Class Design Theory

[Nightgaun7]

I remember seeing some stuff a while back about how the different classes were designed over 4E's life cycle, particularly relating to how you were expected to choose your stats, but I haven't been able to remember the keywords to search for to turn any of those old threads or pages up.

So, discuss 4E class design here, and hopefully it will jog my memory.

[MwaO]

I remember seeing some stuff a while back about how the different classes were designed over 4E's life cycle, particularly relating to how you were expected to choose your stats, but I haven't been able to remember the keywords to search for to turn any of those old threads or pages up.

So, discuss 4E class design here, and hopefully it will jog my memory.

PHB1:
I think the initial idea was that you'd have a primary stat, a secondary stat, and a tertiary stat. You'd use the 16/14/14/13 array and end up with a 16/14/14 in 3 different defenses. You'd likely boost 2 of them for 18/16/14. Then over the course of 30 levels, you'd get a total of +20 via leveling and Epic Destiny. You'd spend 10 on your primary for a 28, say 6 on your secondary for a 22, and 4 on the last for an 18. The easiest example of this might be the PHB Fighter with a Heavy Blade - wants to invest in Dexterity for the critical hit feat, Wisdom for Combat Superiority. A problem showed up in that people realized that they weren't likely getting all the way to Epic - and then focusing on two stats was a better deal usually.

People quickly realized however that unless you think you're actually getting to Epic, there's a lot more usefulness in investing in just 2 stats.

PHB2:
Hmm, maybe we should design around the strong secondary stat or avoidance of. Sorcerer is probably the best example out there of a Strong Secondary stat - contributes to defense & damage! Barbarian is the opposite. Not only are you usually not expected to invest heavily in the stat, there's an option that helps compensate for having say a 28/20/20 in Str/Secondary Stat/Dexterity. Except they didn't spell out Barbarian enough and people made really dumb Barbarian builds having a 12 Dexterity that they didn't then level.

PHB3:
People seem to be giving up on 4e and we're getting two big complaints - Wizards can't win non-combat with spells(okay, let's ignore that for now...), there isn't a 'simple spamming option'. Let's make some spam-only options in Psionics. That's not actually what people were really asking for and we stupidly made options that are too good to spam, but oh well.

Essentials:
Um, people aren't buying books. Hey, Wizards can now win non-combat with the Arcana skill! Fighters can make a single attack per round only that does big damage! And why don't we use an obvious place to fix the math("Hey, everyone now gets a +1 feat bonus to attack/defenses at levels 5/15/25. Done!") as a design space to make 4e more boring("There are these great feats that everyone, and we mean everyone should take at least 2 of by 11th level or you're idiots.")

[Kurald Galain]

PHB1 has "V classes" and "A classes". An example of a V class is the warlock: it either uses con and int, or cha and int (so it has a choice of primary stat). An A class is the rogue: it uses either dex and cha, or dex and str (so it has a choice of secondary stat).

Players quickly realized that a V class basically cannot use half of its printed powers (because a con-build cannot effectively use cha-powers and vice versa), which proved an impopular design decision. Player outcry made them stop printing V classes.

Part of PHB1 design is that you won't have a good melee basic attack unless your strength is decent (for example, the str-rogue can make OAs and the cha-rogue basically cannot). This one took a while longer, but again there was player outcry that this was massively unfair, so eventually they started printing feats and/or at-wills that allowed all classes to use their best stat for OAs. By that time, there were also traits that let you use your best stat for hit points (instead of con), and some options to use your best stat for certain skills (instead of the regular stat), and there were paragon paths and themes that had attack powers of "your best stat vs AC".

Basically, early 4E tries to enforce a difference between the six ability scores, and late 4E lets you do almost everything with your best stat, regardless of which stat that is. Turns out that despite the vocal minority decrying the former, most players didn't actually like the latter.

[Nightgaun7]

PHB1 has "V classes" and "A classes". An example of a V class is the warlock: it either uses con and int, or cha and int (so it has a choice of primary stat). An A class is the rogue: it uses either dex and cha, or dex and str (so it has a choice of secondary stat).

A and V builds, that's what it was! Now I remember there being a couple of other letters - X, Y, L, and maybe one or two others. I think S class was a thing but I never figured out what that was about.

Did the community ever reach a consensus? I recall there being some pretty heated arguments on the subject on the WotC boards.

[masteraleph]

X class would be one with two primaries and two secondaries. An example could be (when you include all of the sources) the Cleric- it can have either STR or WIS as primaries, and has powers with riders requiring CON and CHA.

S-Class doesn't refer to a kind of class based on ability scores- it's based on Japanese video game talk, and refers to the absolute best tier of builds. For example, Warlords might be A-class Leaders (in the sense that they're much better than many other leader classes, rather than based on their ability scores), and Artificers C class (in the sense that they're generally weaker than Warlords, Bards, etc.), but Chordswitch or Flameswitch are S-class- they're the absolute top of optimization. Or Fighters might be A-class defenders, and Battleminds B-class (very nice at higher levels, but Lightning Rush doesn't come online until level 7), but a Swordmage|Warlock/Sigil Carver or any number of Battlemind|Paladin builds would be S-Class. A regular Wizard would be A-class, while a Tiefling Enchanter or Headspin or Illusion of Hope would be A-class, etc. Basically, it's a way of saying "A++++".

[Nightgaun7]

S-Class doesn't refer to a kind of class based on ability scores- it's based on Japanese video game talk, and refers to the absolute best tier of builds.

I thought S-class, insofar as it referred to stats, was supposed to be one that had 3+ possible primary stats, or were ones that cared about one stat only and didn't have a need for a secondary? Not sure what classes actually fulfilled that, but that's what I recall.

L was classes with one primary stat and could use 3+ secondaries, iirc the Warlord was the best example of this.

Don't remember what Y was.

[Kurald Galain]

A and V builds, that's what it was! Now I remember there being a couple of other letters - X, Y, L, and maybe one or two others. I think S class was a thing but I never figured out what that was about.

I've never heard of any of these others.

It seems to me that A and V are clearly intended by the designers and present in the rulebooks, whereas these others are the result of over-analysis by forum users and not an intentional pattern.

[Nightgaun7]

I've never heard of any of these others.

It seems to me that A and V are clearly intended by the designers and present in the rulebooks, whereas these others are the result of over-analysis by forum users and not an intentional pattern.

Agreed, but even so it's interesting to see how things evolved over time - for example, there was a clear tendency, imo, to make the specific stats matter less. For example, all the ways to get around needing STR for MBAs, or the Warlord and Wizard getting more and more secondary stat options, sometimes eclipsing whatever was supposed to be the main stat.

[Kurald Galain]

Agreed, but even so it's interesting to see how things evolved over time - for example, there was a clear tendency, imo, to make the specific stats matter less. For example, all the ways to get around needing STR for MBAs, or the Warlord and Wizard getting more and more secondary stat options, sometimes eclipsing whatever was supposed to be the main stat.

Yes. It is noteworthy that over the lifespan of 4E, the lead designer got fired by WOTC several times (as in, once per year). Of course, his replacement would have a new design philosophy, hence the very different focus between the early books and the late books.

[TheWayofPie]

Fired every year? That's depressing but not surprising in the slightest.

Going back to MBAs: It's kind of funny how there was never really a ranged training feat, but then again the only class I imagine wanting different stats for MBAs would be certain stripes Fighter and Ranger.

[Excession]

Fired every year? That's depressing but not surprising in the slightest.

Going back to MBAs: It's kind of funny how there was never really a ranged training feat, but then again the only class I imagine wanting different stats for MBAs would be certain stripes Fighter and Ranger.

Heavy Thrown weapons give the Fighter and Ranger a Str based ranged option, so they always had the option.

[Waddacku]

There's also Serene Archery for... Not sure who for, archer Paladins and Clerics? But it exists, anyway.

[Nightgaun7]

So what's the preferred style?

[darkbard]

Generally speaking, making your teammates better at killing things quickly: boosting initiative, enabling attacks, pumping damage through boosts, things of that sort...

[Nightgaun7]

Generally speaking, making your teammates better at killing things quickly: boosting initiative, enabling attacks, pumping damage through boosts, things of that sort...

That has nothing to do with whether a class is A, V, or any other flavor of alphabet soup.

[darkbard]

You're absolutely right. I thought I was responding in the "Channel Divinity Optimization" thread and comments about Pacifist Clerics therein. My bad.

[Kurald Galain]

So what's the preferred style?

The preferred style for most players is probably A. It become obvious pretty quickly that players didn't like V classes much, and after PHB1 they stopped printing them.

The preferred style for the designers in the last couple books is closer to "use your best ability score for everything" but that defeats the point of having ability scores in the first place.

[NomGarret]

There are a few ways traditional V class setups can work, but it's a lot more work from a design perspective, so I'm not surprised it was dropped. You can add a lot more powers than other classes get, so that either primary score has almost a full class' worth of powers to choose from. You can drop any secondary stat reference altogether, and just have all the rider effects key off static numbers or the like. Or you can do what I use for Star Pact warlocks and replace the secondary stat with whichever of the two primary stats isn't otherwise used in that power.

Now the last two options no longer make them truly V classes, but the intent is to make them playable, not to adhere to a letter system.

[MwaO]

There are a few ways traditional V class setups can work, but it's a lot more work from a design perspective, so I'm not surprised it was dropped. You can add a lot more powers than other classes get, so that either primary score has almost a full class' worth of powers to choose from. You can drop any secondary stat reference altogether, and just have all the rider effects key off static numbers or the like. Or you can do what I use for Star Pact warlocks and replace the secondary stat with whichever of the two primary stats isn't otherwise used in that power.

Well, you could have a 'primary' stat that isn't one and 'secondary' stats that effectively are primary.

As an example, a Warlock could use Intelligence to hit, but have damage of a power be based on Con or Cha depending on which one it is. That way, a Warlock could basically use any power, but at the same time, they'd have some strong incentive to use either Con or Cha. Especially with Riders.

Issue would be to avoid making Warlocks massive MC'ers of Wizards, but that could be done by making Wizard powers 'sucky', then giving Wizards class features to enhance them. So a Wizard might have a Ranged power 10 that as part of a class feature, could make it area burst 1.

[masteraleph]

Well, you could have a 'primary' stat that isn't one and 'secondary' stats that effectively are primary.

As an example, a Warlock could use Intelligence to hit, but have damage of a power be based on Con or Cha depending on which one it is. That way, a Warlock could basically use any power, but at the same time, they'd have some strong incentive to use either Con or Cha. Especially with Riders.

Issue would be to avoid making Warlocks massive MC'ers of Wizards, but that could be done by making Wizard powers 'sucky', then giving Wizards class features to enhance them. So a Wizard might have a Ranged power 10 that as part of a class feature, could make it area burst 1.

Psion arguably operates that way- hit is INT based, but most of the powerful stuff has either CHA or WIS riders. But that's really just an A class all over again, albeit with really strong secondaries.

[Yakk]

Design wise, the nearly complete lack of "controller" features on Wizards is disappointing.

Defenders have their Mark. Strikers have their damage rider. Leaders have their 2x encounter heal.

Controllers really don't have a controlling class feature. They are controllers due to their powers.

(Barbarian, arguably, is another class built like that).

Of course, striker damage riders tend to get overshadowed by optimization.

---

The reason why A and V where identified is that it turns out that hitting is important to doing stuff in 4e.

In my view, the designers of 4e dropped the ball with that. They underestimated how important it was to hit in game design.

This means that there are multiple easy traps to fall into; not having a 16-20 attack stat at level 1, and not investing a stat-bump in your attack stat every level.

These traps are not all that apparent at low levels, and failure to avoid them gets cumulatively worse, and the failure is irreversible on a given character.

The exceptions to this rule are the lazy-lord and similar builds, where they find a way to contribute without hitting with attacks. But they are serious exceptions.

---

Half the time when I throw together a new 4e class I bypass that mechanism entirely.

The class powers Hit line is special: it is Level+3 vs AC, and enhancement and feat bonuses don't apply (or are capped at +1 each) by special (class specific or keyword-based) rules. (weapon proficiency still fully applies)

Then I have a lot more play to mess around with how stats work for a character for that class. The class, in effect, becomes a self-lazy-lord, where otherwise "secondary" features of the stats determine what mix of stats the character wants.

It is a kind of "revamp 4e design flaws by class design" approach.

---

But that is a tangent.

[Kurald Galain]

Controllers really don't have a controlling class feature. They are controllers due to their powers.

I don't see any problem in that. There's no reason for all roles to work in exactly the same way.

In my view, the designers of 4e dropped the ball with that. They underestimated how important it was to hit in game design.

Or rather, to many classes, the benefits of any ability score other than your primary are too small.

In a game with six main stats, it is not good design that the obvious best allocation for many builds is entirely dumping three of them.

[ThePurple]

I don't see any problem in that. There's no reason for all roles to work in exactly the same way.

I actually agree with Yakk where this is concerned. The lack of truly "controller-y" class features means that the controller classes are all defined *purely* by their power selection, which is why multiclassing/hybriding a controller class can make you just as good at controlling as a full member of that class (e.g. Resourceful Magician).

Of course, part of this is that the controller role itself is confused. It's both the AoE damage role *and* the control role (compare this to the monster roles where Controller and Artillery are explicitly different roles *for good reason*). Imo, most of this is due to the wizards' sacred cow status as being omni-functional (i.e. "wizards should be able to do *anything* with the right spell") which lead them to focus too heavily on recreating existing spells rather than trying to build a truly functional class via a solid set of features.

My own interpretation of the controller role provides them with the unique ability to force rerolls on saving throws ("On hit: The target must reroll its next successful saving throw before the end of your next turn.") or apply weaker versions of a condition if the effect is ended early (e.g. "If a stunned or dominated target ends the effect early, the target is dazed until EoNT"), such that Controllers become defined by their ability to augment save ends effects (and, specifically, make stuff stick to hard targets).

[Kurald Galain]

I actually agree with Yakk where this is concerned. The lack of truly "controller-y" class features means that the controller classes are all defined *purely* by their power selection, which is why multiclassing/hybriding a controller class can make you just as good at controlling as a full member of that class (e.g. Resourceful Magician).

Well, for multiclassing that clearly won't be the case (because you can only get one power of each type, and at an excessive feat cost). For hybriding it could be - but again, why is that a problem? Hybrids of other roles can also made to be just as good as the full class.

And, for that matter, quite a number hybrid combos are just plain bad. It seems to me that on the topic of class design, hybrids need a makeover anyway. If controller hybrids are too good, that's a problem with the hybrid rules, not with controllers in general.

[MwaO]

Well, for multiclassing that clearly won't be the case (because you can only get one power of each type, and at an excessive feat cost). For hybriding it could be - but again, why is that a problem? Hybrids of other roles can also made to be just as good as the full class.

And, for that matter, quite a number hybrid combos are just plain bad. It seems to me that on the topic of class design, hybrids need a makeover anyway. If controller hybrids are too good, that's a problem with the hybrid rules, not with controllers in general.

Most hybrid combos are good. The vast majority in fact. You have twice the options usually for a small sacrifice. The problem usually isn't the hybrid combo, the problem usually is the player. They really just need a small fix to make sure players don't screw up AC - give Leather Armor proficiency to any PC where either class has Light Armor and give Chain where either class has Heavy Armor. And then if both classes have a bonus knowledge skill, the PC gets one bonus knowledge skill.

After that, just don't double up on the same NAD. Defender|Controllers tend to have some issues in play, but beyond that, almost everything is playable and good.

I'm happy to take on 'problematic examples' that don't double-up a defense.

[Nightgaun7]

Funny how satisfying hybrids can be (although a lot of the time it's more frustration than it's worth) versus how pointless multiclassing feels.

[MwaO]

Funny how satisfying hybrids can be (although a lot of the time it's more frustration than it's worth) versus how pointless multiclassing feels.

Hybrids & Multiclassing are about option selection. If you're great, they're great. If you're average, they'll likely be okay. If you suck, you can suck horribly. But in order to go there, you have to ignore all the warning signs.

This is basically my go-to page for picking powers. Not to say I always do, but if I want to build a multi-class or hybrid build, this is the starting point.
http://www.enworld.org/forum/showthr...-other-classes

[Kurald Galain]

Funny how satisfying hybrids can be (although a lot of the time it's more frustration than it's worth) versus how pointless multiclassing feels.

Yes. The fact that a skilled optimizer needs a reference page on a forum in order to build a good hybrid, pretty much underlines my point that in terms of class design, hybrids need a makeover.

Multiclassing was obviously an option the design team didn't really want but was forced to include for legacy reasons. I don't think I've ever seen a character multiclassed beyond the "encounter power swap" feat, and even that was rare. Taking just the first feat ("sneak attack 1/enc") never felt like actual multiclassing.

[Nightgaun7]

Hybrids & Multiclassing are about option selection. If you're great, they're great. If you're average, they'll likely be okay. If you suck, you can suck horribly. But in order to go there, you have to ignore all the warning signs.

It's not that I can't do it, it's that I don't like to. It just doesn't really feel right. One of my very favorite character concepts from 4E is the Paladin/Warlock hybrid, but in practice it's a finicky mess to build and play, even with all the support it got.

Multiclassing was obviously an option the design team didn't really want but was forced to include for legacy reasons. I don't think I've ever seen a character multiclassed beyond the "encounter power swap" feat, and even that was rare. Taking just the first feat ("sneak attack 1/enc") never felt like actual multiclassing.

Honestly, "Hybrid" should just be renamed "Multiclass" because multiclassing, isn't.

[MwaO]

Yes. The fact that a skilled optimizer needs a reference page on a forum in order to build a good hybrid, pretty much underlines my point that in terms of class design, hybrids need a makeover.

I don't need the reference page to build a good hybrid. I need the reference page to easily make hybrids better than base classes when optimizing(and know most of that page by heart because I wrote it up). Because even Fighters, Wizards, and Warlords run out of 'great' powers or have dead levels. As an example, Warlord has probably the best Defender power at 1st level in Vengeance is Mine. So combine it with any Defender class and that's a better Defender than an actual Defender most of the time. Except for the one who MCs into Warlord, takes Novice Power, and power swaps for it.

The problematic hybrids are Defender|Controllers for the most part. And even most of their issues are solved by having a 19-20 AC at 1st level via the hybrid talent feat and picking up mostly control powers, whatever choices those might be.