D&D 4e Retrospective

Your suggestion of tags would help you group abilities by role, and you could still have your default collections of abilities which you call "ranger" and are typical abilities for ranger character concepts.

but yes, an actual point buy system with the classes as example progression paths, with maximums and minimums on important core math balance mechanics like attack bonus and damage and hp would be what i'd view as the best approach to multiclassing.

Because they're classes, not point-bought features

A class is, almost by definition, a sort of specialization, they're features should mesh with each other to make a whole stronger than the sum of it's parts. Multiclassing comes into this as a means to specialize into something not covered by the base classes. Usually something silly, like the iaijutsu katana chucker mentioned in the previous thread.

The worst kind of multiclassing is what we see in 3.X fighters, which exist primarily as a two-level dip for martials to get some extra feats, that's boring as hell

my point is: why should i be punished for not following a single predefined path of progression? what is the justification for such, in terms of game balance?

personally, i see no such reason.

4e multiclassing was good

Even if the majority of weapon-using classes just multiclassed into fighter because fighter feats and PPs were so god damn good

If you're playing a class based system, you're opting to follow a single path of progression with a few branches and choices along the way. That's literally the point of a class based system.

Level by level multiclassing is just a convoluted mess.

You should be punsihed because you weren't following the pre-determined path of progression

That's what the path is for, if you're going to stray off the beaten path, you should only do so when you have a damn good idea of what will come of it, otherwise you're just asking for trouble

It's like you built a fighter with negative strength and massive intelligence and charisma, and now you're complaining about how much worse you are than everyone else, there's a point where it's not "ivory tower game design", it's just you being stupid

You shouldn't be. Classes should have broad concepts, but focused mechanics.

Mage hunter is a narrow concept with broad possible mechanical implications. It's window dressing.

You can be a mage hunter fighter, by being a fighter who resists magic really well. You can be a mage hunter thief who silences mages before they can cast spells. You can be a mage hunting Cleric who breaks enchantments left and right. You could be a mage hunting mage, who sucks the magic out of other mages to make his magic the best.

You don't need all of the above abilities in one package, and having all of those abilities in one package is redundant as well as cripplingly overspecialized.

You don't need a new class for that, and the fact 3.PF trained a generation of gamers think they do is fucking obnoxious.

I'd argue the sole benefit of classes is simpler chargen for the players who want it. level by level multiclassing (while poorly implemented in 3.x) is for players who want more customization; but the best case implementation of such is effectively a well-balanced point buy system.

The continual proof of the '3.PF causes brain damage' meme.

im not talking about choosing abilities that mesh together poorly. im criticizing the notion of designing abilities deliberately to be/make you more shitty if you multiclass at all because they dont stay level-appropriate unless you don't multiclass.

>Your suggestion of tags would help you group abilities by role,

Yes, but it would not eliminate the main problem, which is that instead of having an actual progression with focused roles, you are playing a game where you can basically grab any ability you want.

Even putting aside the mechanical implications, heavily encouraging picking up "combo" abilities like in a freaking card game (i.e. "strikers with defender features", which is why a bunch of the best 4e hybrids that aren't focused on exploiting above par essentials hybrids and battle cleric lore are defender | strikers), you still end up with player characters being all over the place in competency and direction.

I think you misread my post. im not talking about a "mage hunter" class.

I'm asking for a game balance reason why fighter1/slayer4/monk 3/paladin2/whatever6 is any less valid a way to build a character of a particular concept than "ranger 16".

Unless you're ascribing to the class=concept+role premise
seems to be suggesting, in which case, I guess the reason is "I want a game that only allows for this small fixed list of character concepts".

> I guess the reason is "I want a game that only allows for this small fixed list of character concepts".

But that's bullshit you fucking retard

Way to refute him, man. Well done.

>I'm asking for a game balance reason why fighter1/slayer4/monk 3/paladin2/whatever6 is any less valid a way to build a character of a particular concept than "ranger 16".

You dense motherfucker.

You shouldn't NEED to level-by-level multiclass in the game. You should be able to do your concept by picking the right class and sticking with it. Maybe poach a few abilities with feats or IDK.

Your "way of building a character" has no reason to be less valid in 3.PF. But 3.PF is a fucking terrible mess that misses the entire fucking point of classes, in favor of serving the MtG deckbuilding, Ivory tower loving community (which I'm not saying aren't enjoyable, but aren't what class based systems are for).

>a game where you can grab any ability you want
i dont see that as a bad thing.

>striker with defender features = maximum minmaxery. combo features! supermen!
sure, depending on implementation, thats possible.
but if its done well, and you have stats to be a striker, those defender features will be far less effective than if you were statted to fill a defender role.

and if you went evenly split between the two ypu might be adequate at both, but you'd be less effective at either than the character who specialized in one party role.

>What did it do right?
It gave a really solid foundation for a Gamma World game.

Because balancing for that is exponentially harder than balancing for single classes or even split multiclasses.

never said i should "need to", im asking for a reason it is bad to have that option.

people seem to do okay balancing point buy systems, i'd imagine thats the approach you'd want to take to balancing such things, even if the consumers just see packages of individual levels with occasional options rather than the explicit point values and maximums/minimums.

Because having that option means you have level by level multiclassing, which means you have successfully removed the "specialized characters" benefit of class systems.

Agreed. They stuff they trimmed off made it awesome.

You mean niche protection?

Why does that need to be by class, rather than by individual character in a party.

IE: rather than build a party of twf rogues, or a party of somewhat adequate generalists, the group explicitly builds characters to fill different party roles.

I've never had issues with the unspecialized characters in shadowrun 4e, and thats entirely point buy.

If you're doing a class-based system that allows multiclassing and is meticulously balanced so that rampant multiclassing doesn't drop the quality of a character below a single-class character or raise it significantly above a single-class character, you really need to ask yourself "Why aren't I just making a point-buy system?"

Level-by-level multiclassing should be niche, it should be risky, you shouldn't be rewarded for doing it often, otherwise you might as well do away with classes entirely

>You mean niche protection?

If he meant niche protection, he'd have said niche protection.

Winddancer 1 gives you Flight, Assassin 1 gives you Death Strike and Glass Mage 1 gives you Invisibility that isn't broken by attacking, so Winddancer 1/Glassmage 1/Asssassin 1 is super broken because you can invisibly glide over to your target and Death Strike them each round! So OP!

Except this never happens because this is a class based system so you can never have all three abilities at the same time - Glass Mage doesn't get Death Strike, Assassin never gets Flight and the Winddancer doesn't get Invisibility.

That opens up a LOT of design space you would never have in a point based system where everybody can learn everything.

see
The best multiclassing for a d&d-like would be a point buy system with levels setting maximums and classes as example progressions for convenience.

like a more tactical d&d flavored mutants and masterminds, with example progressions and a larger variety of prebuilt abilities, designed to make mediocre generalists rather than being able to fill all roles simultaneously and expertly.

>IE: rather than build a party of twf rogues, or a party of somewhat adequate generalists, the group explicitly builds characters to fill different party roles.

Cool. Why jump through the hoops of doing that when you can just do so through classes, which are already set up to fill those niches?

I'm also going to go ahead and assume (probably wrongly but what the hell) that you are also thinking "class as character". For example, an entire party of two weapon using roguish characters could each have a different class and fill a different role; except if you have been browbeaten by 3.PF into thinking that only characters with "rogue" written on their sheets can be rougish.

>I've never had issues with the unspecialized characters in shadowrun 4e, and thats entirely point buy.

Which is the reason why I think pointbuy is superfluous in SR and most games where character roles are expected; the characters will specialize anyway, so you may as well give them classes.

Sooooo.... not multiclassing at all, but rather point buy with guidelines, gotcha

Why do you keep arguing about "multiclassing" if all you want is point buy then? Come to think of it, why are you even arguing about this in this thread?

>abilities which would be stupid broken if combined with other abilities don't have to worry about those other abilities because the players can't take them if they're in different classes.
ah. I see what you mean. I suppose that is a benefit you wouldn't have in a game with more flexible chargen. I'm not sure it's worth the cost in flexibility, but yes, you wouldn't include such abilities in a game where all of the options are attainable by all characters, they'd need to be designed in a way where they can be combined with other abilities without breaking things.

continued discussion from last thread. 4e multiclassing was criticized, 3e multiclassing was claimed as a great idea/first step with poor implementation, and as better than the 4e option, even in its flawed state. discussion continued over to new thread.

>Which is the reason why I think pointbuy is superfluous in SR and most games where character roles are expected; the characters will specialize anyway, so you may as well give them classes.
the pointbuy gives you the freedom to decide how you're going to fill whatever party role(s) you've chosen for yourself.

>the pointbuy gives you the freedom to decide how you're going to fill whatever party role(s) you've chosen for yourself.

You could still do that with classes. A class only has to make you fill the role, not predetermine how you do it. In SR, I'd probably handle it by having money act as your "point buy" pool to buy the tools of your trade with; spells, cyberstuff, decks, rigs/robots, etc, while class would be your passive benefit granter thing.

Or possibly set it up like Stike! where you pick role+class (although of course SR roles are pretty different from Strike! roles).

i suppose that could work too, sure. but your example has money serving as the pb.

or (talking all pb) you could put points into the roles, determining how good you are at each and how specialized you are, with maximums and minimums by level; you'd be free to choose whichever abilities you had the points for, but the power of those abilities would be keyed off of your role specialization.

striker stats with defender abilities? the defender abilities would be much less effective for you, but you'd have them.

Having mechanical "classes" be roles, and simply provide your stats, and having the abilities be per-level point-buy, with thematic classes as example power packages, could be good, but you would need to have abilities that play nice with eachother and can't "combo off" a la mtg.

I don't see what's wrong with 4e's class system as it is

You could apply the "combo" abilities to the "non-combo" part of the character (basically, the class) and have the rest be handled by feats/equipment/whatever.

This. A lot of the suggestions feel bizarrely over complicated for no real gain.

some people feel its insufficiently flexible.

true, or have the combo abilities in different "trees" of combo abilities, with non-combo capable abilities simple taking points from a general pool.

Really?

With hybridization, multiclass feats, power swap feats, and paragon paths and epic destinies like travelers' harlequin and eternal seeker, you can basically ignore class restrictions in 4e. It's a very fluid class system

Except, its not what happens in SR. Or you become super specialized or you are fucked.

Granted, MCing to any great degree is incredibly expensive, and many hybrids have the issue of class features being barred from cross class interaction

Well, yes, it's costly to do so, because the system is balanced around the classes working as themselves, so some class features are bullshit OP when combined with powers or features from another class

For example, imagine a 4e Ranger, but instead of having the hunter's quarry class feature, she has the Oath of Enmity class feature. That character would be the strongest 4e striker, far outstripping any other Ranger build

Or be a twin strike avenger

Would be far stronger than a twin-strike avenger

Access to all ranger powers is a big deal, the real strength of the ranger isn't just twin strike, it's every level of encounter and daily powers they get. Starting with jaws of the wolf and off-hand strike and culminating in ultimate confrontation

Not to mention saving an entire tier's worth of feats and a multiclass feat along the way

Alright folks, we're talked about what 4E did right and what it did wrong. Now let's consider how to fix it. Let's start by isolating some core mechanics and ideas that worked well and should be improved, but not removed.

>Power Sources (Arcane, Divine, Martial, possibly Psionic and Primal)
>Roles (Controller, Defender, Leader, Striker)
>Hit Points
>Healing Surges
>Encounter Powers
>Utility Powers
>Passive Defences
>Skills (or Backgrounds with the same function)
>Paragon Paths
>Epic Destinies

Let's also list out the mechanics that need heavy improvement and revision, or otherwise should be put as Optional Rules or removed entirely.
>Attributes (Stick with existing 6 and rebalance them or create a new attribute array?)
>Feats (replace with passive Utility powers instead?)
>Magic Items (make optional and not part of core monster math?)
>At-Will Powers (incorporate into Roles or Power Source directly?)
>Daily Powers (consider replacing with more Action Points or something else on a Daily resource schedule)

Also on the to-do list:
>Make Arcane, Divine, and Martial power sources really distinct and different from each other
>Make the benefits of Role more uniform (all Strikers get X, or a choice of Y or Z, or something like that)
>Improve the skill system so that being Trained matters more in the mid to high levels
>Several things I've likely forgotten

>Make the roles more uniform
That's exactly what you don't want. Every Striker has a distinct identity and that's a good thing, not a bad one. There's already something like it,
>Every Defender draws enemy fire and protects important targets
>Every Controler has AoE and Debuff skills good for dealing with lots of weak enemies or softening up big ones
>Every Leader has some form of healing, and provides skills making certain enemies easier to kill, guiding the flow of battle
>Every Striker deals a lot of damage, being the main tool by which enemies fall.

Now, multiple things I read on 4e Party Composition have basically said the whole thing is Striker focused, which we may or may not want to keep.

>Every Striker has a distinct identity and that's a good thing, not a bad one.
I know, that's why I'm proposing something like the following:

Striker: At 1st level choose one of the following:
- Hunter: Once per turn you may spend a minor action to mark the enemy nearest you. Your first successful attack against that enemy each round deals +1d6 damage. This increases to 2d6 at level 11 and 3d6 at level 21.
- Opportunist: The first time each round you hit an enemy that's granting you combat advantage, you deal +2d6 damage to them. You can only deal this bonus damage once per round and only while wielding a light blade, crossbow, or sling.
- Avenger: Once per encounter you may spend a minor action to mark an enemy you can see. While your enemy is adjacent to you, and no other enemies are adjacent to you, you may roll twice to attack them and take the higher result. When your marked enemy is reduced to 0 hit points you regain the use of this ability.

There could be a few others in that list as well, maybe as much as five or six.

I don't like the way this sounds. A role had limited mechanical benefit in the game beyond simple language that would help a player understand how a class might work. Having it be what determines key features, which should instead be used to give each class a distinct mechanical identity to separate them from others of the same role (which was one of the better design aspects in how 4e classes were made) flies in the face of the whole point to a class. It's not good design at that point. Just giving options for options sake. Slapping all Striker features to the Striker role and giving a player free reign of them invalidates all Striker classes out the gate. There's no point to classes at that stage.

4e shined for how each class (despite what some might believe) was distinct from one another at their core. Fighters and Paladins were both defenders, but they used their marks differently, and had different areas of expertise in how they protected their allies. This made them both feel very distinct in actual play despite seeming similar on a surface level. It's one of the more memorable aspects of the game itself. If you're going to refurbish it, keeping that as one of the key things to maintain should be a higher priority.

>I'm not sure it's worth the cost in flexibility

That's fair.

My answer depends on whether you ask me as a player, as a GM or as a game designer.

As a player, I want the ability to make the character I'm imagining. When a class exactly supports my concept, that is best. Second-best is a point-buy system that can be manipulated into almost supporting my concept. Worst is a point-buy system that cannot support my concept. (E.g. I cannot make a D&D wizard in Gurps, or a competent not-spell-using melee warrior in 3E D&D)

As a DM, I appreciate characters being predictable in power, and I appreciate not having to make a lot of decisions about character building.

As a game designer, I definitely prefer the vastly more open design space in strict class based games.

What this actually means is that, as a player and a DM, I prefer 4E (for two different reasons, mind) but as a game designer I prefer OSR editions because you can go as off the ledge as you want and just declare that This Class Doesn't Multiclass if you're in one of the editions that let some classes multiclass.

> Only 10 or so. More levels And Epic Paths in expansion.

Reducing class variety seems like the last thing you want.

And I'd rather not every power source have a single unifying gimmick. They work better as broad sets of mechanics and themes, IMO.

This. While each power source might seem like a good way to umbrella everything, they work better for a qualifier than anything else. The class itself should do the most of the heavy lifting for identity.

>These enemies that are literally intended as unimportant trash to be swept away should require the DM to track their HP values rather than just dying when they get hit.

I'm sorry that 3.x trained your hindbrain to expect game rules to pretend to simulate reality.

>The linked post

Holy fuck there are people that damn stupid?

>There's no point to classes at that stage.
Well, the Ranger and the Rogue aren't terribly different really. They're Martial Strikers with the exact same Hit Points and Surges, and with very similar preferred attributes, armour, weapons, and defence bonuses. Rangers get a bonus feat and bonuses to attack the closest enemy to them, Rogues get dagger and shuriken bonuses and bonuses to attack enemies that grant them combat advantage. The description of both classes says that they're all about hit-and-run tactics, darting in for massive damage before moving on. They both tend towards melee or ranged weaponry.

The only major difference between the two is the skill selection and flavour, that the Ranger travels the rural, wild parts of the world whereas the Rogue navigates urban environments rife with traps. Adjust their skill selection a little and they're practically the same.

Yeah, there's a lot to do in an overhaul of 4E. Personally I'd reduce the levels somewhat, 1-5 for your role's Encounter Powers, 6-10 for your Paragon Path, 11-15 for your Epic Destiny, and maybe something beyond that for demi-gods and the like.

>Right
Looked at the format and was willing to change it.
Keyed skills to level instead of having point buy bullshit.
Clearly defined class niches and worked within them.

>Wrong
Nixed a lot of potential for depth in dungeon crawling, either by having an ambivalent relationship with "utility" content or as a consequence of unrelated shifts. Balance-per-fight instead of balance-per-day makes snowballing too dangerous to be a staple of gameplay. Many turns per fight (admittedly more a problem on release) reduces the value of getting a free turn via initiative. New healing structure made limping through the dungeon unlikely. That sort of thing.

Needlessly kept a lot of focus on attrition. Initially a lot of the daily and encounterly shit was a beefing up of at will shit, healing was more plentiful than it had been, and again (on release) you would get through fewer combats per session. Tracking this shit felt like a waste when plausible options to totally nix attrition (rituals, stance systems) had already existed somewhere in 3x's line.

Within that attrition system 4e limited options needlessly. If you have three dailies you can do each one once, but not one thrice. With the ambivalent relationship to utilities and the dungeon crawl as a game, this feels unnecessary.

Applied excessive standardization to the attritional system. There were no classes (on release) that were fully at will or fully encounterly or fully daily. There wasn't even any leaning one way or another. If the resource management aspect had been excellent that might have worked, but without it? It simply came across as an annoyance for many of us.

Kept a bunch of unnecessary distinctions and cruft. Do you need both abilities and skills? Do you need both powers and feats? Do you need everyone to get a daily at level x and an at will at level y, when you could just have your pick off a list? Do you need so many levels, or so many small variations on similar powers?

Ranger is focused on multiattacks, the rogue is more about big single attacks and. Rogue is also a lot more mobile in practice, but needs to rely on all kind of tricks to get that CA.

Although they are pretty close, they are different enough imo.

I am finding it very hard to parse your overall point.

>Change
Already working on a ground up overhaul that no longer bears a close resemblance to the source.

Almost fully at will. Casting times, a stance system (you can't do two big things or things with duration simultaneously), and interruptions work fairly well as non-attrition limiters.

Mages feel a little vancian without precisely working that way. They'll have stances that take a while to cast and can be interrupted by damage. Some stances allow rituals to be done quickly, which ends the stance. So that's sort of prep-lite and inspired by reserve feats. Also lets me limit shit like charm or sleep, which you should really only have one chance to cast in a fight.

Healing feels a little 4e-ish on hp. In combat healing is weak and some martials have self-healing options. Ritual healing does all of someone's hp, but if the whole party needs it you're likely to trigger random encounters. Some hp works a little like vile damage or Fallout 4 irradiation so you'll eventually need to get to a temple or some shit. Lets you have proper limping through in ways other versions haven't done.

Then there's all kinds of little shit changed. There's an active defense system and AoOs run on the same pool of actions. Everybody's a little sticky, since AoOs interrupt and end what provoked them. Initiative gets you more like 1.5 turns and robs your foes of reactions in combat that can easily end in 2 hits.

I've been working on this shit for ages, but might finally have folks to test it on/with.

Responding to the OP's questions. Seeing 4e as not different enough from prior editions (I know I may be alone in that). Blaming 4e's backlash largely on its poor pacing on release. Shooting the shit. Do I need an overall point?

I guess not. I was just finding it somewhat tricky to figure out from your statements what you actually wanted the game to be.

I was really hopeful when they said it'd be like a combination of SW Saga and Bo9S. I was hoping for something light and streamlined numerically, which I got. But I was really disappointed by the seemingly arbitrary limits placed on AEDU on release. And the combat ran long on release. And most of my favorite character concepts sort of got nixed and stayed nixed near as I can tell.

I played it later when updates and supplements had somewhat improved things, but I had some pretty specific hopes and now I'm stuck with an itch that's probably just not gonna get scratched ever.

>Ranger is focused on multiattacks, the rogue is more about big single attacks
You're absolutely right. Still, what would happen if you let them share the same pool of powers? Characters would gravitate more towards multi-attacks, or big single target damage, or a mix of both.

Personally I think it would be nice if the Ranger filled the conceptual role of Marital Controller. We got to see what that would look like with the Hunter from Heroes of the Forgotten Kingdoms, I really like the idea of a Green-Arrow-style Ranger shooting arrows that explode into nets and create smoke-clouds, maybe even dragging people around the battlefield.

I'd prefer to keep class powers distinct.

I think if you want common power pools, liking utilities to power sources is a better idea.

Yep, my idea is to link Utility Powers to power sources, as well as including a General category that covers things like skill powers, passive bonuses associated with feats, and the like.

>What did it do right?
For the types of games that I run, everything. However, if I step back, and look at it from an outside perspective, the individual benefits are
>The difference between high optimization and mid optimization is not so vast as to make the latter useless.
>Combat is internally fun enough to be a tabletop wargame in its own right.
>Monsters were built off of a fundementally different chassis from PC's, which was easily reskinable, and ULTRA easy to build challenging but not impossible combat encounters.
>Focused on what it was good at (combat) and didn't water down its strengths trying to do everything.

>What did it do wrong?
Personally, I feel like it did nothing wrong, but if I step outside of the type of game I run
>changed too much from an established formula that had dominated the market for a decade, and defined the genre for the majority of living players (despite how many people claim to be different classes played much more similarly to each other than they had in 3e
>Character building was no longer an internally satisfying deck-building game.
>While combat was fun, it took a loooong time compared to the previous edition, especially before you got used to the rules.

>What would you change?
Personally, almost nothing, though I do have a slew of house-rules. Feats should be fun, and not necessary. Equipment should be fun, and not necessary. Essentials can suck a bag of donkey dicks. There is really no advantage to still using the 6 stats, when in 4e there are REALLY just 3 (Power, Finesse, Spirit) with two sub-stat expressions of each primary stat.

If you let them share the same pool of powers, you'd have rogues with ranger powers running around doing even more ludicrous damage than rangers already do

Why do so many 4e houserule suggestions end up buffing ranger powers? 4e Rangers do not need help, they're already the kings of nova and the kings of DPR

Rogue is considerably more of a team player

That sounds suspiciously like RoC's Legend.

No, not really, point buy systems are extremely hard to balance for without forcing structure on the game like Shadowrun does, and even then Shadowrun has heinous balance issues. Games like GURPS and M&M simply label problem abilities and extras instead of balancing them.

Tangentially related, but does anyone recall the name of the site that has the compendium? I can't recall what it was called and I don't know if it's just been taken down or what.

funin.space/

If the problem is that Ranger powers are too strong, maybe those powers need to get nerfed. If one 4E class/build combo is overpowered it should be weakened so it's more in line with the other options.

All the more reason to make the Ranger more of a controller like the Hunter, thus more of a team player.

Why does ranger need to be more of a team player?

You're actively trying to make rangers less defined as their own class, why?

I don't mind certain power sources having a little bit of one role in them though desu.

Like how all Arcane classes had a little controller in them, most divine classes had a bit of Leader in them, etc.

But that kind of works as a broad general slant to their powers, rather than some of the examples given in this thread and the last one of having a single ubiquitous mechanical gimmick for each.

The line could have used more good pre-built adventures. The game's still D&D, so even though the math works better than any other edition, it still requires a lot of heavy lifting on the DM's part, especially on account of the big, setpiece-driven encounter design. And I've got a job and shit to worry about.

If they had put out a Heroic series of Eberron adventures, I'd have been over the moon.

Dungeon magazine had a fair few Eberron Adventures in it

But sadly, most 4e pre-built adventures kind of sucked. They were good for one thing though, and that was giving DMs pre-made places to fight

4e thrives on having interesting battle locales, moreso than other editions thanks to the tactical slant of the game, but making interesting places to fight is much harder than balancing encounters to the party

4e was a great system that just shouldn't have called itself D&D.

>What did it do right?
Balance to roles, which gave classes the freedom to fill those roles in unique ways.

Crunchy as hell, with a shitload of interesting feats to pimp out characters without sacrificing party utility.

>What did it do wrong?
Build around software tools to gen chars, then fucked that up multiple times, only one of which was a murder-suicide.

Emphasis on and nerfing to tournament play. Constantly fucking with shit just because some asshats found an exploit.

Magic items were shit. Just... just shit.

>What would you change?
Give each power crunchily different bits, like how roles work. Example: martial characters get extra at-wills that explode in damage when they roll high or whiff if they roll low, primal characters all have short-duration, transformation-themed burst powers, etc.

Magic items are limited to one or two per character, since the game is balanced without them anyway, and all of them are intelligent with their own desires. Upon activation, the GM decides if the player's satisfied them enough to cooperate. Maybe a persuasion/intimidation roll too.

I like 4e magic items. they were supplemental instead of being the main focus

There was actually an official article where a dev begged DMs to come up with their own items that weren't shit, because all the ones they published had to be tournament-legal and therefore almost completely ineffective.

Boots of Spider Climbing is the only magic item I can ever remember using.

>mount

You know you would

Well I liked the low powered stuff. I was never a fan of things like belts of giant strength or staves full of whole spell lists

You can do both - powers that come from your Power Source and powers that come from your Role.

A Divine character would have access to Channel Divinity powers, which would lean towards Leader-type effects (as they do now). An Arcane character could get a passive that applies forced movement and maybe combat advantage to their attacks in certain situations, and thus a Controller role. Martial characters have better weapons and perhaps an accuracy and/or damage boost, so they make good Strikers. Perhaps Primal characters could have more hit points and healing surges, perhaps good AC as well, so they tend towards Defenders. Of course they'd get maybe a choice of (for example) 10 encounter powers from your source, four for their primary 'role' and two for each of the others, and they'd have a short list of Utility powers and options to choose from.

Roles naturally give their own passive abilities and a list of encounter and utility powers you can choose from. They'd key off of whatever your Basic Attack was, so if you have (for example) an Arcane at-will power that lets you blast fire at your enemies, you'd use the defence it targets, the damage die of the at-will, and the type of damage as the 'base' of the attack.

Ehh, I don't like that degree of overlap. I think it'd just result in things feeling as samey as 4e is often accused of being (even though it isn't).

Well, some people like the idea of Power Sources leaning towards one Role or another. You, and probably others, don't like the idea as much. I could take it or leave it, personally I like the idea of Power Sources determining how you manage your encounter powers.

>Martial characters treat them as Reliable
>Divine characters get a secondary list of Channel Divinity powers. They can use one per encounter, and they're mainly big, immediately, temporary boosts to allies and debuffs/damage to enemies.
>Arcane characters know twice as many (or more) but are limited in how many they can use per encounter
>Psionic characters get Power Points and can spend them to enhance or alter their powers, maybe make their at-will attacks.
>Primal characters get a secondary list of Forms they can enter (like Rage and Wild shape). They can use one per encounter, and they're mainly low-powered but long-term buffs to yourself and allies

This gives the different Power Sources something cool about them, while still having them draw from the same general list of powers based on their Role.

I really don't like every power source having a single mechanical gimmick.

I think power-source feats, paragon paths and epic destinies provide plenty of cool stuff

Like the reincarnate champion ED, or the white lotus arcane feats, or the bard power-source PPs, or the martial style feats

Bump for any more? For all the trolling back and forth, there's also been a lot of interesting ideas shared.