Optimization Ranking system

[WarKitty]

So this is a question that has come up repeatedly: how do the different classes react to optimization? Well, why not build a tier system for them? I'm going to start with a three-step system. Since I'm not the best optimizer here, I'm asking for input on what classes you would put where; classes will get put in after getting 2-4 opinions (depending on who offers the opinion and how consistent they are).


Rank 1: Little to no optimization. Reaches close to full potential with obvious choices. Mistakes are unlikely to seriously impact playability.
Example class: Druid

Rank 2: Some optimization. Optimized characters are significantly better than non-optimized, but non-optimized characters are still decent in comparison. Mistakes hurt but are not huge.
Example class: Rogue

Rank 3: Significant optimization. Wild difference between optimized and non-optimized characters. Build mistakes are near-fatal.
Example class: Fighter

[Yorrin]

I like where you're going... but avoid the word "tier" when describing it. That word will cause more confusion and anger than it is worth.

I'll type something up in a second with my first take at such a ranking, though.

[Saph]

I'd put all three ToB classes in at rank 1. They're very hard to screw up.

Sorcerer and Wizard go at rank 3. They're fantastic in the hands of a skilled player and abysmal in the hands of a bad one. I remember one campaign where we had both in the same party!

I'd put Fighter at rank 2. It's hard to completely mess up a full-BAB class, since they can always be at least kind-of-okay simply by putting your highest score in Strength and taking a two-handed weapon.

[Ernir]

I'd suggest finding a different name than "tiers", if this really takes off, we don't want the confusion...

Anyway. Classes with a small optimization range (Tier 1):
Druid, All the ToB classes, Dragonfire Adept, Beguiler, Dread Necromancer, Duskblade, Dragon Shaman, Soulborn, Healer (always sucking equally qualifies, right?).

Middle of the road (Tier 2):
Totemist, Psion, Psychic Warrior, Binder, Warlock, Scout, Spellthief, Rogue, Paladin, Cleric.

Classes with a big optimization range (Tier 3):
Sorcerer, Archivist, Wizard, Bard, Monk, Factotum (ever seen a poorly built Factotum? It SUCKS.), Incarnate, Barbarian, Warmage, Ranger, Hexblade, Marshal, Fighter, Favored Soul, Wilder, Ninja (CA), Samurai (CW), Swashbuckler, Knight, Truenamer.

[Yorrin]

I'd put all three ToB classes in at rank 1. They're very hard to screw up.

Sorcerer and Wizard go at rank 3. They're fantastic in the hands of a skilled player and abysmal in the hands of a bad one. I remember one campaign where we had both in the same party!

I'd put Fighter at rank 2. It's hard to completely mess up a full-BAB class, since they can always be at least kind-of-okay simply by putting your highest score in Strength and taking a two-handed weapon.

I think you and I are thinking at very different levels/definitions of optimization. A Fighter with Greater Weapon Specialization and Power Attack who swings his big stick around is VASTLY different from a Dungeoncrasher Fighter with Improved Bull Rush, Improved Trip, Knock Down, Knockback, etc.

Likewise, a Sorcerer is really easy to screw up- if you don't know what the good spells are you can mess yourself up pretty bad (ie- picking three damage spells at every level). But a Wizard can always spend a little money and have a whole new spellbook instantly, and is thus much harder to seriously screw up.

[WarKitty]

Wizard should probably be split depending on how strictly your DM enforces the payment for new spells and how hard they are to find.

[Saph]

I think you and I are thinking at very different levels/definitions of optimization. A Fighter with Greater Weapon Specialization and Power Attack who swings his big stick around is VASTLY different from a Dungeoncrasher Fighter with Improved Bull Rush, Improved Trip, Knock Down, Knockback, etc.

You've got a point. I'd still argue for rank 2, though, since I generally find a badly played Fighter is more useful than a badly played Sorc/Wizard.

Likewise, a Sorcerer is really easy to screw up- if you don't know what the good spells are you can mess yourself up pretty bad (ie- picking three damage spells at every level). But a Wizard can always spend a little money and have a whole new spellbook instantly, and is thus much harder to seriously screw up.

Here I don't agree. I've seen more terribly played Wizards than I care to remember. The ability to re-pick your spell selection isn't much use if you don't know which spells are the good ones.

[JeminiZero]

I would argue that all the tier 1/2 fullcasting classes (even Druids) would Rank 3, (or 2 at the very least, in the case of Druids), on the Optimization scale, for 1 reason: Spell Selection (and indirectly, the feat selection to modify their spell selection).

The spell choice difference between an Ur-optimizer and that guy who closes his eyes, opens the PHB to a random page and puts his finger down, can spell a massive world of difference. While its true that an utterly unoptimized Druid is reasonably strong (Wild Shape and Animal Companion), if the guy has any idea about what hes doing, soon Venomfire Fleshrakers, Control Weather and various other shenanigans start to appear.

[WarKitty]

Do I need to add more rankings? I can expand it to a 5-rank system if needed.

[Saph]

I would argue that all the tier 1/2 fullcasting classes (even Druids) would Rank 3, (or 2 at the very least, in the case of Druids), on the Optimization scale, for 1 reason: Spell Selection (and indirectly, the feat selection to modify their spell selection).

The spell choice difference between an Ur-optimizer and that guy who closes his eyes, opens the PHB to a random page and puts his finger down, can spell a massive world of difference.

Remember though that Druids can spontaneously convert any spell into a Summon Nature's Ally. It's pretty hard to mess up your spell selection when every spell doubles as "make an extra fighter".

[Prime32]

I'd put Fighter at rank 2. It's hard to completely mess up a full-BAB class, since they can always be at least kind-of-okay simply by putting your highest score in Strength and taking a two-handed weapon.

IIRC the playtest fighter's feats were
Exotic Weapon Proficiency (bastard sword), Weapon Focus (bastard sword), Weapon Specialisation (bastard sword), Weapon Focus (longbow), Weapon Specialisation (longbow), Toughness

[Yorrin]

The ability to re-pick your spell selection isn't much use if you don't know which spells are the good ones.

I'll respectfully disagree and we'll leave it at that.

Given that such is my thinking, however, here are my categories:

Nearly Impossible To Optimize
aristocrat
commoner
noble
warrior
truenamer
eidolon
hexblade
samurai

Hard to Optimize
marshal
ninja
rogue
soulknife
swashbuckler
dragon shaman
knight

Moderate Optimization Possible
artificer
bard
favored soul
healer
psion
psychic warrior
shaman
shugenja
sohei
sorcerer
spellthief
warlock
wu jen
dread necromancer
shadowcaster
ardent
divine mind
lurk
beguiler
duskblade
dragonfire adept
monk
paladin
ranger
scout
expert
factotum

Very Optimizable
barbarian
fighter
crusader
swordsage
warblade

With Proper Knowledge Can Optimize Mid-Game
adept
cleric
druid
eidoloncer
mystic
wizard
archivist
incarnate
soulborn
totemist
binder
erudite (variant psion)


Let the criticism begin

EDIT: I should probably clarify my assumption that this is based on "I completely screwed up my build for the first several levels, so how screwed am I?"

[Saph]

IIRC the playtest fighter's feats were
Exotic Weapon Proficiency (bastard sword), Weapon Focus (bastard sword), Weapon Specialisation (bastard sword), Weapon Focus (longbow), Weapon Specialisation (longbow), Toughness

You do wonder sometimes . . .

[WarKitty]

Just as an aside: I'm not considering anything that can change by day for free to be included in the ranking system. Mostly, this means prepared casters such as the cleric or druid that have access to all spells should be considered as using good spells at all times.

Basically, if you got handed a completely unoptimized character with no retraining allowed, how hard would it be to get that character to work?

[molten_dragon]

I like this idea. Here's my thoughts on the core classes.

Babarian: Rank 2. There's a decent amount of optimization potential there (spirit lion totem anyone?) but they're hard to completely screw up.

Bard: I'm not too familiar with bards, but I would go for rank 3, simply due to the fact that bad spell choices are permanent.

Cleric: Rank 2. There's some definite potential for optimization with a cleric, but the most important part of the class (spells) is hard to screw up really badly, since you can change them every day.

Druid: Rank 2. Not taking natural spell, poor spell selection, poor selection of forms for wild shape, and a poor choice of animal companion can all reduce a druid's power significantly. Some of those problems can be easily fixed, but others aren't quite so simple. There's also quite a bit of room to optimize above a standard core druid as well. Greenbound summoning, rashemi elemental summoning, dragon wild shape, and several other excellent feats offer a druid a lot of potential beyond core.

Fighter: Rank 3. The differences between a well-optimized fighter and a poorly optimized one are vast.

Monk: Hard to say honestly. Probably tier 2. It's harder to judge when it's the difference between bad and worse rather than good and bad.

Paladin: Rank 2. For similar reasons to the barbarian.

Ranger: Rank 2. Ditto the barbarian and paladin.

Rogue: Rank 2. There's optimization potential, but they're hard to completely screw up.

Sorcerer: Rank 3. Poor spell choices are nearly permanent, and since they're all you get, poor choices are crippling.

Wizard: Rank 2. Poor spell choices are just as crippling as for a sorcerer, but since wizards can always write new spells to their spellbook, poor choices can be fixed much easier.

[Lans]

I think two systems might be appropriate here.
One system to rank the classes out of how hard it is to get full potential

Another to rank the difficulty in fixing a crap build.

[JaronK]

I think two systems might be appropriate here.
One system to rank the classes out of how hard it is to get full potential

Another to rank the difficulty in fixing a crap build.

This is worth considering. Druids are almost impossible to mess up (you get an Animal Companion that's solid no matter what) but you actually have to work a bit to make them really good (like, read through their spells, and actually search for decent Wild Shape forms).

Likewise, Rogues can be hard to totally mess up (just put your skill points in the skills you plan to use) but you actually have to know what you're doing to avoid being worthless in a fight.

Fighters, meanwhile, are really easy to make completely worthless. So are Wizards evidently, but they can rocket to the top just as easily as Druids (hey, I just read this spell called Glitterdust...).

JaronK

[Lans]

I figure things like clerics, druids, incarnates, binders can get there game on after 24 hours IG and a trip to optimization land.
Wizards needs to gain a level or two and/or find a library.
Sorcerer, favored soul, shadow caster may need several levels
Truenamer needs Energy Drain or magic mart, and likely several levels and splat books, and multiclassing, and maybe house rules.
Fighter is SOL out side of retraining

The list casters can't really be screwed up, and are pretty resistant to change.

[Pechvarry]

I don't think "how hard to salvage a terrible character" is really a worthy enough criteria. However, "how good can you make it" vs "how easy is it to make a useless one in the first place?" is quite relevant.

On the topic of casters: even when clerics are preparing cure spells and the wizards are blasty, they're still easily seeing invisibility and fly on their lists which say "no" to melee.

[JaronK]

The important thing to remember here is that the initial 3.5 and 3.0 playtesters thought core was balanced... and we've just shown what the Fighter was. In other words, it's quite possible to mess up all the classes to make them just as bad as that Fighter... since that's exactly what the Playtesters did. Interestingly, this means it's hard to mess up the Monk, since even the Playtesters made a pretty average Monk. But evidently it's quite possible to make a Druid as bad as that Fighter (IIRC the playtest Druid that we see in the PHB has Weapon Focus: Scimitar, because Wild Shape wasn't used in combat).

So really all classes in core can be made to suck as much as that Fighter.

JaronK

[Prime32]

The important thing to remember here is that the initial 3.5 and 3.0 playtesters thought core was balanced... and we've just shown what the Fighter was. In other words, it's quite possible to mess up all the classes to make them just as bad as that Fighter... since that's exactly what the Playtesters did. Interestingly, this means it's hard to mess up the Monk, since even the Playtesters made a pretty average Monk. But evidently it's quite possible to make a Druid as bad as that Fighter (IIRC the playtest Druid that we see in the PHB has Weapon Focus: Scimitar, because Wild Shape wasn't used in combat).

So really all classes in core can be made to suck as much as that Fighter.

JaronK

I recall stories of wizards with 8 Str who ran into melee wielding scimitars one-handed and never cast spells...

[FMArthur]

Druids are not anywhere close to being impossible to fail at playing, which is pretty much what the highest ranking entails, isn't it? You say Wild Shape and Animal Companion set the bear minimum power level pretty high? I say you've never seen a Druid who likes cute little critters and never gives any thought to taking the big burly animals. That is, by the way, a perfectly reasonable and normal fantasy archetype that players can and will try to fulfil, either not knowing or not caring how useless it makes them. "All animals of the forest are your friends" is basically the standard Druid in fiction, not a mutant bear army on a rampage.

[WarKitty]

The important thing to remember here is that the initial 3.5 and 3.0 playtesters thought core was balanced... and we've just shown what the Fighter was. In other words, it's quite possible to mess up all the classes to make them just as bad as that Fighter... since that's exactly what the Playtesters did. Interestingly, this means it's hard to mess up the Monk, since even the Playtesters made a pretty average Monk. But evidently it's quite possible to make a Druid as bad as that Fighter (IIRC the playtest Druid that we see in the PHB has Weapon Focus: Scimitar, because Wild Shape wasn't used in combat).

So really all classes in core can be made to suck as much as that Fighter.

JaronK

Point taken. I'm going to say low optimization level is probably someone who looked through the books and didn't do anything absolutely stupid (like not wearing armor on a fighter, for example, or trying for weapon feats on a wizard) but didn't really make much attempt to optimize either.

Druids are not anywhere close to being impossible to fail at playing, which is pretty much what the highest ranking entails, isn't it? You say Wild Shape and Animal Companion set the bear minimum power level pretty high? I say you've never seen a Druid who likes cute little critters and never gives any thought to taking the big burly animals. That is, by the way, a perfectly reasonable and normal fantasy archetype that players can and will try to fulfil, either not knowing or not caring how useless it makes them. "All animals of the forest are your friends" is basically the standard Druid in fiction, not a mutant bear army on a rampage.

Wild shape and Animal Companion are being discounted because they are easy to change, not because you can't make them suck. If I go into a higher-op game and pick a poor wild shape form and end up not doing anything, I can go back and re-do it the next day.

If anyone has better ideas on a standard I'm open to them, btw. Mostly I wanted to measure the difference between, say, picking bad spells on a cleric and picking bad spells on a sorcerer.

[Zaq]

Out of curiosity, what's the source for these playtest fighters with bastard swords and druids with WF: Scimitar? I'd be quite interested in seeing some of those, if they're around anywhere.

[Pechvarry]

Now we're getting into even more problems:

How easy is it to build a subpar character of this class?
vs
How easy is it to play a character of this class so poorly it's terrible?

A druid could do nothing but summon kittens, and that's a bit more realistic than an 8-str melee wizard (the latter type definitely exists, but that's a player thing, not a class thing). But it's easier for a Wizard to realize flying+invisible = untouched than it is for a Druid to crack open a couple Monster Manuals and actually compare combat stats of different potential animal forms.

I feel like the Druid playtesting is an example of playing the class idiotically: they were silly to assume we wouldn't turn into BEARS.

On the other hand, changing into a bird to fly through a window, then into a mouse to creep under the door, is a perfect example of the poorly-played Druid completely stealing the Rogue's show. They're still obsoleting other roles, like a true tier 1.

Anyway, I DON'T think their playtest conclusively shows a PHB-only fighter to be as potent as a PHB-only Druid. Maybe in 3rd edition (though unlikely), but definitely not in 3.5

[Amphetryon]

Nitpicking, and probably falling into the "I like them so I overvalue them" paradigm, but I think Hexblades are hard to optimize, not nearly impossible. They need the Dark Companion variant, IMO, but they are reasonably good as debuffing melee or ranged full BAB types.

[Fiery Diamond]

Point taken. I'm going to say low optimization level is probably someone who looked through the books and didn't do anything absolutely stupid (like not wearing armor on a fighter, for example, or trying for weapon feats on a wizard) but didn't really make much attempt to optimize either.



Wild shape and Animal Companion are being discounted because they are easy to change, not because you can't make them suck. If I go into a higher-op game and pick a poor wild shape form and end up not doing anything, I can go back and re-do it the next day.

If anyone has better ideas on a standard I'm open to them, btw. Mostly I wanted to measure the difference between, say, picking bad spells on a cleric and picking bad spells on a sorcerer.

I think this means that we have multiple different variables that should be considered, perhaps splitting this ranking into multiple rankings. Basically, "easy to change" doesn't mean it should be discounted when determining optimization range, but it does mean that it should be discounted when determining how hard it is to permanently screw up a character, which are two different things.

If I have a player who wants to be a druid with a hawk animal companion, who uses wildshape to go hang with the animals and only uses communication-oriented spells, then this is poorly optimized. With regards to optimization range, you have to consider this character, because "But he can change it!" doesn't matter if that's the character he wants to play. The availability of change only matters if the player is more concerned with being competent in multiple situations (such as combat) than with keeping his character intact - and he might not be. And it isn't bad gaming if he doesn't want to change it.

[WarKitty]

Well, here's the primary idea I had here:

Many of us run reasonably optimized groups. Now, suppose we want to induct a new player. Someone who's really never played D&D before and doesn't know what optimization is. What class do you give them?

Conversely, suppose you want to introduce an optimizer into a non-optimized group. If you're like me or one of the other high-optimizer people, there's a little twinge when choosing a less than optimal selection. If I want to optimize to my heart's content, what do I pick?

It might be best to steal the model of the PrC tier system here. A non-optimized sorc can drop two or three tiers easily. An optimized fighter can go up one. An optimized adept can probably go up two or three.

[JaronK]

Many of us run reasonably optimized groups. Now, suppose we want to induct a new player. Someone who's really never played D&D before and doesn't know what optimization is. What class do you give them?

Crusader. Seriously, they can't go wrong. See, even a Druid is easy to mess up, because you might care about your animal companion and thus see it as horribly wrong to put it in danger by putting it in combat. And you might not realize Wild Shape is for fighting in combat, since you'd lose your weapons to do that. Yes, I've seen both of these.

But Crusader? You pick 5 maneuvers, and you have 6 total choices, at level 1. Go ahead, pick randomly if you want (though note that two of the maneuvers are the same, and one of them does something about AoOs that new players often don't understand, so one of those three is probably what you don't take). You know you're supposed to hit things in melee. Go hit them. Hey look, you have your choice of three random maneuvers. Pick one, and hit them with it. Neat. It's REALLY tough to go wrong, and even on new players I've never seen them struggle with a Crusader.

Conversely, suppose you want to introduce an optimizer into a non-optimized group. If you're like me or one of the other high-optimizer people, there's a little twinge when choosing a less than optimal selection. If I want to optimize to my heart's content, what do I pick?

Commoner. I've done it. It was overpowered. Sorry, at the end of the day, magic items and feats can get you REALLY powerful regardless of class (for the record, Orc Commoner with Headlong Rush, Spirited Charge, Shock Trooper, and the prerequisites. Skillful Valorous Lance. A really nice Warbeast mount. Maxed out Ride and Handle Animal. A bunch of items to make him more likely to hit).

JaronK

[WarKitty]

For completeness should probably figure out how to handle magic item use. UMD can go a long way towards rescuing a character, but it's not the same as the raw power of the class.

And ok, you win on the commoner. Still, it's a decent idea in principle despite needing some working out.