What am I forgetting about 4e D&D?

[Grod_The_Giant]

I've spent a probably unhealthy amount of time thinking about roleplaying systems and "fixes" and "overhauls" of various Dungeons and Dragons editions. And the more I do, the more I find myself admiring things I remember from 4e and thinking that a hypothetical 4.5e could have been really, really good. But it's been a long, long time since I read anything from 4e, much less played it, and I also remember finding a lot of things about it frustrating at the time.

So, uh... what am I getting wrong? My rapidly-self-destructing brain insists that:

• The basic framework of Powers was great. The specific powers usually wound up being bland and hyperfixated on combat, but it meant that every character was getting options at every level. That most of the design work was going into active abilities, not passive class features. That you could easily compare apples to oranges, that you didn't have to repeat basic rules about targeting and partial effects over and over again, that you had a template for easily creating any sort of new magic item or weird ability and slotting it neatly into place alongside the others... you could have used a bit more variability between classes in the distribution of At-Will/Encounter/Daily options and a hell of a lot more (read: any) noncombat abilities, but still.
  
• Skill challenges! They didn't quite work and were all too easy to reduce to just rolling dice over and over again, but they were a step towards more nuanced interaction rules that 5e took a sharp turn away from.
  
• Paragon Paths and Epic Destinies were more than a way to differentiate characters--they were an "excuse" for bigger, better, and weirder abilities at higher levels. People may gripe about the basic Fighter getting superhuman strength, but not so much when he's getting that strength because he's descended from giants, or has a trademark magic item, or covered himself with runes, or... you get the idea. If half your powers come from your class and half from your Race/PP/ED, you wind up with a huge variety of character options.
  
• The tightly bound system math is honestly good--predictable numbers make it way easier to improvise and homebrew things.
  
• The obsession with forced movement and area effects in combat meant that battles naturally encouraged dynamic positioning, without demanding that the GM constantly come up with new ways to keep the combat from devolving into two sides standing still and hacking at each other.
  
• Keywords make abilities easier to read, and reduce the potential of weird broken edge cases.

[animorte]

The more I’ve been reading and discussing, I have also been seeking something simpler (even been working on my own basic system). I never played 4e, but I’ve heard mention enough awful things and good things about it that inspired 5e, One, and several other modern TTRPGs. I do kind of want to see for myself.

[Actana]

The biggest frustration in my memory is simply tracking all the various effects in combat. There eventually just ends up being too many of them, and tracking which durations happen when is a pain in larger combats.

[PhoenixPhyre]

The biggest frustration in my memory is simply tracking all the various effects in combat. There eventually just ends up being too many of them, and tracking which durations happen when is a pain in larger combats.

Yeah. Conditional stacking modifiers are horrible. Especially in a "tightly balanced" game because those end up really mattering.

The other issue with the tight balance is that it was quite fragile--easy to fall behind the curve and be utterly useless. This was due to needing numbers coming from magic items and feats, which pretended to be "choose what works" but really were "you need these specific options to meet system expectations".

[Tanarii]

You've nailed a few of the upsides, but here were the downsides:
- powers etc were overly focused on combat. And you can't just swap out a combat power vs a non-combat power without some notes on equivalence.
- combat required a battlemat to play. This was a huge problem because battlemats slow down combat, which ruins pacing.
- required constant referencing of powers on a card or in the books or online, both character and magic items, during combat. Also slowed down combat, ruining pacing.
- Many modifiers and conditions, again slowing down combat.
- bonuses exceeded the range of the die at high levels.
- tightly bound math meant only a very narrow level range could adventure together, and content had to be designed for a party's level. Unlike 5e, you couldn't have characters 5 levels apart effectively adventuring together.

Replacing 5e spells with 4e-ish layout would be fine. Spell cards are already a necessity for a 5e full caster. And very common keywords referring core terms would be fine. As long as it didn't break things again by changing bonuses to exceed the die, narrowing the range of characters that could adventure together, require a battlemat, or otherwise ruin 5e's blazing fast combat (for WotC D&D combat speeds).

[Grod_The_Giant]

The more I’ve been reading and discussing, I have also been seeking something simpler (even been working on my own basic system). I never played 4e, but I’ve heard mention enough awful things and good things about it that inspired 5e, One, and several other modern TTRPGs. I do kind of want to see for myself.

For whatever it's worth, one of my most recent insane giant homebrew projects (a d20 adaptation of Exalted) ended up looking kind of 4e-like, if you squint-- straightjacket-tight math based almost entirely on your level, lots of forced movement, and 90% of abilities coming in the form of semi-standardized Charms powers picked off a big list. But with waaaayyyy more noncombat stuff.

The other issue with the tight balance is that it was quite fragile--easy to fall behind the curve and be utterly useless. This was due to needing numbers coming from magic items and feats, which pretended to be "choose what works" but really were "you need these specific options to meet system expectations".

Yeah, that's a good point to remember. If you're relying on characters to have certain numbers, you better be damn sure they're getting them. You certainly shouldn't rely on them emerging as the sum of five completely separate features .

[Kurald Galain]

Well, my best advice for you is to use the archive function of this very forum and look through your own posts in the 4E forum from years ago.

That said, from my own memory (having not played 4E for years now),
(1) the basic framework was great, and templates are great, and they should have templated a bit more. For instance, it doesn't help the game that there are easily a dozen "fireball"-type effects, some bigger than others, and some targeting fort instead of ref, or dealing some status condition or whatnot. What 4E does very right is that, if you describe enemies as (e.g.) "eladrin", the players will immediately expect them to teleport all over the place (and they do). What 4E does very wrong is that, if you yell "fireball!", the players will wait for you to explain what this particular fireball does (as opposed to yesterday's fireball).
(2) the biggest issue with skill challenges is probably that they have no way of dealing with items or power usage, and in some situations an item or power is just the best solution.
(3) there are all kinds of issues with items, from option paralysis, to the fact that 90% are vendor trash, to the reuse of "daily" powers by buying copies of cheap low-level stuff, to the christmas tree effect. My experience at least is that 4E plays better and smoother without items (using inherent bonuses).
(4) PPs are great, EDs not so much (well, the fluff is great). This is largely because of what you describe: PPs grant more (and more visible) powers than EDs do.
(5) forced movement is fun! What's also fun is crowd control magic. Or both: use forced movement to bunch enemies together, action point to cast massive crowd control.

HTH.

[PhoenixPhyre]

Yeah, that's a good point to remember. If you're relying on characters to have certain numbers, you better be damn sure they're getting them. You certainly shouldn't rely on them emerging as the sum of five completely separate features .

Agreed. Hard (meaning "things fall apart if not met") system expectations should not be optional.

Personally, I'd be much happier if
* Every "number" (core capability) that a particular class was supposed to have was baked into the class itself. You can have choices within that, but you shouldn't need anything else just to hit the expected bar.
* Everything else (such as feats and magic items) gives horizontal progression, not vertical progression. No feats that make you better at what you're already good at, but feats that let you do things not baked into the class.
* Multiclassing (if it had to exist) is an extreme example of this and is "baked in" so you either get deeper/faster goodies (above system baseline expectations) in your "main" class OR you get horizontal growth, getting specific capabilities from other classes.

--------

In general, 4e had a lot of great ideas...and executed them poorly.
* regularized system (especially monster) math? Great idea. Implementation was rough and often tended toward very interchangeable monsters
* Everyone on the same basic system of powers? Great idea, in principle. In practice, not so much.
* Keywords? They can be great, but they actually add more loopholes. Or they homogenize everything.
* skill challenges? The idea of "this isn't just one roll and done" is good, but the implementation sucked. It should have been a guideline for DMs to make engaging non-combat scenes, where each choice moves the narrative along and it's not just "roll lots of dice at the problem."
* Etc.

[Telok]

The biggest frustration in my memory is simply tracking all the various effects in combat. There eventually just ends up being too many of them, and tracking which durations happen when is a pain in larger combats.

This was big for us. Having four or five effects with that 50/50 save ends each turn on a piece was murder for keeping stuff straight and smooth play. Payed merry hob with stuff that gave/got off turn saves too. A lot of it could have easily been cut to "until the end of your/target's next turn", and would have massively improved some stuff. One of our group started getting pissed because the DM was running support monsters granting off turn saves and all that pc has was save-ends effects. Half the time their power's good effect got stripped off before it made a difference on any important monsters.

The mounted combat rules and mounts were pure bad jank to us.

OAs were awful if you weren't a str based character or didn't have a power to use for them.

Hp bloat was an issue because you expected to blow your main useful powers in the first 4 or 5 turns, but if people missed a bit then you could assume like 3+ turns of at-wills to make up for one missed daily power. So (butt pull some potential example numbers) the seemed to assume around 75% hit rate, at-will being 2x basic attack, encounter being 2x to 3x at-will, and daily being 2x to 3x encounter. If you slammed out your 3 encounters & a daily but instead of missing any one you missed your two best you'd need roughly 3 to 6 turns of at-will to replace your extra miss. If 2 of 4 players have that happen you're generally looking at like two or three extra turns of the whole party doing just at-will spam to get even with where they were supposed to be at the end of turn 4. And that was assuming good decisions on the players side all the time & effective clumping of monsters for the AoEs.

[Tanarii]

(1) the basic framework was great, and templates are great, and they should have templated a bit more. For instance, it doesn't help the game that there are easily a dozen "fireball"-type effects, some bigger than others, and some targeting fort instead of ref, or dealing some status condition or whatnot. What 4E does very right is that, if you describe enemies as (e.g.) "eladrin", the players will immediately expect them to teleport all over the place (and they do). What 4E does very wrong is that, if you yell "fireball!", the players will wait for you to explain what this particular fireball does (as opposed to yesterday's fireball).

Heinsoo (mostly) fixed this in 13th Age. Instead of having a close to identical version with bigger numbers or slightly boosted effects at higher levels, with a new name, many lower level powers could just be selected at the higher level for greater effect. The upgrade was directly in the original power. This saved a lot of space. But it does mean paging back and forth a bit when selecting powers at higher levels, instead of just seeing your 4 possible options right next to each other.

* skill challenges? The idea of "this isn't just one roll and done" is good, but the implementation sucked. It should have been a guideline for DMs to make engaging non-combat scenes, where each choice moves the narrative along and it's not just "roll lots of dice at the problem."

Having imported them to 5e, the biggest issue remains the same: Instead of the players engaging challenges and the DM calling for checks when there is a question of resolution, the DM has to figure out problems with possible approaches ahead of time that are 1 check each, then string together sufficient problems to total the maximum success + failure before the skill challenge fails, but also build in a break out of the skill challenge to continue the game after the minimum successes is reached. And think about how failure won't be a blocker.

It kinda works for something like "our spaceship engine is failing and we need to get it working again" single problems and the players try whatever they can think of to analyze it then fix it. But it's much harder to design a good one for something like "traverse the swamp to the enemy cave" wrapped around several encounters.

[RandomPeasant]

So, uh... what am I getting wrong? My rapidly-self-destructing brain insists that:

Basically, that the implementation sucked. And it extra sucked at the beginning, when people formed their opinions on the edition. RAW Core 4e sucks. It offers an "Epic" progression that steals space from everything else, but isn't fleshed out enough to be worthwhile. Skill Challenges are absolutely miserable (IIRC the problem in that iteration was that the DCs were so high you always failed, but I can't keep the versions straight). A bunch of classes and races were cut, meaning that people couldn't translate characters well. Combat is grindy, and becomes more grindy in boss fights and at high levels. Some classes don't have the abilities they need to make their expected builds work. Some classes don't have races that work with them. Some races don't have classes that work with them. Some of that improves over the course of the edition, but it's never really enough to get it to good (especially because there are some fundamental tensions it's saddled with, like being a stripped-down game with thirty levels of progression).

But, yes, 4e has several good ideas for how to make D&D better, and it is unfortunate that the quality of the implementation has soured people on a lot of them. One thing you don't mention is 4e's monster classes, which while far from perfect are way better than anything else D&D has proposed for that. Saying that a monster is a "Lurker" or a "Brute" conveys more than saying it is a "Dragon" or a "Fey", and the things the latter classifications to convey can be handled more than adequately by subtypes.

The basic framework of Powers was great.

I think this depends a lot on what you mean by "basic framework of Powers". Consistent progression is certainly good, but the obsessive focus on AEDU absolutely was not. I think the absolute commitment to a single progression was also too much, as different ways for class abilities to work naturally push towards different progressions. The Wizard (who has a big book of spells they prepare out of) and the Warlock (who has a pact that grants them specific powers) want to get different amounts of powers and get them in different ways. I think the sweet spot is probably to pick a progression at which people get new levels of powers, and let individual classes set things within that (I'm partial to 1/4/7/10/13/16, vacillating on whether you need the additional level of powers at 19).

Skill challenges! They didn't quite work and were all too easy to reduce to just rolling dice over and over again, but they were a step towards more nuanced interaction rules that 5e took a sharp turn away from.

Skill Challenges are quite close to working. You just need to flip it from "the challenge ends after X failures" to "the challenge ends after X rounds of rolling", and then you have a robust and extensible system for "non-combat stuff". But in all the iterations I paid attention to, they messed with every damn nob but that one.

Paragon Paths and Epic Destinies were more than a way to differentiate characters--they were an "excuse" for bigger, better, and weirder abilities at higher levels.

4e's Paragon Paths and Epic Destinies were the single best way D&D has ever had to deal with the Fighter Problem, and it's not close. By far the worst outcome of 4e's failure is that the well's been poisoned on that.

(2) the biggest issue with skill challenges is probably that they have no way of dealing with items or power usage, and in some situations an item or power is just the best solution.

The biggest issue with skill challenges is that the math doesn't work and the incentives are wrong. Once you fix that, fitting in items and powers isn't too bad. You just let people cash them in for auto-successes (or just to bypass the challenge).

(4) PPs are great, EDs not so much (well, the fluff is great). This is largely because of what you describe: PPs grant more (and more visible) powers than EDs do.

EDs and PPs really have opposite problems. EDs nail the flavor, upgrading everyone into Heroes of Ragnarok and Demigods and whatnot. But they're mechanically lackluster. Whereas PPs have the mechanics, but a bunch of them are things like upgrading your "Fighter" into a "Pit Fighter" (feel the paragon power!). Doesn't help that the game has thirty levels of progression when D&D has never needed more than twenty.

This was big for us. Having four or five effects with that 50/50 save ends each turn on a piece was murder for keeping stuff straight and smooth play. Payed merry hob with stuff that gave/got off turn saves too.

Didn't help that 4e wasn't great at standardizing durations. I think "things end on a save" is fine if you're willing to commit to it, as it sidesteps the problem of things ending without the relevant party noticing, but you need to commit to it. And, yes, that includes standardizing when people get saves (and any "make a new save" effect should fire saves for everything at once).

[PhoenixPhyre]

Having imported them to 5e, the biggest issue remains the same: Instead of the players engaging challenges and the DM calling for checks when there is a question of resolution, the DM has to figure out problems with possible approaches ahead of time that are 1 check each, then string together sufficient problems to total the maximum success + failure before the skill challenge fails, but also build in a break out of the skill challenge to continue the game after the minimum successes is reached. And think about how failure won't be a blocker.

It kinda works for something like "our spaceship engine is failing and we need to get it working again" single problems and the players try whatever they can think of to analyze it then fix it. But it's much harder to design a good one for something like "traverse the swamp to the enemy cave" wrapped around several encounters.

As I said, the concept is good. The idea that a good non-combat scene should involve multiple people doing multiple different things that all advance the narrative/situation toward one of many possible outcomes[1] is good. Necessary, in fact, to make non-combat scenes as engaging as combat scenes...which do that exact thing.

But the implementation as "succeed at X before failing at Y checks" is fraught with fundamental issues. It focuses on pressing buttons (another issue with 4e, which was very hard in the "must have a properly-labeled button to do <thing>" camp) instead of engaging in the fiction. Especially with the emphasis on "reskinning" a keyword-driven core. Instead of acting at a narrative level most of the time, players were expected to act at the game level, playing the rules, not the game. And each individual check didn't actually move the narrative much, making it an exercise in "make good gauge go up before bad gauge goes up too far", like a bad quick-time (in slow motion) event. I'd say that if, instead of "skill check succeeded/skill-check failed", each success and each failure would leave the situation in different states.

Spoiler: A stupid example

Show

Situation--party is convincing the king and his advisors to send aid to a village. Four advisors (economic, military, religious, and magical) + the king. Each one has a small list of things that will convince them (with various DCs) and another small list of things that will make them unpersuadable (with various thresholds for failure). Other things just don't have any effect. Scene ends when there isn't anyone left to change their opinion (or after a certain number of "turns" of no progress in either direction to prevent deadlocks). The magical guy might be more convinced by intellectual arguments (Intelligence (Persuasion)), the economic guy might be vulnerable to blackmail, the military guy might want hard evidence of the threat and the party's capabilities (more martial avenues), the religious guy might want omens (etc) and the king might be weak to having his womenfolk nag him. Or whatever.

Perfect success, then, is convincing everyone to help. Perfect failure is locking all of them into unpersuadable. Depending on who exactly is in what state when the scene ends, the help they send differs. Convince just the king and he'll help, but the help he can send is limited because all the other powers slow-walk things. Convince all the advisors but not the king and you might get no official help, but conveniently a few hundred gold and some mercenary groups, plus some free-lancing wizards and clerics might show up to help, paid for out of personal funds. Etc.

[1] effectively, their actions start filtering out some of the myriad of possible end states and possibly introduce new ones. These aren't pre-determined "must use X to do <X>, Y to do <Y>, etc" steps, these are individual micro outcomes that are all (or at least mostly) interesting in their own right.

[Grod_The_Giant]

Agreed. Hard (meaning "things fall apart if not met") system expectations should not be optional.

Personally, I'd be much happier if
* Every "number" (core capability) that a particular class was supposed to have was baked into the class itself. You can have choices within that, but you shouldn't need anything else just to hit the expected bar.
* Everything else (such as feats and magic items) gives horizontal progression, not vertical progression. No feats that make you better at what you're already good at, but feats that let you do things not baked into the class.
* Multiclassing (if it had to exist) is an extreme example of this and is "baked in" so you either get deeper/faster goodies (above system baseline expectations) in your "main" class OR you get horizontal growth, getting specific capabilities from other classes.

Entirely on the same page here. Numbers are the least interesting thing thing on a character sheet, and tend to be the most confusing part of character building. I'm a fan of using "low/medium/high" proficiency as the only thing to keep track of-- three tiers of bonus is enough to let you differentiate between dabbler, professional, and master, and that's really all you need to do.

As I said, the concept is good. The idea that a good non-combat scene should involve multiple people doing multiple different things that all advance the narrative/situation toward one of many possible outcomes[1] is good. Necessary, in fact, to make non-combat scenes as engaging as combat scenes...which do that exact thing.

The best way I've come up with to handle such things is to structure skill challenges like combat scenes. "Goals" instead of monsters, skill checks instead of attack rolls, environmental hazards and ticking clocks instead of enemy attacks, at least a rough turn order, that sort of thing. The key is to make sure there's a back-and-forth element to the scene.

This tends to be easier in games with less focus on limited-use abilities and damage rolls that aren't completely different than everything else in the system.

[animorte]

For whatever it's worth, one of my most recent insane giant homebrew projects (a d20 adaptation of Exalted)

I appreciate that and I’m giving it a look through just now (might take a couple days).
Excellent introduction, by the way.

The best way I've come up with to handle such things is to structure skill challenges like combat scenes. "Goals" instead of monsters, skill checks instead of attack rolls, environmental hazards and ticking clocks instead of enemy attacks, at least a rough turn order, that sort of thing. The key is to make sure there's a back-and-forth element to the scene.

I recently discovered Worlds Without Number that I’ve been looking through and started a thread on. It directly refers to any type of encounter as a scene and a lot of things seem to key off once-per-scene. I don’t have much experience with it yet, but it looks promising.

[Kurald Galain]

One of our group started getting pissed because the DM was running support monsters granting off turn saves and all that pc has was save-ends effects. Half the time their power's good effect got stripped off before it made a difference on any important monsters.

This is important. The rules are built on the idea that save-ends effects are more powerful than EONT (that's why dailies tend to be save-ends and encounter powers tend to be EONT), but in practice it's the opposite (because of support abilities granting extra saves).
And yes, for smoother gameplay it would be better to have one of them but not both. I'd suggest effects that last until the end of their victim's turn, so that you don't have to remember which character created the effect.

The mounted combat rules and mounts were pure bad jank to us.

I concur. Any time there was a mounted battle (for plot reasons), on the first turn every player would immediately dismount so they wouldn't have to deal with mounted combat rules.

the biggest issue remains the same: Instead of the players engaging challenges and the DM calling for checks when there is a question of resolution, the DM has to figure out problems with possible approaches ahead of time that are 1 check each, then string together sufficient problems to total the maximum success + failure before the skill challenge fails, but also build in a break out of the skill challenge to continue the game after the minimum successes is reached. And think about how failure won't be a blocker.

My question is, why would you want this in the first place? Instead of engaging the players it creates more work for the DM... I'm not seeing the benefit here?

The biggest issue with skill challenges is that the math doesn't work and the incentives are wrong. Once you fix that, fitting in items and powers isn't too bad. You just let people cash them in for auto-successes (or just to bypass the challenge).

Thank you for mentioning incentives here. Basically every purported fix I've seen just tweaks the math without considering the incentives.

EDs and PPs really have opposite problems. EDs nail the flavor, upgrading everyone into Heroes of Ragnarok and Demigods and whatnot. But they're mechanically lackluster. Whereas PPs have the mechanics, but a bunch of them are things like upgrading your "Fighter" into a "Pit Fighter" (feel the paragon power!). Doesn't help that the game has thirty levels of progression when D&D has never needed more than twenty.

I concur.

[Cluedrew]

I don't have much to add about the flaws of 4e, but I will applaud its ability to change things that are broken. The fact the things that it changed to were... well you have plenty of other posts to read about that. But if they had taken the "spirit" of what became 5e and used 4e's structure, the result would have been pretty good.

[Sneak Dog]

4e was great. Awesome design philosophies. The execution was found a little lacking by me though. Most importantly I think the mechanics and fluff were disconnected.

Combat could feel like picking which daily to use, then spending your turns using encounter powers and maybe some at-wills to clean up.
Power bloat. Powers were added not for niches, but for the sake of having more powers?
Non-combat was barely supported. Utility powers turned out to be (mostly) combat utility. I never 'got' skill checks either, and they seem weird. More mechanics than fluff at a type of encounter where I want the reverse.
Skill DCs had a table detailing a DC for every level, implying you should scale DC exactly by PC level. Rather than say having a table for each tier or half-tier to indicate challenges becoming higher tier and warranting higher DCs. (E.g. kicking down a nice wooden door is for your basic chump level 1~5 barbarian, kicking down a proper iron door for a level 11~15 barbarian. And a wooden door is an appropriate challenge for a character of level 2 and one of level 5.)

Monsters were great. Easy overview, no need for cross-referencing. Non-AC defences were awesome. Clear secondary defences to target, by mage and martial alike.
Most importantly though: Paragon paths and epic destinies. A clear indicator that your character goes up a tier, a customisation point to make a character your own. I love them. It's akin to getting a 3e/PF prestige class on top of your own class.

[Quertus]

So, I think I’ll address “the math just works”, Skill challenges, Arangee, and my signature closing.

(It’s mostly already been said, I’m just saying it again)

The Math Just Works

On the one hand, you have tight requirements for exact math.

On the other hand, you have choice wrt your stats, feats, and items that can change those numbers. There’s also buffs, debuffs (I assume), situational bonuses, and… probably more.

Pick one.

Skill Challenges

It’s hard for me to not call 4e Skill Challenges “an idiot’s guide to failing”. Let’s break that down.

Yes, sure, maybe the 4e devs eventually got Skill Challenges “right” (for the definition of “right” that means a) the math actually works; b) it actually achieves their stated goal of getting everyone involved) if you squint hard enough, but when they first came out (ie, when I played 4e), “X successes before Y failures” painfully obviously translates to “only your best character should ever roll, and you’re either an idiot or guilty of PvP levels of sabotaging the party if you even look at the dice when you don’t have the highest bonus in the group”.

But eventually they “fixed” that, to where you’re a liability just by existing. The 4 person party only needs X successes in Y rounds, but if they add The Load, they now need X+Z successes in those same Y rounds. Or is that 5e?

But, worst of all (IMO), they’ve mistaken “rolling the dice” for “playing the game”. The key is “making meaningful decisions”, which Skill Challenges decidedly lack.

And the worst part of that is (and I’m going to word this more poorly than it’s already been said) it’s focused on “playing the system” rather than “roleplaying the character”.

If your goal was to kill the fiction and the narrative, there’s few better tools than 4e Skill Challenges.

Arrange

downsides:
- bonuses exceeded the range of the die at high levels.

Here’s what I just heard: gradeschool math whiz could out-punch gradeschool bully, but Bruce Banner cannot out-punch Hulk; gradeschool bully could out-think gradeschool math whiz, but Hulk cannot out-think Bruce Banner.

Is that what you meant? If so, why is that a problem? If not, what is the problem inherent in what you meant?

Closing Comment

I’ve heard people complain that 4e wasn’t D&D. A lot. I think that they’re wrong. I think that 4e was D&D, but it wasn’t an RPG. (If I didn’t say it, you’d think I was an imposter).

The big difference between a war game and an RPG is that everything in a war game exists “inside the box”, as discrete buttons with defined in-game effects; in an RPG, you have access to “outside the box” thinking. The extent to which the system discourages the GM from accepting non-button, “outside the box” actions is the extent to which the system is unsuited to being an RPG. The extent to which the players are encouraged to play the system rather than roleplay their characters (EDIT: for the outcome of their actions being possible to adjudicate, rather than merely being suboptimal) is the extent to which the game is unsuited to being played as an RPG.

RPG or war game; inside the box & outside the box or buttons only. Pick one. And describe what you produce accordingly.

[Slipjig]

Going to the non-mechanics side for a moment, one thing that I thought was really cool about 4e was that different classes were viable to perform different roles in a party. This meant that you could have a viable party composed entirely of arcane characters. Or a viable party made of only divine characters. Or a viable party made up of non-magical characters.

[CarpeGuitarrem]

Also keep in mind that as time passes, the light that you view older games through changes. As you spend time with a game, you learn to accept or work around things that initially irritated you, and it becomes a lot easier to have fun with it, without being bothered by the negatives. And, since there isn't an edition war raging any more, it becomes easier to distinguish your own opinions from the RPG hivemind. So that's another likely factor. Now that 5E is out, there's not really any reason to view 4E as a competitor against the old guard, so you can approach it on its own terms.

[RandomPeasant]

And yes, for smoother gameplay it would be better to have one of them but not both. I'd suggest effects that last until the end of their victim's turn, so that you don't have to remember which character created the effect.

I think save ends is better. You want stuff to last multiple rounds, but still be able to end without someone spending resources. IMO your durations should be "save ends" (most status conditions) or "doesn't end on its own" (for out-of-the-fight conditions like "dead", but also easy-to-remove ones like "prone"). But, again, I broadly agree that you want to minimize the number of types of duration.

Combat could feel like picking which daily to use, then spending your turns using encounter powers and maybe some at-wills to clean up.

To expand on this, fights tended to be long enough that you would use all your Encounter powers in each one, meaning that the only real agency you had was "what daily do I use" and "how do I sequence my encounter powers". And neither of these were really engaging questions. As a result, you had very long fights with very few real decision points, which is a pretty miserable combination.

Skill DCs had a table detailing a DC for every level, implying you should scale DC exactly by PC level.

Stripping out the one layer of abstraction between "level" and "DC" ended up being really hard for people to swallow.

Easy overview, no need for cross-referencing.

But also very little consistency between entries, and a great deal of difficulty doing customizations. It was easier to understand one 4e monster than one 3e monster, but the scaling was awful because so much was unnecessarily bespoke.

but when they first came out (ie, when I played 4e), “X successes before Y failures” painfully obviously translates to “only your best character should ever roll, and you’re either an idiot or guilty of PvP levels of sabotaging the party if you even look at the dice when you don’t have the highest bonus in the group”.

This is exactly the incentive issue I was alluding to. I genuinely do not understand how a professional game designer approved that set of mechanics for that design goal. It's the closest thing I'm aware of to an objective failure in the subjective realm of game design, because they told us what they were trying to do and then did not do it.

But, worst of all (IMO), they’ve mistaken “rolling the dice” for “playing the game”. The key is “making meaningful decisions”, which Skill Challenges decidedly lack.

I actually think that's somewhat unfair, as you're looking at the basic engine and complaining that it doesn't have choices in it, when that's true of the basic engine of most parts of most games. If you strip it down to the very basics, combat is "roll attack, roll damage, see who falls over first", and that doesn't have any agency to it either. The problem was that, because the core of Skill Challenges didn't work, they spent all their time iterating on it instead of adding flourishes that could be used to make individual skill challenges more interesting.

Suppose, for a moment, we had a "you roll for X rounds, then you count up the successes over those rounds" framework for skill challenges. What are some ways you can add complexity on top of that? An obvious one would be to add rounds of something else between skill challenge rounds. Maybe you are navigating a jungle, and you do a combat encounter between rounds of skill challenge, allowing you to decide how you prioritize your resources between those components. Another option would be to change the expected value of rolling specific skills. Maybe one skill counts for two successes, but can only be tried once. Maybe one skill counts for the normal +1 on success, but counts for -1 on failure. Or you could allow people to cash in limited resources for successes. Or you could give people abilities that interact with skill challenges directly. There's plenty of stuff you can add once the basic framework works.

Going to the non-mechanics side for a moment, one thing that I thought was really cool about 4e was that different classes were viable to perform different roles in a party. This meant that you could have a viable party composed entirely of arcane characters. Or a viable party made of only divine characters. Or a viable party made up of non-magical characters.

That wasn't really unique to 4e. Well, the all-martial party largely was, but that's because 3e gave martials the short end of the stick. But your all-arcanist party was just fine once they printed the Beguiler for an arcane trapfinder (healing was a bit of a pain, but could be handled through items, PrCs, or a Bard). The all-divine party was similarly fine once they printed that trapfinding Domain (or just if you played a Rogue dip, but you could reasonably call that cheating). I actually think, while 4e's monster roles were good, 4e's player roles largely were not. People should be allowed to play the class they want, fitting that into whatever specific framework of "roles" you've devised almost never makes them happier.

[Telok]

Here’s what I just heard: gradeschool math whiz could out-punch gradeschool bully, but Bruce Banner cannot out-punch Hulk; gradeschool bully could out-think gradeschool math whiz, but Hulk cannot out-think Bruce Banner.

Is that what you meant? If so, why is that a problem? If not, what is the problem inherent in what you meant?

Here's what I think was meant, taken from my group's actual game. By 12th level we hsd a druid character with a +10 perception bonus over the next highest perception bonus in the party, and a rogue with a stealth bonus more than +10 over the next highest stealth bonus in the party.

Since 4e's skills was basically 5e's skills & bounded accuracy, just with the +1/2 level on top, it has/had effectively all the same issues if the DM followed the suggested DC chart. Sure the description of the door you kicked in or the lock getting picked changed, but in practice since you never went back to repeat old tasks it was a Red Queen's Race... unless you hit on a few bonuses that stacked. 3e intended stacking bonuses and made efforts at keeping game play formed around that, 4e just assumed you were always on the d20 RNG.

So we had a rogue for whom the take-10 option was equal to anyone else's nat-20 roll, and a druid with similar perception. The DM had problems with that. Especially with the stealth & perception numbers hard coded into the monsters. You can sort of do similar in 5e, because the monsters are all basically hard coded DCs to hit, but they did manage to reduce the number of stacking bonuses plus remove the player's option to take 10 and how generally useful the checks are.

[Grod_The_Giant]

On the other hand, you have choice wrt your stats, feats, and items that can change those numbers. There’s also buffs, debuffs (I assume), situational bonuses, and… probably more.

Though it's important to note that you can still have choice, it'll just have to manifest in ways other than "+2 attack."

I’ve heard people complain that 4e wasn’t D&D. A lot. I think that they’re wrong. I think that 4e was D&D, but it wasn’t an RPG. (If I didn’t say it, you’d think I was an imposter).

The big difference between a war game and an RPG is that everything in a war game exists “inside the box”, as discrete buttons with defined in-game effects; in an RPG, you have access to “outside the box” thinking. The extent to which the system discourages the GM from accepting non-button, “outside the box” actions is the extent to which the system is unsuited to being an RPG. The extent to which the players are encouraged to play the system rather than roleplay their characters (EDIT: for the outcome of their actions being possible to adjudicate, rather than merely being suboptimal) is the extent to which the game is unsuited to being played as an RPG.

RPG or war game; inside the box & outside the box or buttons only. Pick one. And describe what you produce accordingly.

Was 4e really that much worse than, say, 5e in that regard? In both editions, noncombat interaction rules boil down to skill checks and a subset of utility spells-- 4e botched the execution in a lot of ways, but fixes are pretty easy to come up with.

[Kurald Galain]

So we had a rogue for whom the take-10 option was equal to anyone else's nat-20 roll, and a druid with similar perception. The DM had problems with that.

The solution to that strikes me as obvious: don't allow taking 10. In fact, IIRC "take 10" was never a part of the 4E rules in the first place.

Other than that, there's no difference in practice between "the rogue automatically makes his stealth checks against most enemies" and "the wizard can cast Invisibility". In principle I have no problem with either of those.

[Kurald Galain]

Was 4e really that much worse than, say, 5e in that regard? In both editions, noncombat interaction rules boil down to skill checks and a subset of utility spells-- 4e botched the execution in a lot of ways, but fixes are pretty easy to come up with.

No, 5E is largely the same. For all its talk of "three pillars", 5E largely lacks rules for out-of-combat situations (other than "roll a skill and make something up"). Personally I think that's a loss for both 4E and 5E.

[Tanarii]

My question is, why would you want this in the first place? Instead of engaging the players it creates more work for the DM... I'm not seeing the benefit here?

1) It helps keep me focused on the idea that everything that is a challenge is an encounter. Doesn't matter if it's a tricks/traps challenge, a social challenge, or an exploration challenge wrapped around several combat / tricks&traps encounters, etc.
2) It allows me to determine the difficulty of the challenge (similar to using the combat encounter difficulty guidelines), and if I'm creating for a specific level design non-combat encounters to a specific difficulty by judging the number of component challenges or decent DCs to pick. Also resource usage to bypass skill checks for an automatic success can be built in, meaning I'll think about "expected" resource usage for non-combat challenges, which is a requirement in 5e for encounter difficulties above Easy. (Even if they use skills to avoid spending resources.)
3) Knowing the difficulty means I can accurately award XP for non-combat encounters, since that's what the award is based on.

It's a useful game structure in many ways. But that usefulness breaks rapidly, as with any game structure, if the players stop playing their characters and start playing pieces, or if they have to start pixel bitching to guess the right path forward. (The "playing pieces" thing is problem that the D&D combat game structure already suffers from in many ways.)

[PhoenixPhyre]

Here's what I think was meant, taken from my group's actual game. By 12th level we hsd a druid character with a +10 perception bonus over the next highest perception bonus in the party, and a rogue with a stealth bonus more than +10 over the next highest stealth bonus in the party.

Since 4e's skills was basically 5e's skills & bounded accuracy, just with the +1/2 level on top, it has/had effectively all the same issues if the DM followed the suggested DC chart. Sure the description of the door you kicked in or the lock getting picked changed, but in practice since you never went back to repeat old tasks it was a Red Queen's Race... unless you hit on a few bonuses that stacked. 3e intended stacking bonuses and made efforts at keeping game play formed around that, 4e just assumed you were always on the d20 RNG.

So we had a rogue for whom the take-10 option was equal to anyone else's nat-20 roll, and a druid with similar perception. The DM had problems with that. Especially with the stealth & perception numbers hard coded into the monsters. You can sort of do similar in 5e, because the monsters are all basically hard coded DCs to hit, but they did manage to reduce the number of stacking bonuses plus remove the player's option to take 10 and how generally useful the checks are.

Except that 4e also had
- Bear lore making a hash out of the world
- nothing like bounded accuracy--in fact, it had the opposite. If you weren't specialized in <skill>, you had no real chance of making an on-level skill check. And if you were specialized, below-level skill checks were trivialized. This meant that you had to carefully plan out as a party who was going to specialize in what, and then those people (and only those people) got to make checks using that skill. Which meant skill challenges didn't work unless you had a wide range of "levels" (DCs) involved or you contrived it to hit everyone's specialized skill. At which point you might as well just say "ok, each of you roll your highest skill. If more than X succeed, you succeed."

It also had severe issues with ludo-narrative dissonance due to overuse of generic keywords. Proning an ooze, for example.

[kieza]

4e was great. Awesome design philosophies. The execution was found a little lacking by me though. Most importantly I think the mechanics and fluff were disconnected.

This is exactly how I've felt about it for years. I hated the proliferation of subsystems and progression rates in 3.5e, and I loved that 4e made:

• Martial classes get cool things like magical classes.
• All powers work the same (attackers roll to hit).
• Attacks, defenses, skills, etc. nominally improve at the same rate.

But I feel like there were three massive, far-reaching flaws:

• Having all classes get cool things (powers) doesn't mean they all have to get the same kind, or at the same rates.
• A lot of powers felt like they were designed by picking effects out of a hat and writing fluff around them. (Consider, e.g., the Sorcerer power Acid Typhoon: "Okay, the hat says we have to write an area power with a non-standard area of effect, that targets Fortitude, deals acid and thunder damage and causes ongoing damage as a side effect.") End result: lots of random powers that are super-powerful because they got the right combination of effects, but have outright weird fluff that makes them hard to fit into a character's theme. And lots of powers that are "lower-level power upgraded for high-level play."
• The math didn't work out as well as it was supposed to. Maybe the designers just couldn't do math, but I read some convincing speculation early on that the reason PC attacks and defenses didn't increase as rapidly as monster attacks and defenses might be that the designers assumed at high levels PCs would have near-constant boosts from a leader-type character that made up for the difference, which...didn't necessarily happen. When it did, the fiddly little bonuses were a pain to track, too.

I have a system I've been writing for a while that started out as a 4e retroclone, and those are the major flaws I set out to avoid replicating.

[Kurald Galain]

1) It helps keep me focused on the idea that everything that is a challenge is an encounter. Doesn't matter if it's a tricks/traps challenge, a social challenge, or an exploration challenge wrapped around several combat / tricks&traps encounters, etc.
2) It allows me to determine the difficulty of the challenge (similar to using the combat encounter difficulty guidelines), and if I'm creating for a specific level design non-combat encounters to a specific difficulty by judging the number of component challenges or decent DCs to pick. Also resource usage to bypass skill checks for an automatic success can be built in, meaning I'll think about "expected" resource usage for non-combat challenges, which is a requirement in 5e for encounter difficulties above Easy. (Even if they use skills to avoid spending resources.)
3) Knowing the difficulty means I can accurately award XP for non-combat encounters, since that's what the award is based on.

I'm curious if we, as the playground, can make rules that meet those goals without requiring the DM (as you mentioned earlier) "to figure out problems with possible approaches ahead of time that are 1 check each".

[PhoenixPhyre]

I'm curious if we, as the playground, can make rules that meet those goals without requiring the DM (as you mentioned earlier) "to figure out problems with possible approaches ahead of time that are 1 check each".

I really don't think you can create generalized rules, at least if you want to have any kind of level of detail/granularity. There's just way too many possible (and incompatible) situations.