose a skill challenge, or to run something "the traditional way". But if you write something the "traditional way" a group that prefers a mechanics-first playstyle can just opt to use Athletics to swim the river, and off they go; whereas a group that prefers a narrative-first, problem-solving approach can swim, or build a raft, or fell a tree across the river, or follow the river trying to find a ford or a bridge.

But if you instead make a skill challenge, only the mechanics-first playstyle is supported. The problem-solving playstyle is not supported because the obstacles don't exist until a skill has been chosen. A DM could try to make it into a series of concrete obstacles on the fly, but the mechanics don't support it.

So when you choose to put a skill challenge into a published adventure for activities that could be resolved in the traditional way, you are lessening the play experience of people who prefer that narrative-first approach to gaming. It limits your market, is all.

Hypothetically, you could limit skill challenges to situations that do not lend themselves to being resolved in the traditional way. But if you also require that the SCs employ a lot of different skills, I am having trouble imagining what those situations might be. I suspect they don't arise very often, certainly not at-least-once-in-every-adventure, which makes it kind of a corner-case mechanic.

Isn't it the other way around? What you call narrative-first and problem-solving is what skill challenges looks a lot more like how skill challenges should be handled than what you call mechanics-first. It sounds like your issue here is much more with the pre-written, published SCs than with the system itself.

[RandomPeasant]

While the rules encourage the DM to say yes if possible to player ideas, they also emphasize the importance of those actions making sense. That is to say, the coherence of the fiction is prioritized.

Again, "punt it to the DM" is not the sign of a good system. "Have the DM ask the players to make a bunch of rolls that seem like they make sense" is an option that is available in any system. Skill Challenges must be evaluated based on what they bring to the table beyond that.

The X-before-3 structure serves as a pacing mechanism for this narration, it tells you when the scene is or isn't resolved yet.

You are once again describing something that seems vastly better handled by having the Skill Challenge last a given number of rounds, rather than ending after a given number of failures. This has the added benefit of providing a natural mechanism for degrees of success (which I recall being another Skill Challenge design goal).

I would generally recommend against allowing substituting one skill or ability for another without exceptional reason, since it leads to exactly the kind of collapse into one person rolling their highest skill all the time that you are concerned about.

No it doesn't. It leads to a collapse of each player rolling their highest skill, but that's really much less of a problem, because it still means everyone is participating. What leads to the collapse into one player rolling is counting failures, and it does that no matter what skills you're letting people roll. Seriously, do the math: are you more likely to get six successes before three failures rolling at a 40% chance of success or a 60% chance?

As for repeated usage of one skill, well, repeating the same action over and over when it doesn't make sense to do it more than once is nonsensical.

What if it does make sense to keep doing? The solution here seems to be "the DM needs to craft every Skill Challenge so that the optimal skill changes over time", but that seems like a much bigger ask of the DM than simply providing a system where you don't have to roll the optimal skill to make a positive contribution.

Relevant powers and rituals can even give you multiple successes at once, instead of or in addition to a roll.

Can you point to any mechanically-supported examples of this?

I have to disagree with you.

I would love for just one of the people disagreeing with my mathematical claims to present some mathematical analysis of why they are wrong.

[Beoric]

Isn't it the other way around? What you call narrative-first and problem-solving is what skill challenges looks a lot more like how skill challenges should be handled than what you call mechanics-first. It sounds like your issue here is much more with the pre-written, published SCs than with the system itself.

Well, no, but I think I get the confusion. The problem is I am making up defined terms where AFAIK none currently exist, and we have a limited number of relevant words that could be included in the defined term. So I get why you might think the phrase "narrative-first" could be applied to SCs, because SCs involve a lot of narration; but a SC is not driven by the narration, it is driven by the mechanics. That is why I call a SC "mechanics first". But maybe it would be clearer if I call it "mechanics driven".

I chose the phrase "narrative-first" to describe traditional encounters because in earlier versions of D&D, before skill systems, ALL noncombat issues (and many combat issues) were resolved nonmechanically, through narration. So a player would ask questions about the situation until he thought he understood what was going on, and then he would describe what action his character attempted narratively, with no reference to mechanics whatseover. The DM might make up an ad hoc mechanic, assigning the odds of success he thought was reasonable and rolling a die. But the player's description of the action was purely narrative.

SCs do not involve problem solving because the problem is too abstracted. You are choosing a skill, making a check, and then describing how the character helped solve the problem; that is you are narrating a solution or failure, you are not solving or failing to solve the problem. But a traditional encounter, or series of encounters, the obstacles to resolving the encounter(s) are concretely defined, and invite players to come up with creative solutions.

Let me illustrate this with an example. It isn't equivalent to a skill challenge, but maybe just a couple of skill checks to keep it simple. So it isn't an example of a complex encounter, it is just an example of the "narrative" part of a narrative driven encounter, and the "problem solving" part of a narrative driven encounter.

Let's say you find a chest which might contain treasure, and you are concerned it might be trapped. With late editions of D&D you would probably have the rogue make a skill check to detect the trap, and a skill check to disarm it if one was found. In very early D&D, by contrast, there were no rogues. Even when rogues (actually thieves) were introduced, they had really poor chances of detecting traps at low level, and had a really high chance of being instantly lethal. So if you could find a way to detect and disarm a trap without relying on the rogue, that is what you might do.

So you start inspecting the chest closely - without touching it if you can get away with it. Are there any visible holes where darts or poison needles could come out, or decorations that might double as a dart-hole cover? Can the lid be opened slightly, and to there appear to be any trigger mechanisms resting against it? If you slide a piece of parchment between lid and chest, will it move all the way around, or does it bump into something that might be a trigger? Does it rock slightly, like it was sitting on a pressure plate?

If you fail to detect a trap this way - narratively - then you let the thief make a check. But if the thief, unsurprisingly, fails to find a trap, then what? You can take a chance and open the chest. You can see if you can open it from a distance by pushing it with a ten-foot pole or pulling with a rope. You can bash it open with a mace or axe instead of touching the latch. You could try the latch but make sure you are wearing steel gauntlets in case of a poison needle. You could drag it to some stairs and throw it down them (and risk breaking any treasure. Or do anything else you think will improve your chances.

If you do find a trap, then before the thief tries to disarm it you can try to figure out how it works and try to circumvent it. Maybe the dart-hole is on the front so you open it from the side. Maybe you find the trigger and figure out a way to hold it down when you open it. Maybe you plug the holes that shoot darts or poison gas. Maybe, like Indiana Jones, you make sure that the weight in the chest never changes, so as not to trigger the pressure plate it is sitting on; or maybe you tie a rope to it and pull it off the pressure plate from a distance. Maybe you break it open from a distance, or throw it down the stairs.

Now I'm guessing most people reading this thread aren't interested in playing a game with this kind of chest, and that's fine. I'm not advocating for the chest, I'm just trying to give an example of an encounter that is driven by the narrative, not by mechanics; is probably resolved by the narrative, not by mechanics; and requires active problem solving on the part of the player, not the character. Which is why I chose those terms to describe a traditional encounter.

Turning back to skill challenges, even if you make up a SC on the fly, and you accept any reasonable pitch regarding applicable skills, the SC is mechanics driven, and you are solving the problem only in the most high level and superficial way. And that supports a different play style than a narrative driven encounter does, although a narrative driven encounter may also be resolvable by resorting purely to mechanics.

I'm not saying it's a bad playstyle, I'm just saying that it isn't a universal playstyle, and it will exclude some players. Whereas narrative encounters still somewhat support a mechanics-driven playstyle as long as there are skill mechanics that can be engaged; so narrative encounter can support both types of playstyle.