Community Tiering for all 3.5 Base Classes

[Gnaeus]

And, I mean... if you're going to factor Arcane Disciple/Runestaffs/PrCs that heavily into the discussion, you have to extend all that to the Dread Necro and Warmage, who have all of the same advantages and merely start with inferior lists.

I think Dread Necro is very likely T2. Warmage is more likely T3. You can't do everything with arcane Disciple or runestaffs. It doesn't make up for having a 1 dimensional list. But DN and Beguiler have very strong base lists.

[Cosi]

It's about quality, not quantity. A lot of the utility of planar binding is that you get to pick up all these neat SLA's that are sometimes overleveled, and a lot of the other utility is that you can just generally get a pretty high end caster with all the caster-action doubling that implies. If you're only getting beef, or lower end casting, then that's significantly less interesting.

Using planar binding to get downtime utility casting is just not all that impressive. You can buy spells for money. Caster minions are certainly better than non-casters, but you can get a lot more minions off of charm and dominate shenanigans (particularly in concert with Diplomacy).

I suppose that element is accurate, though the fact that you specifically need to win the fight against this opponent and in this specific way is definitely a downside.

You kind of need to win all your fights against all your opponents, so I don't know that "you need to win" is a downside. Needing to do non-lethal damage is a little annoying, but it can be done after the fight is functionally ended by glitterdust or something.

I'm not really talking about wish looping though, or looping in general. Just that these high quality spells are, y'know, high quality. Shapechange is still great even with zodars as non-existent. Ice assassin, beyond its game breaking power, is also just a really good spell.

It's one of those "where do you draw the line" things. Ranking wish cheese is absurd. But ignoring planar binding is also kind of stupid. Best solution is probably to decide on a list of broken powers, and ignore those for the purposes of ranking to avoid having to figure out exactly how much power it's fair to squeeze out of simulacrum.

The entire focused list full casters like Beguiler and Dread Necromancer are indeed better in their area of specific focus than a Sorcerer who has focused exclusively on doing the same thing. But the Sorcerer doesn't have to focus on doing that one thing. Admittedly, the Sorcerer's limited spells known is a rather painful limitation, and requires careful consideration when selecting spells known, but the ability to select any spells from the Sorcerer/Wizard list is a useful one. But the Sorcerer can take the Planar Binding line, Shapechange, Wish, etc, without expending list-expansion resources.

That's a different argument. It's true that Sorcerers can have more versatility, but Beguilers catch up very easily and once they do, having more baseline spells is a better deal than having a marginal extra free choice or two. greater planar binding, polymorph any object, and eight other spells is probably better than greater planar binding, polymorph any object, and the third coolest eighth level spell.

Cleric: 1
Divine Metamagic counts for a lot, and is not remotely an obscure feat. Spell list is surprisingly limited for how large it is, but there are more than enough tricks on it to work around that.

If we are counting Divine Metamagic for the Cleric, we should also be counting Arcane Disciple for the Beguiler, at which point any question of versatility goes out the window. You can get anything from animate dead to wall of stone to teleport, and that's just in core.

Beguiler: 342
I don't see that a spell that you get outside of the level 6-16 range that the Tier System mainly considers is enough to save you from ultimately being a one-trick pony, and a lot is immune to your one trick.

The things that are immune to your trick are mindless creatures (which lose to silent image), high level Celestials with magic circle against evil auras (which you are probably not fighting), and non-mindless Undead (for which you have minions). Plus, it's trivially easy to expand your spell list for almost nothing.

[Pleh]

My personal view is that if you don't account for non-class options, you don't get a useful ranking. If the class on its own is Fighter Tier, but the class with the right feats is Wizard Tier, in practice it's Wizard Tier, or at least two different rankings. Saying it's Fighter Tier just results in people being sad when they try to play it next to a Fighter. You can tell that people do believe that, because people talk about DMM: Persist in the context of the Cleric, despite the fact that it requires three feats (and some items to do it more than once or twice a day).

Okay, but assuming non class options can produce a misleading list as well. People could see a class that is rated high because of its access to prcs and acfs, but without knowing why it was rated high, they play it straight out of the book and wonder why it doesn't live up to its tier.

We do need a line somewhere.

[Cosi]

Okay, but assuming non class options can produce a misleading list as well. People could see a class that is rated high because of its access to prcs and acfs, but without knowing why it was rated high, they play it straight out of the book and wonder why it doesn't live up to its tier.

Isn't that exactly what "Why Each Class Is In Its Tier" is supposed to do? If you just present a set of rankings, people are going to be able to misread it regardless of what went into it. If you don't include PrCs and ACFs, people might assume those don't have an effect. If you don't specify which options put a class in its tier, you'll end up with people assuming blaster Wizards are as good as battlefield controllers.

[AnachroNinja]

Okay, but assuming non class options can produce a misleading list as well. People could see a class that is rated high because of its access to prcs and acfs, but without knowing why it was rated high, they play it straight out of the book and wonder why it doesn't live up to its tier.

We do need a line somewhere.

Something worth noting is that Arcane Disciple is a feat, not an ACF or PRC. A single feat is very much so within the possible standard build options for Beguiler20. And has been noted, that single feat opens up a huge variety of options to the point where most Beguilers should end up picking it up, even if the domain chosen is likely to vary. It's very close to the no brainer status of druids taking Natural Spell, which somehow is fair to always assume in this kind of debate.

More to the point, I don't remember any arbitrary distinction for Tier2 that required you have a variety of game breakers at every level. Tier2 is the "Your build can be designed so that it breaks the game, but it will be in narrow or predictable areas." The reality is, if one Beguiler20 build breaks the game, even if that's the ONLY Beguiler build that breaks the game, or even the only one that exists, it is Tier2. That is because Tier2 is primarily a function of power. And honestly, if anyone thinks that a Beguiler who spends his 1-20 career charming, dominating, diploma congratulations, and similar, is not going to break the game in fundamentally the same way as a planar binding sorcerer... Well you haven't tried to run a game with an enthusiastic Beguiler lately.

An argument can be made that their ceiling is not quite as high as a Sorcerer, and I'd probably agree with that, in a directly competitive sense. But if we aren't counting any PRCs, ACFs, or similar, then no, the sorcerer is not going to pull any huge distance ahead. And if we are counting that stuff, the Beguiler wrecks the sorcerer. And you have to acknowledge that is because the base chassis of the Beguiler is pretty good.

Bard - 3
Battle Dancer - 5
Beguiler - 2
Binder - 3
Cleric - 1
Commoner - 6

[eggynack]

Using planar binding to get downtime utility casting is just not all that impressive. You can buy spells for money. Caster minions are certainly better than non-casters, but you can get a lot more minions off of charm and dominate shenanigans (particularly in concert with Diplomacy).

But money costs money and whatnot. Planar binding can pretty easily be free. And you can be gaining access to a lot of money here. After all, potential uses for planar binding include a nightmare for astral projection or a mirror mephit for simulacrum. There're a lot of other uses too, ones that are really strong without going over the obviously borked wish loop line. Also, haven't noted this yet, but this overall strategy is weak to dispelling in a way that bound minions are not.

You kind of need to win all your fights against all your opponents, so I don't know that "you need to win" is a downside. Needing to do non-lethal damage is a little annoying, but it can be done after the fight is functionally ended by glitterdust or something.

My point here is mostly just that you're increasing the difficulty of fights somewhat by pursuing this, and generally expending more daily resources. Binding has no such issue. A lot of its power is in how low cost a maneuver it is.

It's one of those "where do you draw the line" things. Ranking wish cheese is absurd. But ignoring planar binding is also kind of stupid. Best solution is probably to decide on a list of broken powers, and ignore those for the purposes of ranking to avoid having to figure out exactly how much power it's fair to squeeze out of simulacrum.

My argument is basically that if you assume some theoretical variable positioned line that exists in each game, then ice assassin can get you to that line, wherever it is, without going over. The value associated with that line approximation capacity is obviously table dependent, but I certainly don't think it's a non-object.

Edit:

It's very close to the no brainer status of druids taking Natural Spell, which somehow is fair to always assume in this kind of debate.

To be fair, while natural spell and DMM frequently see mention when discussing the tier of these classes, because they're amazing, neither class really needs these feats to be in their tier.

[Muggins]

Bard: Tier 3/4.
The bard's versatility means that whatever the situation, they always have something to contribute. Not because they're particularly potent on their own - although they definitely, unabashedly are - but because they simply make everyone around them better. Poor decision-making with spells and ACFs can lower their overall power, but it's difficult to conjure a situation in which they're entirely useless.

Battle Dancer: Tier 5.
An unarmed fighter who benefits from charisma, similar to a monk and wisdom. Except that, unlike the monk, they have no splatbook support and little versatility in their core chassis, making damage their only function. And, unfortunately, they end up at the bottom of the martially-inclined, damage-dealing totem pole. At least you're better than the NPC Warrior, right?

Beguiler: Tier 2/3/1.
The definition of Tier 2 is incredible power with low versatility, and I think the Beguiler fits this definition perfectly. They're a fixed-list caster with a bevy of illusion magic, enchantment magic, and old standbys like Haste, Dispel Magic, Freedom of Movement, True Seeing, Mind Blank, and Time Stop filling their list. They compare better to a sorcerer than a bard, and they become even stronger when they add to their lists via methods like rainbow servant, sand shaper, and domain access. Any concerns about mind-affecting immunity can be worked around by a canny player that knows what their doing. Meanwhile, a badly-played beguiler falls to the floor of Tier 3; held back not by the nature of the class, but by bad decision-making and low competence.

Binder: Tier 3/4.
The binder's versatility means that they're always the best - or, otherwise, very good - at whatever they're trying to do that day. Their baseline level of power and competence is pretty average, with a good chassis and some reasonable class features (notably pact augmentation and soul guardian), but it's the vestiges that make them truly shine. If the bard is the jack-of-all-trades, then the binder is a master-of-one; from combat, to skills, to rogueplay, to spells, to summoning. A player can choose bad vestiges and neglect to use his granted abilities, but they're only ever a few bindings away from doing better.

Cleric: Tier 1/X.
The cleric gets no class features, but that's fine; they get spells instead. They also get domains, but those are just extra spells with a passive or daily ability tacked on. Their solid chassis (medium BAB, good Fort and Will saves, a half-decent skill list, d8 HD) makes up for any perceived inferior choices between the cleric and wizard spell lists, both of which have spells capable of dominating any social or combat encounter. Select ACFs and feat choices (looking at you, Divine Metamagic) offer ways to either break the game further or weaken the class only slightly, since you're never locked into your spell choices for more than a day. I'd daresay that the only limit to this class' power is the competence of the player driving it.

Commoner: Tier 6.
Completely unnoteworthy, uninteresting, and worthless. No skills, no class features, and a bad chassis. Only receives recognition because of a poorly-worded Flaw from an esoteric splatbook which has the potential to break the game.

[Doctor Despair]

Isn't that exactly what "Why Each Class Is In Its Tier" is supposed to do? If you just present a set of rankings, people are going to be able to misread it regardless of what went into it. If you don't include PrCs and ACFs, people might assume those don't have an effect. If you don't specify which options put a class in its tier, you'll end up with people assuming blaster Wizards are as good as battlefield controllers.

Aren't we supposed to be ranking these classes on their own merit and not on the merit of PrCs? We are ranking base classes, after all.

[eggynack]

Aren't we supposed to be ranking these classes on their own merit and not on the merit of PrCs? We are ranking base classes, after all.

ACF's are still clear factors of power, especially the lower key ones that aren't essentially whole new classes, and while I agree that PrC's are probably outside the bounds of this, there's definitely an argument to be made that some classes get more out of PrC's than others. For example, barbarian, which has access to stuff like runescarred berserker and champion of Gwynharwyf, is going to do better in a PrC inclusive environment than a fighter which... doesn't have those things.

[Pleh]

Isn't that exactly what "Why Each Class Is In Its Tier" is supposed to do? If you just present a set of rankings, people are going to be able to misread it regardless of what went into it. If you don't include PrCs and ACFs, people might assume those don't have an effect. If you don't specify which options put a class in its tier, you'll end up with people assuming blaster Wizards are as good as battlefield controllers.

Yes and no. The "Why..." thread can explain the tier either way. If we assume vanilla base classes, the why thread would say that prcs can affect tier greatly. If we assume non class features, the why thread should explain where such things have been assumed.

But I think we want the tier list to best represent the average 20 level single class build with consideration for floors and ceilings. The tier list is supposed to tell you what the class itself has to offer so that you know how much help it needs.

Because most anything can be optimized if you allow multiclassing. Barbarian is probably a higher tier if you only consider it a 1 level dip for building other classes.

[Cosi]

But money costs money and whatnot. Planar binding can pretty easily be free. And you can be gaining access to a lot of money here. After all, potential uses for planar binding include a nightmare for astral projection or a mirror mephit for simulacrum. There're a lot of other uses too, ones that are really strong without going over the obviously borked wish loop line. Also, haven't noted this yet, but this overall strategy is weak to dispelling in a way that bound minions are not.

Back of the envelop calculation says wall of iron produces enough iron to afford a 9th level spell with change left over. That's less casts than binding an Efreet, but with money left over and free choice of spell.

My argument is basically that if you assume some theoretical variable positioned line that exists in each game, then ice assassin can get you to that line, wherever it is, without going over. The value associated with that line approximation capacity is obviously table dependent, but I certainly don't think it's a non-object.

But if we're talking about loops (even chained ones) the best aren't class based.

Aren't we supposed to be ranking these classes on their own merit and not on the merit of PrCs? We are ranking base classes, after all.

If you rank things in a way they are not used, you don't get good rankings. No one actually takes 20 levels of a base class with no feats, items, PrCs, dips, or ACFs. You could try to do what the old tier system did, and have a separate ranking for PrCs, but that causes immediate problems when you try to rank things like the Rainbow Servant that are heavily dependent on the class chassis. Rainbow Servant is nuts if you cast like a Warmage, decent (but outclassed) if you cast like a Wizard, and awful if you cast like a Sorcerer. If that information isn't baked into base class rankings, it's really hard to represent effectively.

[eggynack]

Back of the envelop calculation says wall of iron produces enough iron to afford a 9th level spell with change left over. That's less casts than binding an Efreet, but with money left over and free choice of spell.

It seems kinda hypocritical to be fully opposed to things that can loop infinitely but then be fine with exploding WBL to such an extent that can rather trivially get to any quantity of money you could plausibly desire. The parallel between wizard wish loops as compared to commoner wish loops, and wizard WBL breaking as opposed to commoner WBL breaking, seems like a clear one to me.

But if we're talking about loops (even chained ones) the best aren't class based.

I'm not really talking about loops though. I'm talking about just, y'know, casting ice assassin. For value. You get what's still a sort of caster singularity, just because you can approximate any other spell effect, just as they can approximate any spell effect of yours.

[BaronDoctor]

Bard: 342

Bards are Tier 3. Water is wet. They can contribute almost anywhere in a meaningful fashion (Inspire Courage, Skills, Spells), but they don't distort the game due to their ability in any one area. Usually. Words of Creation DFI (Sonic) bards start throwing around reliable bonus damage. Jack of All Trades + Bardic Knack lets you do literally any skill halfway decently (at least enough to properly Aid Another on literally any skill check). They're not massively busted all by themselves, but they're an outstanding force multiplier that could contort the game.

Battle Dancer: 56

It's a monk. Without the monk benefits, but with full BAB and a single good save. Ick.

Beguiler: 23

9th level casting. Sorcerers are pretty widely accepted to be T2 (Phenomenal Cosmic Power in one significant dimension that can alter the shape of a campaign). So how do Beguilers compare? Better HD, more skills (and good skills), Int casting, ability to use light armor. They've got my personal favorite casting mechanic (spontaneous casting off a limited, but all known, list; "which button do I push?") which also enables the best use of list-expanders like Arcane Disciple, bonus domains from prestiges, and the already-discussed Rainbow Servant (which would push them into T1). They keep pace with sorcerers on the magic front and they can do more without magic than sorcerers can. If you were to make a bard undeniably T2, it would end up looking a lot like a beguiler.

And as far as mind-affecting immunity? Silent image wins against mindless things, dispel can crack spell immunity. Bringing a lot of those in, though? That's the game being distorted by the capabilities of the class/character. Which is a clear mark of T2.

Binder: 34

A bunch of tricks you can do all day (but generally not more than twice a minute for the best ones). Plenty of neat effects here, nothing game-distorting.

Cleric: 1

Remember that bit I mentioned about casting off a list where you know all of it? The cleric list is one of the biggest in the game; anything you can't solve today you can probably solve tomorrow by picking the right spell. They also have a bunch of "you can probably solve today" buttons for general preparation, along with turning and an entirely adequate chassis which means you're probably going to make your fort saves and will saves.

Commoner: 6

Fat, drunk, and stupid is about this class's only way to go through life. It's also no way to go through life, at least according to Animal House.

[Tiri]

Fat, drunk, and stupid is about this class's only way to go through life.

Well, that's certainly a bit insensitive. Most of us here would probably be commoners if built with D&D rules.

[JBarca]

Well, that's certainly a bit insensitive. Most of us here would probably be commoners if built with D&D rules.

Sure, but we don't live in a world that prizes combat ability above literally everything else.

[Jormengand]

Well, that's certainly a bit insensitive. Most of us here would probably be commoners if built with D&D rules.

Most of us would be experts. Too many skills too high to be a realistic commoner.

[Pleh]

If you rank things in a way they are not used, you don't get good rankings. No one actually takes 20 levels of a base class with no feats, items, PrCs, dips, or ACFs. You could try to do what the old tier system did, and have a separate ranking for PrCs, but that causes immediate problems when you try to rank things like the Rainbow Servant that are heavily dependent on the class chassis. Rainbow Servant is nuts if you cast like a Warmage, decent (but outclassed) if you cast like a Wizard, and awful if you cast like a Sorcerer. If that information isn't baked into base class rankings, it's really hard to represent effectively.

Feats and items (wbl) are part of a base class, in my mind. You haven't built a complete character without those things.

Multiclassing is where I draw the line. Dips and prcs are ways to optimize your build, not your class. Feats and items can easily optimize your class without just taking some other class.

Acfs are the fuzzy line.

[eggynack]

Just noticed, the class list is definitely missing magewright (ECS, 256), and it's my opinion that urban druid (DrC, 57) is sufficiently an entirely new class, as opposed to an ACF (no matter how I have it listed in the handbook), to merit inclusion. That's how it seems to be listed, anyway.

[Doctor Despair]

Jormengand, what's your take on the PrC situation? Should we allow them in consideration for classes? Personally, I'd rank bard 23 if you could consider PrCs, but 32 if you can't. Small difference, since I'm certain they'll end up a resounding 3 regardless of how I rank them, but still, for accuracy's sake. :p

[Karl Aegis]

What spells in core were you thinking of that put the Cleric list ahead of Wizard? Not to put words in your mouth here, but all I can think of is the divination line, glyphs, and gish buffs. I'm not overly familiar with the Wizard list, though, so maybe I'm missing something.

Clerics get some cool things that Wizards don't. Evocations like Spiritual Weapon, Invisibility Purge (when the enemy is too close to Glitterdust), Blade Barrier Hallow/Unhallow and Consecrate/Desecrate. Travel options like Water Walk, Air Walk, Wind Walk. Protections like Hide From Undead, Aid, Spell Immunity, Meld into Stone, and Death Ward. Even things like Heal/Harm, the Restoration line, the Holy Word line, Raise Dead and Resurrection shouldn't be underestimated. Sure, they don't get Shrink Item, but overall some very good effects.

[Troacctid]

We're just ranking the base classes here. Prestige classes are a different beast.

Bard: 3
This shouldn't be controversial—high versatility and medium spellcasting make for a strong class. Poor spell selection or tactics can bring you down a little, and inspire courage optimization or ACFs like the animal companion can bring you up, but it's all still comfortably within the bounds of T3.

Battle Dancer: 5
Class is very bad—worse than Monk—but at least it's still better than NPC classes.

Beguiler: 2
Putting the Beguiler in a different tier from the Sorcerer and Favored Soul is madness. Anyone who thinks those classes are T2 but Beguiler is T3, I challenge you to craft a list of known spells for them at various levels that is clearly better than the Beguiler's. You can't do it. Beguilers have game-breaking spells at low, medium, and high levels even if you don't consider what you can get with feats, and they also have incredible skill access on top of it. They're extremely powerful—overpowered, even—and if Sorcerer is your benchmark for T2, then Beguiler has to land there as well.

Binder: 32
I don't think this is controversial either. Binders have decent skills combined with a nice suite of abilities and pretty solid scaling for the high levels. We can talk about whether Zceryll makes them T2—which IMO rests solely on whether you rate it with or without the "permanent duration" dysfunction—but I can't see an argument for any lower than T3.

Cleric: 1
One of the most powerful and versatile classes in the game.

Commoner: 6
Literally the #1 worst class in the game.

[Doctor Despair]

Binder: 32
I don't think this is controversial either. Binders have decent skills combined with a nice suite of abilities and pretty solid scaling for the high levels. We can talk about whether Zceryll makes them T2—which IMO rests solely on whether you rate it with or without the "permanent duration" dysfunction—but I can't see an argument for any lower than T3.

What "permanent duration" dysfunction are you referring to? My Google-fu is weak this time and I've not heard of it before.

[Troacctid]

What "permanent duration" dysfunction are you referring to? My Google-fu is weak this time and I've not heard of it before.

There is no duration on Zceryll's summons ability, so you get to keep your summoned aliens indefinitely unless they are dispelled. There's also no limit on the number of summons you can have at once, so you can just summon more, and more, and more, and more, until you've amassed an infinite army. This is probably not the intent.

[Jormengand]

Magewright will be added to the list.

I don't think we should assume people are taking prestige classes or variants. Variants with a significantly different tier can certainly be nominated to be handled at the end. I may add a few things which aren't strictly base classes (hit dice, for example - I'm genuinely not sure where outsider and dragon go) at the end as well.

Prestige classes aren't a helpful metric when determining class tier, because almost any class can go into Ur Priest. You need to stand on your own feet, not the shoulders of prestige classes.

[Troacctid]

Urban Druid needs to be separate too. It's at least as different from Druid as Jester is from Bard. And it's presented as an entirely different class.

[Jormengand]

Urban Druid needs to be separate too. It's at least as different from Druid as Jester is from Bard. And it's presented as an entirely different class.

Fair point on that one.

[Pleh]

Prestige classes aren't a helpful metric when determining class tier, because almost any class can go into Ur Priest. You need to stand on your own feet, not the shoulders of prestige classes.

I agree with one caveat.

Prcs are a useful metric in determining the flexibility of a base class. A base class can qualify for lots of good quality prcs easily, but not need to take any of them to shine alongside prcs. This is the mark of a strong base class.

Some base classes require prcs to function and don't easily access many good prcs (often because of stringent requirements for useless feats).

A prc can be evaluated on its own merits, but it can also measure a base class by its need and access to prcs.

It's a bit of a nebulous measurement, but then so is feat and spell selection.

[Nifft]

There is no duration on Zceryll's summons ability, so you get to keep your summoned aliens indefinitely unless they are dispelled. There's also no limit on the number of summons you can have at once, so you can just summon more, and more, and more, and more, until you've amassed an infinite army. This is probably not the intent.

As an aside...

Spoiler: What I Do about Zceryll

Show

- You can summon one monster.
- The monster sticks around until it dies or until you dismiss it.
- After the monster dies or is dismissed, you must wait 5 rounds, and then you can summon another one.

[Zanos]

I would just call it summon monster [x] where caster level = binder class level. Seems fair.

[Tiri]

Most of us would be experts. Too many skills too high to be a realistic commoner.

Well, by 'here' I meant 'in the world' not 'on GiantitP'.