Towards a Better Tier System

[Cosi]

Discussion in the Community Re-Tiering thread has brought up a number of issues with the Tiers as formulated by JaronK. As such, I think it's appropriate to attempt to develop an alternative tiering system that addresses those issues. For example:

The Tiers are ranking something that is not strongly predictive of observed ability: class in isolation. People don't play without feats, items, or PrCs. At the same time, evaluating those things poses some potential problems. Is it fair to give a class credit for a PrC? What about classes that are less than twenty levels long (basically every class except Druid)?

On a related note, it's not clear which and how many splatbooks should be evaluated. Core + Class Book? Everything? Something in between? Should web enhancement, setting specific, or Dragon Magazine material be included in rankings?

Class power varies across level. A 1st level Fighter is able to contribute to a 1st level party, even if he's less effective than a Druid. But you wouldn't give a 10th level Fighter the time of day. A ranking that says the Fighter is in such-and-such a tier at all times isn't reflecting the reality that the Fighter's value changes dramatically throughout the game.

There isn't any kind of objective standard for tier membership, which makes it difficult to judge whether or not a class rises to a particular level. What does "campaign smasher" mean? Should we evaluate build options?

Some tiers are covering multiple kinds of class. For example, Tier Three is supposed to be home classes that are "capable of doing one thing quite well, while still being useful when that one thing is inappropriate, or capable of doing all things, but not as well as classes that specialize in that area," which is two entirely different groups of things.

Simply saying "class X is in tier Y" is not very helpful for evaluating class power. Knowing that a class can smash the campaign doesn't tell you anything about the circumstances where that might be done, or how to prevent it from happening.

So, I think we need a new system to rank classes in a way that is more detailed and more reflective of actual play. My proposal looks something like the proposal to rank classes by complexity that was posted some time ago, with rankings at several breakpoints.

But instead of jumping directly into a discussion of how to rank things, we should first explore the problem space. What are the flaws with current proposals for ranking classes? What do we want a system for ranking classes to do?

[Muggins]

Classes at Face Value
I don't think it's helpful to ignore class variants and build options beyond what is provided by the base class. ACFs should definitely be included in the discussion, and Feats probably so, but when inspecting the baseline power of the class I think most PrCs should be ignored. Classes don't exist in a vacuum, and ignoring things like ACFs and feats just doesn't happen in actual play.*

In cases where the presence of a particular variant actually does cause a Tier change on its own, it seems reasonable to provide a separate ranking. As good as Erudite is, Spell-to-Power Erudite is just earth-shatteringly better. In regards to prestige classes, the only significant, important example I can think of are the fixed-list spontaneous casters (beguiler, dread necromancer, and warmage) with list-expansion PrCs like rainbow servant.

*Speaking of which..

Level of Play
I've always considered the Tier System to be based around at least mid-OP play, if not high-OP. There's a reason why Wizard is Tier 1, even though a player could choose awful spells and be barely more competent than the fighter or monk, and it's because of what the class is capable of when some serious thought has been put into it. And I think it should stay that way.

If someone wants to make an Idiot's Guide to the Tier System based on low-OP actual play, though, I wouldn't say no.

What do the Tiers Mean?
I'll admit, some of the Tiers are kinda weird, and some of the classes fall through the cracks. The original Tier System has a few classes lit in blue (best of its class) and red (worst of its class) to help deal with this, and I think that's probably the best way to go about it. We don't need a Tier 3.5 for the psychic warrior or crusader, we need to mark them in red to clarify that they have much less versatility than what's offered by classes like the bard or binder. Or possibly a blue Tier 4, to demonstrate they do a better job than the rogue or hexblade.

Unfortunately, the other thread which is active right now doesn't include this system of grading the tiers, so we're getting results which aren't terribly accurate and it can feel like we need extra tiers to compensate.

Tier 1: Can break the game in a new way every day.
Tier 2: Can break the game, but only in one or two ways.
Tier 3: Power and minor utility, or major utility.
Tier 4: Power but no utility, or minor utility.
Tier 5: Sub-par in its field, or unfocused.
Tier 6: Worthless.

[Nifft]

I'd like to see at least two axes: Versatility on one side, Power on the other.

Versatility:
A - Can whip out whatever set of abilities is needed to solve the current problem.
B - Has a flexible set of powers on demand which are applicable to a variety of different situations.
C - Has an inflexible set of powers, but it's a pretty large set.
D - Has an inflexible and small set of capabilities.
E - Has one or two different capabilities.
F - Has no special abilities.

Power:
1 - Can use raw power to push through nearly any situation, including those significantly above the character's level.
2 - Can use raw power to push through some situations above the character's level.
3 - Can use raw power to push through nearly any situation appropriate for the character's level.
4 - Cannot use raw power to push through many situations appropriate for the character's level; requires consumable resources or cleverness to handle some level-appropriate situations.
5 - Cannot use raw power to push through most level-appropriate situations. Cannot contribute meaningfully to some level-appropriate situations.
6 - Cannot contribute meaningfully to the majority of level-appropriate situations.


... then a Commoner might be F6, while a Wizard is A1, and other classes will fall somewhere in between.

[Red Fel]

The Tiers are ranking something that is not strongly predictive of observed ability: class in isolation. People don't play without feats, items, or PrCs. At the same time, evaluating those things poses some potential problems. Is it fair to give a class credit for a PrC? What about classes that are less than twenty levels long (basically every class except Druid)?

First, the Tiers aren't supposed to be predictive. They're supposed to be descriptive. They're designed to highlight the strengths and weaknesses, or more accurately the versatility, of a given class. A good Tier assessment includes the caveat that, in the hands of a pro, a class like Monk can be insane, while in the hands of an amateur, a CoDzilla can be pathetic.

Further, classes necessarily must be examined in a vacuum. The options are simply too varied, too diverse, to consider it any other way. If you're willing to consider multiclassing or PrCs, almost any class goes up a Tier if you add casting, or down if you take it away; certain templates are basically golden irrespective of class; and so forth. There are just too many ways to take even the worst classes and make them amazing, so you have to examine the class on its own merit.


On a related note, it's not clear which and how many splatbooks should be evaluated. Core + Class Book? Everything? Something in between? Should web enhancement, setting specific, or Dragon Magazine material be included in rankings?

For a system to be useful, it must be comprehensive. It should either include every splat, or include sufficiently specific criteria that every splat can be evaluated constructively. It does no good, for example, to create a rating system that evaluates only the PHB classes and no others, if I decide to play a Warblade.

Now, a sufficiently nuanced Tier system might stratify its ratings based on available materials. For example, a so-so class can become much more versatile with the options in a particular splat; in that case, it's worth noting that versatility is based on available materials.


Class power varies across level. A 1st level Fighter is able to contribute to a 1st level party, even if he's less effective than a Druid. But you wouldn't give a 10th level Fighter the time of day. A ranking that says the Fighter is in such-and-such a tier at all times isn't reflecting the reality that the Fighter's value changes dramatically throughout the game.

Well, yes and no. Assuming two high-op players playing, say, a Fighter and a Wizard, there will always be a gap between them. It simply becomes more pronounced at higher levels.

Further, in most cases, the Fighter's versatility - which the current Tier system measures - won't vary significantly between levels 1 and 20. What changes are his numbers - damage, attack, AC, and so forth. And while those become less and less valuable with levels, the point is that he's not doing less than what he previously did, it's simply not as useful. By contrast, some other classes do become more versatile as they level - for example, the Druid with its Wild Shape progression - but their versatility is generally enough to have put them well ahead of the curve to begin with.


There isn't any kind of objective standard for tier membership, which makes it difficult to judge whether or not a class rises to a particular level. What does "campaign smasher" mean? Should we evaluate build options?

Again, there is and there isn't. For instance, if a class can have, with minimal preparation, the tools to instantly overcome any challenge or encounter, that's pretty campaign-breaking. It's not that there's an explicit definition, although certain spells (e.g. Wish, Miracle, Gate) are generally agreed to check all the boxes.

And with respect to build options, as previously mentioned, that way lies madness. Fighter 1 can instantly jump up several tiers if the build reads Fighter 1/ Wizard 19, for example. There are just too many variables to provide any reasonable advice or explanation.


Some tiers are covering multiple kinds of class. For example, Tier Three is supposed to be home classes that are "capable of doing one thing quite well, while still being useful when that one thing is inappropriate, or capable of doing all things, but not as well as classes that specialize in that area," which is two entirely different groups of things.

This is a valid concern. If you were to break this Tier out, though, would you put one ahead of the other? Or would you split them into two parallel Tiers (e.g. Tier 3A and 3B)?


Simply saying "class X is in tier Y" is not very helpful for evaluating class power. Knowing that a class can smash the campaign doesn't tell you anything about the circumstances where that might be done, or how to prevent it from happening.

Simply saying "class X is in Tier Y" is not very helpful, period. It's a meaningless statement, like "The pen is blue." Unless someone is asking "What color is the pen?" the statement has no intrinsic value.

Most discussions of classes and Tiers include someone explaining why a class is - or should be - in a given Tier. And that is helpful. Knowing, for example, that a Cleric is considered Tier 1 because it is capable in melee, strong at buffing, able to cast spells in armor, and has access to certain game-breakers at higher levels (e.g. Miracle), is extremely detailed and helpful. Moreover, that's the kind of context that tells you what might make a class into a campaign-breaker.


So, I think we need a new system to rank classes in a way that is more detailed and more reflective of actual play. My proposal looks something like the proposal to rank classes by complexity that was posted some time ago, with rankings at several breakpoints.

But instead of jumping directly into a discussion of how to rank things, we should first explore the problem space. What are the flaws with current proposals for ranking classes? What do we want a system for ranking classes to do?

I think the biggest flaw is consensus. Simply put, there are - and will probably always be - disagreements over what class goes in what Tier. And while making the Tiers more detailed, or breaking them out into more granular divisions, might allow us to use more detail, I don't think that will resolve the disagreements.

As for what we should want a system to do, I think it's important to realize the nigh-impossibility of ranking classes according to power. That's because power varies based on the player, rarely the class. There are a few classes that are almost powerless, it's true, but the vast majority, in the right hands, can be optimized into something very potent. That's a measure of the player, not the class.

The only thing intrinsic to the class is what the class can or cannot do - that's versatility. A Fighter, absent multiclassing, magic items, or racial features, will never fly. That's a limitation inherent to the class, an upper bound on its versatility. That's something the current Tier system is designed to measure.

Does it do so perfectly? No. Obviously, there is still a lot of disagreement. Could it be refined? Certainly. But is the idea sound? I think so. I think that measuring the versatility of the class, itself, in a vacuum, is the only way to consistently rate these things absent the influence of the player.

[Aimeryan]

I don't think assigning a class a single tier works to be honest, not if the idea is to be actually informative on how that class will play in a campaign. It gives some indication of power, but even that can fail in certain scenarios; a Fighter (using ACFs mixed probably with a Barbarian dip and some PrC) can Ubercharge a decent degree of combat encounters to such an extent as to cause the DM to have to utterly play around it. In that situation the Fighter is incredibly powerful, possibly more so than say a Sorcerer.

If the idea is that you Fighter is in the low/weak tiers and, therefore, as a DM you should have no problem with the Fighter being too powerful in your campaign that would be wrong. Indeed, worse, if the Fighter does go full Ubercharger it can be bad for both the DM and the Fighter because now the DM has a problem finding something the Fighter is good at without being far too good at.

A tiering system that was there to inform players and DMs would need to specify that the Fighter was a) capable of a taking up one of a number of builds [Variable], b) was locked into a build once taken [Inflexible], c) could break encounters if taking some builds to the extreme [Potentially Powerful], d) would find it difficult to be useful outside of that build [Vulnerable to being Impotent]. Anything less and I think you are losing valuable information - in fact, it should probably be stated where the edges of the class can fall so that players/DMs can avoid those if appropriate. For Fighters that would be something like: Ubercharger build can break encounters, Grappler builds can find they become useless in pretty much all encounters after a certain point, either way no real strengths/uses of the class outside of combat encounters.

[Pleh]

Fundamental questions first.

1. What is the purpose of the tier system? For the most part, people seem to agree that it serves the need to communicate succinctly the real play value of the various character options.

More simply, it is a shorthand reference to relative power/effectiveness.

If we wish to make a useful reference, we have to consider for whom we are writing. Who is the audience? Are we writing a reference for high op players who are already familiar with most of the material, for mid op players who understand the concept and just need directions as to what options they need, or low op players who need their awareness of the balance issues to be raised?

If we want to write to any of these, all of these, or any other group on combination, writing to multiple groups will diminish the focus and complicate the results. Telling an experienced player the class they've never played is T3 will communicate something different than telling a new player the bard they play every game is only T3 and they might insist that you just haven't played the bard enough to know all it can do.

Answering the question of exactly what we want the tiers to say and to whom they are speaking will inform us greatly in how it needs to be written.

[Cosi]

I've always considered the Tier System to be based around at least mid-OP play, if not high-OP. There's a reason why Wizard is Tier 1, even though a player could choose awful spells and be barely more competent than the fighter or monk, and it's because of what the class is capable of when some serious thought has been put into it. And I think it should stay that way.

Assuming this was meant as a reply to my post, I think you're sort of missing the point. I wasn't talking about optimization, I was talking about level. Optimization does have an effect, but I think "complexity" is likely better than "power" there.


Tier 1: Can break the game in a new way every day.
Tier 2: Can break the game, but only in one or two ways.

I am very strongly opposed to using "break the game" as a primary ranking. It just doesn't mean all that much, because anyone can break the game. Insofar as it's useful, I think accidental power swings are what people care about.


I'd like to see at least two axes: Versatility on one side, Power on the other.

First, rankings should all use the same scale. I like (S/)A/B/C(/F) because it lends itself to the use of + and - modifiers for extra precision.

Second, I think ranking versatility is sort of dumb. Having a bunch of tools that are almost good enough is massively worse than having even one tool that is good enough. Insofar as it means anything useful, it's just another term for "able to solve problems" which is the same thing as power.

I would probably go with "complexity to build" and "complexity to play". Then secondary terms for "how likely are you to build something horribly overpowered or underpowered" or "how dependent is this on splatbooks" or whatever.


Further, classes necessarily must be examined in a vacuum. The options are simply too varied, too diverse, to consider it any other way. If you're willing to consider multiclassing or PrCs, almost any class goes up a Tier if you add casting, or down if you take it away; certain templates are basically golden irrespective of class; and so forth. There are just too many ways to take even the worst classes and make them amazing, so you have to examine the class on its own merit.

At the same time, class in a vacuum creates a system that misses on huge swaths of the game. Hyperbolically, vacuum rankings would put a class with "if you take Power Attack, gain wish as an at-will SLA" and no other abilities in Tier Six, which is obviously stupid. There needs to be a middle ground somewhere more inclusive than "no feats, no items, class only, Final Destination" but less inclusive than "all possible builds"


For a system to be useful, it must be comprehensive.

No, for a system to be useful, it needs to reflect the space it's evaluating. If most games don't allow 3.0 material or web enhancements, those things should not be ranked. If most games do allow Eberron material or Dragon Magazine, those things should be ranked.


Well, yes and no. Assuming two high-op players playing, say, a Fighter and a Wizard, there will always be a gap between them. It simply becomes more pronounced at higher levels.

That's not a counter-argument. If Fighter/Wizard is closer at level 5 than it is at level 15, that difference should be accounted by different rankings for those two use-cases.


Again, there is and there isn't. For instance, if a class can have, with minimal preparation, the tools to instantly overcome any challenge or encounter, that's pretty campaign-breaking. It's not that there's an explicit definition, although certain spells (e.g. Wish, Miracle, Gate) are generally agreed to check all the boxes.

What about marginal cases? What about non-abusive uses (e.g. planar binding to get utility spells)? It's also worth noting that all your listed options are 9th level spells, and having two tiers worth of delineation for the power of 17th or higher level characters is inane.


And with respect to build options, as previously mentioned, that way lies madness. Fighter 1 can instantly jump up several tiers if the build reads Fighter 1/ Wizard 19, for example. There are just too many variables to provide any reasonable advice or explanation.

I think Fighter 1/Wizard 19 is pretty clearly a Wizard build rather than a Fighter build, and by the time you get something that is close to a Fighter with a Wizard dip rather than vice versa, you're probably back to the Fighter in terms of power.


This is a valid concern. If you were to break this Tier out, though, would you put one ahead of the other? Or would you split them into two parallel Tiers (e.g. Tier 3A and 3B)?

Personally, I would not use a single axis ranking. If I were forced to for some reason, it seems obvious that "competent in one area" is better than "almost competent in all areas".


Simply saying "class X is in Tier Y" is not very helpful, period. It's a meaningless statement, like "The pen is blue." Unless someone is asking "What color is the pen?" the statement has no intrinsic value.

And yet, that is exactly what the re-tiering thread is creating and what the original tiers were.


As for what we should want a system to do, I think it's important to realize the nigh-impossibility of ranking classes according to power. That's because power varies based on the player, rarely the class.

But it is undeniable that some classes are easier to optimize than others. Making a competent Truenamer is an exceedingly complicated endeavor that requires several different splatbooks to trawl for skill boosts. Making a competent Beguiler involves putting your highest stat into your casting stat. Maybe that's not "power", but it's certainly not "versatility".


The only thing intrinsic to the class is what the class can or cannot do - that's versatility. A Fighter, absent multiclassing, magic items, or racial features, will never fly. That's a limitation inherent to the class, an upper bound on its versatility. That's something the current Tier system is designed to measure.

I'm not at all convinced that you can usefully measure "versatility" in a way that ends up being different from "power".


If the idea is that you Fighter is in the low/weak tiers and, therefore, as a DM you should have no problem with the Fighter being too powerful in your campaign that would be wrong. Indeed, worse, if the Fighter does go full Ubercharger it can be bad for both the DM and the Fighter because now the DM has a problem finding something the Fighter is good at without being far too good at.

IMHO, optimized versions of "low-tier" classes are harder to DM than optimized versions of "high-tier" ones. The Ubercharge is a binary character. Either you have a problem that can be solved by "charge it really hard" and it is solved instantly, or you have a problem that can't be solved by "charge it really hard" and there's nothing for the character to do. Wizards on the other hand have abilities that are varied and not usually instant win buttons, which makes it much easier to write encounters that are challenging but not impossible.

[johnbragg]

What you're proposing is essentially going back to the drawing board.

So I'd argue for starting over, and focusing on the fact that the game is played around a table with other party members.

Tier 1 is bad not just because it has campaign-smashing abilities, but because Tier 1 classes very easily make the rest of the party irrelevant. Tier 3's can usually keep up with enough splats and bookdiving, but you have to work to keep Tier 3s relevant when Tier 1s get in gear.

Tier 5 is bad because Fighter and Monk can't do their jobs effectively past the first few levels.

Tier 2-4 work fairly well together, in JaronK's original text and in most people's experience.

So is it really important whether a Psychic Warrior is really upper Tier 4 or lower Tier 3, etc?

Side Note: Maybe next time you do a fighter fix, you start with the Barbarian chassis, and figure out the non-barbarian equivalents of Rage should be.

OP: "What do we want a system for ranking classes to do?"

IMO, a few basic tests.
1. Does this class need an adventuring party help kill monsters and take their stuff?
2. Does this class do an important job in an adventuring party?
3. Does this class do an important--and fun--job in an adventuring party?

I know there are people who like high-OP Tier One campaigns. I still don't know from interacting with them on forums what table-play is like in their campaigns, why Reed Richards and Charles Xavier and Doctor Fate and Tony Stark actually need each other.

[exelsisxax]

I'd like to see at least two axes: Versatility on one side, Power on the other.

Versatility:
A - Can whip out whatever set of abilities is needed to solve the current problem.
B - Has a flexible set of powers on demand which are applicable to a variety of different situations.
C - Has an inflexible set of powers, but it's a pretty large set.
D - Has an inflexible and small set of capabilities.
E - Has one or two different capabilities.
F - Has no special abilities.

Power:
1 - Can use raw power to push through nearly any situation, including those significantly above the character's level.
2 - Can use raw power to push through some situations above the character's level.
3 - Can use raw power to push through nearly any situation appropriate for the character's level.
4 - Cannot use raw power to push through many situations appropriate for the character's level; requires consumable resources or cleverness to handle some level-appropriate situations.
5 - Cannot use raw power to push through most level-appropriate situations. Cannot contribute meaningfully to some level-appropriate situations.
6 - Cannot contribute meaningfully to the majority of level-appropriate situations.


... then a Commoner might be F6, while a Wizard is A1, and other classes will fall somewhere in between.

This formulation is objectively superior to the single dimensional tier system. Conflating different terms makes those tier lists useless tools for anything but continuous arguments. Either you separate them(maybe add a third axis if one can be clearly delineated from the others) or you give up on tiers entirely because it's worthless. The other option is something like niche ranking (http://www.giantitp.com/forums/showthread.php?314701-Person_Man-s-Niche-Ranking-System) with whatever your precice niches/categories are.

[javcs]

I'd like to see at least two axes: Versatility on one side, Power on the other.

Versatility:
A - Can whip out whatever set of abilities is needed to solve the current problem.
B - Has a flexible set of powers on demand which are applicable to a variety of different situations.
C - Has an inflexible set of powers, but it's a pretty large set.
D - Has an inflexible and small set of capabilities.
E - Has one or two different capabilities.
F - Has no special abilities.

Power:
1 - Can use raw power to push through nearly any situation, including those significantly above the character's level.
2 - Can use raw power to push through some situations above the character's level.
3 - Can use raw power to push through nearly any situation appropriate for the character's level.
4 - Cannot use raw power to push through many situations appropriate for the character's level; requires consumable resources or cleverness to handle some level-appropriate situations.
5 - Cannot use raw power to push through most level-appropriate situations. Cannot contribute meaningfully to some level-appropriate situations.
6 - Cannot contribute meaningfully to the majority of level-appropriate situations.


... then a Commoner might be F6, while a Wizard is A1, and other classes will fall somewhere in between.

I fully agree with this.
Although, I'd consider a few other axes, even though that would make rating classes more complicated.
For the Versatility Axis, I'd also suggest another value or chain of values - "flexibility in selecting powers/abilities that are then permanently locked in once selected"; on the strong end of this subchain would be Sorcerer, Favored Soul, and also probably Psion and the other psionic full manifesters - and then at the low end of the subchain would be Fighter.



On the Power Axis, I'd suggest make it explicit that the Power value is for areas where their abilities/versatility can be applied, and possibly give a second rating for their power when their primary abilities/versatility are unusable or hard to make applicable.
Or maybe just expand it - right now, it doesn't have a middle ground where a class can use its raw power for some level appropriate situations, when the situation is one that the class's power can be applied to, but can't depend solely upon raw power - as is, I think that's falls between a 3 and a 4, or ... oh, nuts, that's a 3.5. Uh-oh.






Also, the very terminology of "Tier" has connotations and implications that are both unfortunate for what's being looked for here and actively counterproductive.
I therefore suggest using the term "Type" or "Category" to avoid the connotations and implications that go with the term "Tier".




IE, sticking with Nifft's system:
Wizard: Type A1


Factotum: Type B3

Rogue: Type D4

Fighter: Type E5?E6? I think

Commoner: Type F6

Most of the not primary casters are probably Type D, with some Type E, and vary on their power. The Specialist/focused full casters are probably mostly types C and D, depending on their spell lists.





Also, any rating system must assume the base class, ACFs (although ACFs that make a major difference probably warrant their own distinct rating entry).
Maybe an axis that measures how well it usually PRCs, but a base class should not be assumed to PRC for anything other than a specific to PRCing rating.
Feats cannot inherently be assumed, but perhaps there should be something measuring interactions with feats.
Plus "Why" commentaries/explanations.

[Aimeryan]

IMHO, optimized versions of "low-tier" classes are harder to DM than optimized versions of "high-tier" ones. The Ubercharge is a binary character. Either you have a problem that can be solved by "charge it really hard" and it is solved instantly, or you have a problem that can't be solved by "charge it really hard" and there's nothing for the character to do. Wizards on the other hand have abilities that are varied and not usually instant win buttons, which makes it much easier to write encounters that are challenging but not impossible.

Absolutely agree, however, if the tier system makes no effort to explain where potential problems lie with a class, such as an Ubercharger build (or any damage optimisation via one method that oneshots encounters), then it doesn't really help a prospective DM/player in making sure that the game will be challenging in a good and fun way. The point of a tier system from what I can see is to help the DM/players in choosing a balanced party that is capable of running a good campaign without being overshadowed by certain classes/builds.

If someone sees "Fighter 5" and thinks "Well everyone else is around Tier 5 or 4 so Fighter will fit, hmmm, how should I build this?" then goes and checks up good fighter builds, sees Ubercharger, bam! Unbalanced campaign. If that person/DM is informed that Fighter is Tier 5 but with a warning that Ubercharger builds are problematic and unlikely to be suited to a campaign balanced around Tier 5 and the reasons why then such a thing may be avoided.

[Pleh]

I think people here have different definitions of competency for characters.

There is competency where you only need to roll above a 5 and competency where you need to roll above a 10. There is a lack of competence where you need a natural 20 for any hope for success, and then there's "game breaking" where you don't have to roll because you've got a trick that bypasses the problem.

What level of competence do we expect classes to attain?

[Hurnn]

I think people here have different definitions of competency for characters.

There is competency where you only need to roll above a 5 and competency where you need to roll above a 10. There is a lack of competence where you need a natural 20 for any hope for success, and then there's "game breaking" where you don't have to roll because you've got a trick that bypasses the problem.

What level of competence do we expect classes to attain?

Competence to me is 10+ on a d 20 is a success. 5+ is superior ability, 15+ is inferior. 1+ is makes the challenge irrelevant, and 20+ is just incompetent/no ability.

[Hurnn]

I am not sure the tier system as a whole needs to be reworked. However it definitely has some room for improvement.

Tier 1 classes have no need of a nuanced ranking and really neither do tier 2 they tell the rules to shut up and sit down and can do what they want. A wizard being 1x and a druid being 1y is mostly irrelevant.

Tiers 3, 4 , and 5 are where we can use some nuance but that is as simple as adding rankings A, B, and maybe C to them. A would be excels at their thing and is ok at some others Barbarian as an example would be 4A they are great at smashing faces. B would be ok at a lot of things but not great at any of them Ranger would be 4B.

[SirNibbles]

I think you should identify all the individual roles that exist- both in and out of combat- so that you can see which classes are able to fill those roles, how often, and how well.

My lists are not exhaustive.

Out of Combat Roles/Abilities:
Using Skills to make money
Crafting (Magic) Items
Gathering Information
Making NPC allies
Talking your way out of trouble
Avoiding Enemies
Avoiding Traps
Disabling Traps
Travelling Abilities


In-Combat Roles/Abilities:
Buffing (Self and Party)
Healing (Self and Party)
Single Target Melee Damage
Multi-Target Melee Damage
Single Target Ranged Damage
Multi-Target Ranged Damage
Debuffing
Field Control
Movement/Versatility
Running Away



Edited to add Travel/Movement.

[Pleh]

You forgot travel abilities, sir nibbles

[OldTrees1]

I'd like to see at least two axes: Versatility on one side, Power on the other.

Versatility:
A - Can whip out whatever set of abilities is needed to solve the current problem.
B - Has a flexible set of powers on demand which are applicable to a variety of different situations.
C - Has an inflexible set of powers, but it's a pretty large set.
D - Has an inflexible and small set of capabilities.
E - Has one or two different capabilities.
F - Has no special abilities.

Power:
1 - Can use raw power to push through nearly any situation, including those significantly above the character's level.
2 - Can use raw power to push through some situations above the character's level.
3 - Can use raw power to push through nearly any situation appropriate for the character's level.
4 - Cannot use raw power to push through many situations appropriate for the character's level; requires consumable resources or cleverness to handle some level-appropriate situations.
5 - Cannot use raw power to push through most level-appropriate situations. Cannot contribute meaningfully to some level-appropriate situations.
6 - Cannot contribute meaningfully to the majority of level-appropriate situations.


... then a Commoner might be F6, while a Wizard is A1, and other classes will fall somewhere in between.


Second, I think ranking versatility is sort of dumb. Having a bunch of tools that are almost good enough is massively worse than having even one tool that is good enough. Insofar as it means anything useful, it's just another term for "able to solve problems" which is the same thing as power.



Character ability is much more complicated than Nifft's system can capture and versatility, as I will explain later, has two values that Cosi highlights and overlooks with their criticism. I will address both of these in this post.


Parties(multiple characters) can be faced with countless varieties of challenges. In each of those challenges an individual character's ability to contribute is some real number(positive or negative). While negative values are possible, I am going to ignore them for simplicity. This ability to contribute can be near 0 (basically the character is sidelined or otherwise negated for the challenge), low (able to provide low skill contributions), high (able to provide high skill contributions), total (able to handle the challenge on their own), or excessive (their mere presence makes the challenge a waste of the player's time).

Combining those two variables we can see that each class is a bell like curve with various regions. It has a central region where it is the highest. This region's width is the range of their competency and the height is the magnitude of their competency (which can be excessive). Next we may have a region with lower but not near 0 magnitude. This region's width is the range of their ability to contribute and the height is the magnitude of their contributions. Finally we may have a region with near 0 magnitude. This region's width is the range of the times the character is unable to contribute.

Even in the simplified form that I describe, it would take no less than 4 variables to describe(height and range of the 1st two categories implies the range of the 3rd, ~0 height, category).

Now why is the versatility relevant? Higher Versatility is the layman's colloquial term for decreased range of the 3rd category, although it occasionally is used as the range of the 1st category. Cosi is right that there is a difference between competence and ability to merely contribute. However it is also important to measure the range of inability to contribute.

So no, a 2 variable ranking would be insufficient to the point of being hopelessly inaccurate in its description of capability. However I will admit that listing a bunch of curves is way to over complicated for such a subjective matter. Instead I will highlight some values we might find very important:

1) We want to know the height of their capability. This marks both what kind of campaigns they can participate in as well as what kinds of campaigns they can break.
2) We want to know the breadth of their capability. How often can they lead a contribution rather than merely assist.
3) We want to know the breadth of their ability to contribute. How often is their participation of negligible worth or worse?
4) Rather than list the magnitude of their ability to contribute when they are merely assisting, we can set a standard for what we expect that level of contribution to be. Thus the information stored by such a value is instead stored by the range of that area.

There, with 3 variables and 1 constant we can have a better description that might have some real merit.

[javcs]

I think you should identify all the individual roles that exist- both in and out of combat- so that you can see which classes are able to fill those roles, how often, and how well.

My lists are not exhaustive.

Out of Combat Roles:
Using Skills to make money
Crafting (Magic) Items
Gathering Information
Making NPC allies
Talking your way out of trouble
Avoiding Enemies
Avoiding Traps
Disabling Traps

In-Combat Roles:
Buffing (Self and Party)
Healing (Self and Party)
Single Target Melee Damage
Multi-Target Melee Damage
Single Target Ranged Damage
Multi-Target Ranged Damage
Debuffing
Field Control
Running Away

That's basically just redoing the Niche system. And while that's a perfectly fine system, it's also a lot more in depth, and it's a lot harder skim it quickly.


The Tier system, as laid out by JaronK, is easy to skim, and fairly straightforward, for the most part.



I think what's really being looked for is a rating system that is somewhat more granular and descriptive than the existing Tier system, but still fairly easy to understand at a glance, which means less granular than the Niche system.
Nifft's proposal might or might not be granular enough, but simplicity is important for anything to get traction and wider acceptance in the community.

[Pleh]

Yes, a new tier system will need a tight control on its focus.

If you want perfect accuracy in your ranking, it will become a library of handbooks.

JaronK became famous for something simple enough to grasp at a quick glance. Critics weren't satisfied with the accuracy, so it needs to focus to get dialed in.

We're dealing with the age old question of acceptable tolerances. Too much accuracy is too much information to be useful (like the 10000th decimal of pi). Too little accuracy is unhelpful and misleading.

We need a goldilocks approximation that gives us the information we need quickly without being so quick we get thrown off the trail.

This is why I was arguing earlier that we need to establish who our audience really is and what purpose the tier is really serving. There is a range of acceptable possible tiers. To understand how to best build the tier, we have to figure out what problems we want to solve with it.

[Lans]

How should we use feats and items in this? Some classes can afford to take things that can expand there repertoire with out comprising what they do as a class, and some can't.

Take a fighter and a whirling frenzy barbarian as examples. The fighter has to use all his resources to remain relevent in combat, while the barbarian could potentially afford to take shape soulmeld or bind vestige feats.

[ChaosStar]

I personally favor a system of five to six points with each point ranked from 1-5 or 1-10 then arranged in an easy to glance at circular graph.

http://i.imgur.com/TEysh9X.jpg

[SirNibbles]

That's basically just redoing the Niche system. And while that's a perfectly fine system, it's also a lot more in depth, and it's a lot harder skim it quickly.


The Tier system, as laid out by JaronK, is easy to skim, and fairly straightforward, for the most part.



I think what's really being looked for is a rating system that is somewhat more granular and descriptive than the existing Tier system, but still fairly easy to understand at a glance, which means less granular than the Niche system.
Nifft's proposal might or might not be granular enough, but simplicity is important for anything to get traction and wider acceptance in the community.

My idea was to work with Niff's idea to Tier each class based on:
-how many roles it can fill
-how well it fills each role
(Power)
and

-in how many situations it can fill that role (to the same extent)
(Versatility)

For example, let's look at a Rogue (a pure Rogue, for simplicity).




Out of Combat Roles/Abilities:
Using Skills to make money
Crafting (Magic) Items
Gathering Information
Making NPC allies
Talking your way out of trouble
Avoiding Enemies
Avoiding Traps
Disabling Traps
Travelling Abilities

In-Combat Roles/Abilities:
Buffing (Self and Party)
Healing (Self and Party)
Single Target Melee Damage
Multi-Target Melee Damage
Single Target Ranged Damage
Multi-Target Ranged Damage
Debuffing
Field Control
Movement/Versatility
Running Away



It has a variety of non-combat roles which it can do, some to a smaller extent than others, but its skill points mean it can do them well enough.

In combat, it can mainly damage a single enemy at a time, either with melee or ranged attacks. It has very minor debuffing abilities. Just based on the roles it can fill and how well it can do them, it would probably be a Tier 3 class (in my opinion). However, it has poor combat versatility. Its abilities depend on the enemy being vulnerable to sneak damage and other special Rogue attacks. There are so many enemies that are completely immune to this, in which case you end up with a 3/4 BAB Fighter but with fewer feats and Evasion. Its low versatility in combat would rank at D.

So, it's a Tier 3D class (medium power but easily countered).

Compare that to a Wizard who can do a lot of things well and can do those things well pretty much all the time (1A). If the same Wizard could only cast spells when the sun was up, his versatility would be very poor, probably 1D or 1E.

[javcs]

How should we use feats and items in this? Some classes can afford to take things that can expand there repertoire with out comprising what they do as a class, and some can't.

Take a fighter and a whirling frenzy barbarian as examples. The fighter has to use all his resources to remain relevent in combat, while the barbarian could potentially afford to take shape soulmeld or bind vestige feats.

That's why barbarian would get categorized differently from fighter.

Item dependency is a factor that would be taken into account.
So too is how many build resources can be spent on broadening one's versatility before loosing focus when it matters.

Fighters are a ball hammer. Barbarians are a claw hammer.
Rogues are a multi-tool.
Factotums are a toolbelt.
Beguilers (and the other focused/specialist casters) are electrical toolkits. Barring higher op shenanigans (Rainbow Servant) to massively expand their spells known.
Sorcerers are a toolbox.
Wizards are an entire machine and electrical shop.

[D&DPrinceTandem]

Tier 1: Can break the game in a new way every day.
Tier 2: Can break the game, but only in one or two ways.
Tier 3: Power and minor utility, or major utility.
Tier 4: Power but no utility, or minor utility.
Tier 5: Sub-par in its field, or unfocused.
Tier 6: Worthless.

Tier 7: Truenamer

[Aimeryan]

Really liking the look of Personman's Niche system and similarly SirNibbles'. The Niche system does come up with some oddities, however, even though I don't dispute the rankings Personman has given.

Take Barbarian, total points 60 - lower is better, and 60 is really really high, so very bad. For the most part this is because the only thing they actually do is deal melee damage and take damage in his chart. A Monk gets 55 total. The reason for this is because it gets a lot of 3's in the Niche system, being able to do a number of things somewhat slightly. This is an issue because the Monk being a poorman's Jack Of All Trades (like, really poor) is not even close to the usefulness and the overall ranking I would give to the Barbarian's Master Of One.

Therefore, I find it to be an issue of only ranking 1-4; getting a 1 or 2 in something is not far enough ahead of getting a 3 in something. I would probably use 0-10 instead, with 10 being exemplar in that field and 0 being completely oblivious to that field even existing. This is the opposite direction to Personman's Niche system, so high is good and low is bad; I find it easier to evaluate in that direction - but I'm not wedded to it.

Furthermore, I find the Niche system to be missing a rating for how much resources a class has to use (out of its total resources) in order to fulfil its potential. For example, many mundanes have to use a lot of their Feats, Skills, and/or WBL to fulfil their potential in even one category - which leaves them unable to be versatile. Where as, to be an accomplished melee damager a Wizard may only need to use one spell. So I would add a Resourceful category.

There is also an issue with ratings changing over character levels, but the only way to solve that would be to include a rating for several points of the game - say low (1-7), mid (8-14), high (15-20).

Lastly, Personman's table needs to have regular headers!


I'll make an attempt at some categories myself (with judicious recycling of what others have come up with!).

Rate 0-10 for each category, total at the end. 0 is non-existent, 10 is exemplar. Do this three times, denoting L, M, and H respectively for Low (~1-7), Mid (~8-14), and High (~15-20) character levels. This is to be relative to what is effective at the time; if you are a Tripper build (Manipulator) you may start off fairly effective, reaching a peak just at the end of low levels: 6L; fall behind as enemies become more resistant and care less about being tripped at mid levels: 3M; become almost completely ineffective at high levels: 1H.

1. Resourceful
Making efficient use of the party's WBL - crafting, using skills to make money, spells to buff the party rather than using items, etc.
2. Party Face
Interacting with and influencing NPCs - including by use of spells to dominate/charm, as well as skills, etc.
3. Ninja
Ability to remain unnoticed - including the art of theft (or, indeed, the theft of art).
4. Detective
Ability to gather information - for example, from the environment (searching, trapfinding, etc.) or from the locals (knowledge skills, etc.).
5. Pathchanger
Ability to manipulate or bypass the environment - including making walls, picking locks, teleportation, etc. Combat and out of combat use here.
6. Elusive
Ability to avoid danger - being at range tends to help quite a lot, as do contingencies and immediate actions.
7. Tank
Ability to absorb punishment - should give some thought to Saves as well as AC, DR, HP.
8. Manipulator
Ability to change how the combat plays out - including battlefield control, buffing, debuffing, healing, dominating etc. High rating here should show great variability, effectiveness, and reliability.
9. Damage Dealer
Ability to inflict damage - include ability damage as well as HP damage. High rating here should show great damage potential and great reliability to apply that damage.
10. Versatile
Being able to effectively fulfil multiple roles at the same time - this means still having resources (spells, skills, feats, wealth, etc.) left over after becoming effective at one thing, so that they can become effective at something else as well. High rating here should indicate being able to fully take on many roles to the potential indicated in by those categories.


Example (numbers to be much refined):



Class
Resourceful
Party Face
Ninja
Detective
Pathchanger
Elusive
Tank
Manipulator
Damage Dealer
Versatile
Total


Psychic Warrior
1L / 3M / 5H
0L / 0M / 0H
0L / 1M / 1H
0L / 1M / 2H
4L / 4M / 6H
6L / 4M / 4H
2L / 4M / 4H
7L / 4M / 3H
7L / 6M / 4H
2L / 3M / 3H
29L / 30M / 32H



These numbers are with some optimisation only; having maybe looked at a handbook and decided on a build. I feel this is the most useful point for a tier system for DMs and Players who are not already themselves quite familiar with the classes and how they play in campaigns.

In particular for the example Psychic Warrior, this is looking at having perhaps taken the Mantled Warrior ACF, but only choosing one of the obvious ones that fits what they want to do; so say Creation if thinking about being a crafter; Freedom if looking at being Elusive (and probably as a default choice); etc. It wont involve any particularly high optimisation to do with Mantles.

[bekeleven]

There's also my tier rankings by optimization level (http://www.giantitp.com/forums/showthread.php?459141).

I think at this point we need a tier system for tiering systems.

[Pleh]

Really liking the look of Personman's Niche system and similarly SirNibbles'. The Niche system does come up with some oddities, however, even though I don't dispute the rankings Personman has given.

Take Barbarian, total points 60 - lower is better, and 60 is really really high, so very bad. For the most part this is because the only thing they actually do is deal melee damage and take damage in his chart. A Monk gets 55 total. The reason for this is because it gets a lot of 3's in the Niche system, being able to do a number of things somewhat slightly. This is an issue because the Monk being a poorman's Jack Of All Trades (like, really poor) is not even close to the usefulness and the overall ranking I would give to the Barbarian's Master Of One.

Therefore, I find it to be an issue of only ranking 1-4; getting a 1 or 2 in something is not far enough ahead of getting a 3 in something. I would probably use 0-10 instead, with 10 being exemplar in that field and 0 being completely oblivious to that field even existing. This is the opposite direction to Personman's Niche system, so high is good and low is bad; I find it easier to evaluate in that direction - but I'm not wedded to it.

Furthermore, I find the Niche system to be missing a rating for how much resources a class has to use (out of its total resources) in order to fulfil its potential. For example, many mundanes have to use a lot of their Feats, Skills, and/or WBL to fulfil their potential in even one category - which leaves them unable to be versatile. Where as, to be an accomplished melee damager a Wizard may only need to use one spell. So I would add a Resourceful category.

There is also an issue with ratings changing over character levels, but the only way to solve that would be to include a rating for several points of the game - say low (1-7), mid (8-14), high (15-20).

Lastly, Personman's table needs to have regular headers!


I'll make an attempt at some categories myself (with judicious recycling of what others have come up with!).

Rate 0-10 for each category, total at the end. 0 is non-existent, 10 is exemplar. Do this three times, denoting L, M, and H respectively for Low (~1-7), Mid (~8-14), and High (~15-20) character levels. This is to be relative to what is effective at the time; if you are a Tripper build (Manipulator) you may start off fairly effective, reaching a peak just at the end of low levels: 6L; fall behind as enemies become more resistant and care less about being tripped at mid levels: 3M; become almost completely ineffective at high levels: 1H.

1. Crafter
Making efficient use of the party's WBL - including the use of skills to make money.
2. Party Face
Interacting with and influencing NPCs - including by use of spells to dominate/charm, as well as skills, etc.
3. Ninja
Ability to remain unnoticed - including the art of theft (or, indeed, the theft of art).
4. Detective
Ability to gather information - for example, from the environment (searching, trapfinding, etc.) or from the locals (knowledge skills, etc.).
5. Pathchanger
Ability to manipulate or bypass the environment - including making walls, picking locks, teleportation, etc. Combat and out of combat use here.
6. Elusive
Ability to avoid danger - being at range tends to help quite a lot, as do contingencies and immediate actions.
7. Tank
Ability to absorb punishment - should give some thought to Saves as well as AC, DR, HP.
8. Manipulator
Ability to change how the combat plays out - including battlefield control, buffing, debuffing, healing, dominating etc. High rating here should show great variability, effectiveness, and reliability.
9. Damage Dealer
Ability to inflict damage - include ability damage as well as HP damage. High rating here should show great damage potential and great reliability to apply that damage.
10. Resourceful
Having and effectively using their resources to be able do multiple things - resources include skills, spells, feats, etc.


Example (numbers to be much refined):



Class
Crafter
Party Face
Ninja
Detective
Pathchanger
Elusive
Tank
Manipulator
Damage Dealer
Resourceful
Total


Psychic Warrior
1L / 3M / 5H
0L / 0M / 0H
0L / 1M / 1H
0L / 1M / 2H
4L / 4M / 6H
6L / 4M / 4H
2L / 4M / 4H
7L / 4M / 3H
7L / 6M / 4H
3L / 4M / 5H
30L / 31M / 33H



These numbers are with some optimisation only; having maybe looked at a handbook and decided on a build. I feel this is the most useful point for a tier system for DMs and Players who are not already themselves quite familiar with the classes and how they play in campaigns.

I tried for a bit, but my eyes glazed over pretty quick. And I'm not just talking about how the first half explains what you're doing. The demonstration half is pretty heavy reading considering we want this to be a quick reference.

If we're looking to capture that simple staying power JaronK had, it's got to communicate its message faster and more efficiently than this.

[Aimeryan]

I tried for a bit, but my eyes glazed over pretty quick. And I'm not just talking about how the first half explains what you're doing. The demonstration half is pretty heavy reading considering we want this to be a quick reference.

If we're looking to capture that simple staying power JaronK had, it's got to communicate its message faster and more efficiently than this.

I would say the only hard reading part of it is the xL / yM / zH - but I feel this information is valuable. I guess we could just have a separate table for each:

Low levels (~1-7):



Class
Crafter
Party Face
Ninja
Detective
Pathchanger
Elusive
Tank
Manipulator
Damage
Resourceful
Total


Psychic Warrior
1
0
0
0
4
6
2
7
7
3
30



Medium Level (~8-14):



Class
Crafter
Party Face
Ninja
Detective
Pathchanger
Elusive
Tank
Manipulator
Damage
Resourceful
Total


Psychic Warrior
3
0
1
1
4
4
4
4
6
4
31



High Level (~15-20):



Class
Crafter
Party Face
Ninja
Detective
Pathchanger
Elusive
Tank
Manipulator
Damage
Resourceful
Total


Psychic Warrior
5
0
1
2
6
4
4
3
4
4
33





There's also my tier rankings by optimization level (http://www.giantitp.com/forums/showthread.php?459141).

I think at this point we need a tier system for tiering systems.

Just read through it, quite enjoyed the read! I agree that optimisation is another point to look at in a tiering system.

[Florian]

@Cosi:

My problem with the Tier system as established by JaronK always was, that there was no baseline assumption on what performance a character (solo/team) should have and what the expectations to this actually are, so the terms "breaking the game" or "do a thing well" are pretty much meaningless.

I mainly gm Paizo APs, so I've got a very good idea of whatis necessary and what does never come up at all, so I've been faced with characters that can do stuff that is never going to be used.

When edoing the Tiers, I think it is necassary to look at the baseline erformance, as well as overperformance and rate that, too.

[Pleh]

@Cosi:

My problem with the Tier system as established by JaronK always was, that there was no baseline assumption on what performance a character (solo/team) should have and what the expectations to this actually are, so the terms "breaking the game" or "do a thing well" are pretty much meaningless.

I mainly gm Paizo APs, so I've got a very good idea of whatis necessary and what does never come up at all, so I've been faced with characters that can do stuff that is never going to be used.

When edoing the Tiers, I think it is necassary to look at the baseline erformance, as well as overperformance and rate that, too.

This is a good point. Where is the line between competent Heroes and game breaking? Where is the line between game breaking and game changing?

I think it might be more helpful (in terms of tier theory) to think of it as power in relation to the DM's toolbox. If the only tools a DM has to beat a character is copying the PC and fighting fire with fire, or to use Rule 0 and kill him with Rocks Fall, then the game is broken because the DM has lost control to the mechanics. The DM is supposed to play and have fun, too. When teleportation kills the DM’S plans to set his Heroes on a world spanning trek, it's a form of game breaking (probably better viewed as game changing because in that case, a different game with its own balances can still be played).

[exelsisxax]

Aimeryan,

the main drawback is the level-dependent rankings. It's a lot of extra work and complication that, in the vast majority of situations, won't provide much real information. I don't see why it's at all necessary. It might be useful as accessory information, but that's better placed in a handbook where it can be fully explained. Not really "quick overview" sort of material.

It should, if properly thought out, be simply better in every respect than the old conflating tiers. The total score would be a more appropriate proxy than tiers for the noobs(they couldn't build a T1 cleric anyway) and the individual niches/roles provide more meaningful information to anyone with either experience or a goal in mind.

I'd try to collapse the niches though. For one, i see no reason at all why "crafter" is separate from "resourceful"

[Cosi]

Although, I'd consider a few other axes, even though that would make rating classes more complicated.

If you are ranking more than two things, Nifft's 1 - 6/A - F ranking becomes immediately unacceptable, because what are you using for your third ranking? Roman numerals? You need a single set of grades (I like S/A/B/C/F) that you apply to several categories, rather than a different set of equivilent rankings for each category.


For the Versatility Axis, I'd also suggest another value or chain of values - "flexibility in selecting powers/abilities that are then permanently locked in once selected"; on the strong end of this subchain would be Sorcerer, Favored Soul, and also probably Psion and the other psionic full manifesters - and then at the low end of the subchain would be Fighter.

What? "Here are a bunch of spells, select a small number" is exactly the same as "here are a bunch of feats, select a small number". Also, options you don't have don't make your character any better.


On the Power Axis, I'd suggest make it explicit that the Power value is for areas where their abilities/versatility can be applied, and possibly give a second rating for their power when their primary abilities/versatility are unusable or hard to make applicable.

Power and versatility aren't meaningfully different. They both simply measure one's ability to solve problems. Having a lot of useless abilities doesn't mean anything important for evaluating a character.


Maybe an axis that measures how well it usually PRCs, but a base class should not be assumed to PRC for anything other than a specific to PRCing rating.

It seems to me that PrCs are a very solvable problem if you have different sets of rankings by level. If a PrC merits its own ranking (e.g. Shadowcraft Mage, Gish PrCs), create a separate entry for it at relevant levels. So you might have something like this:

Shadowcraft Mage
Something A
Another Thing B
A Third Thing A

<Explanation>

Then for PrCs where base class is meaningful (e.g. Warmage/Rainbow Servant, Beguiler/Rainbow Servant, Ur-Priest/Mystic Theurge) rank that specific combination.


I think people here have different definitions of competency for characters.

Competency is fairly simple for combat. There are a bunch of monsters, and combat is about defeating those monsters. You are competent at combat if you can defeat level appropriate monsters at an appropriate rate. You are overpowered if you are better than that, and underpowered if you are worse.

Non-combat is harder to judge, because there isn't a canonical list of non-combat encounters. plane shift is a useful non-combat ability, but there's no way to determine (from the rules) whether you're supposed to get it at 9th (like a Cleric), 13th (like a Wizard), or never (like a Fighter).


Tier 1 classes have no need of a nuanced ranking and really neither do tier 2 they tell the rules to shut up and sit down and can do what they want. A wizard being 1x and a druid being 1y is mostly irrelevant.

I don't think you're taking a sufficiently outside view for this. Nuance may not be necessary with rankings heavily based of breaking the game (such as JaronK's), but those rankings are not particularly helpful. They aren't helpful to DMs because they don't explain what abilities are broken. They aren't helpful for players because players mostly don't use "how broken is this" as a selection heuristic. They aren't helpful overall because they don't look at what happens when people aren't allowed to break the game.


In-Combat Roles/Abilities:
Buffing (Self and Party)
Healing (Self and Party)
Single Target Melee Damage
Multi-Target Melee Damage
Single Target Ranged Damage
Multi-Target Ranged Damage
Debuffing
Field Control
Movement/Versatility
Running Away

I think this is a bad way of looking at things, because it doesn't reflect the game. You don't fight an encounter with "you need AoE damage". You fight an encounter with a bunch of monsters.


How should we use feats and items in this? Some classes can afford to take things that can expand there repertoire with out comprising what they do as a class, and some can't.

Feats and items should, IMHO, be evaluated. There's a trade-off between accuracy and usability, and I think feats and items are pretty clearly on the useful side of that trade-off.


Really liking the look of Personman's Niche system and similarly SirNibbles'. The Niche system does come up with some oddities, however, even though I don't dispute the rankings Personman has given.

I dislike niche ranking type systems because they mostly end up being complicated to evaluate (by virtue of ranking like 30 things) or not terribly useful (combining scores on tasks of varying value). Being really good at running away is not nearly as valuable as being really good at defeating enemies.

[Aimeryan]

Aimeryan,

the main drawback is the level-dependent rankings. It's a lot of extra work and complication that, in the vast majority of situations, won't provide much real information. I don't see why it's at all necessary. It might be useful as accessory information, but that's better placed in a handbook where it can be fully explained. Not really "quick overview" sort of material.

It should, if properly thought out, be simply better in every respect than the old conflating tiers. The total score would be a more appropriate proxy than tiers for the noobs(they couldn't build a T1 cleric anyway) and the individual niches/roles provide more meaningful information to anyone with either experience or a goal in mind.

I'd try to collapse the niches though. For one, i see no reason at all why "crafter" is separate from "resourceful"

I may not have adequately shown in that example (Psychic Warrior) why the Low, Mid, High levels matter, although it did for the Manipulator category where the uses of things like Tripping tend to fall off. The totals by chance happened to end up balancing out - some stuff went down, some went up. Still, perhaps a bad example for this.

I do feel for a number of classes that the level can make a vast difference, but I could see that this could be dropped from consideration. The question would then remain as to what level should be used - perhaps level 20, although the game is mostly over then (unless you are going Epic, which is a whole different discussion). Perhaps level 10, being half way through?

I tried to collapse the niches as I could, while retaining some adequate differentiation, but its merely a first pass and could change. Resourceful is meant to be a measure of versatility while the character is in play (so not a measure of possible different builds, which I don't feel warrants being rated). I'm not stuck on the name - it could change! Maybe just Versatility.

As for Crafter, hmm, I felt it sufficiently game-changing enough to be included since a lot of power for some classes is derived from their wealth and items. You may be right that Resourceful would make a better name for this category. Oh, it also made the number of categories equal 10, which made for a nice 0-100 scale.

[I think I'll make those name changes]


I dislike niche ranking type systems because they mostly end up being complicated to evaluate (by virtue of ranking like 30 things) or not terribly useful (combining scores on tasks of varying value). Being really good at running away is not nearly as valuable as being really good at defeating enemies.

I hear what you are saying; thirty categories would be overload, indeed. However, I don't think they are more complicated to evaluate, just more thorough. I wouldn't say the total score at the end is the most important part of them, either. If anything a tier system is worse for that - without a breakdown you can't separate out the parts that you consider important.

[Pleh]

@Cosi. Access to spells and access to feats is immensely different. When you take spells, you can change your selection tomorrow. When you take feats, you can change them next month if you're lucky. A spellcaster can equip themselves daily. A feat monger is pretty much stuck with their choices.

This actually is the perfect demonstration of power vs versatility. A fighter with an optimized feat selection might be able to keep up with the wizard for a half a dozen levels (of they're lucky). Power is comparable at those levels. But the Fighter would have to undergo an intense process to change his feats to solve a different scenario. The wizard of the same power level just prepares a different set of spells and loses no power

[exelsisxax]

I may not have adequately shown in that example (Psychic Warrior) why the Low, Mid, High levels matter, although it did for the Manipulator category where the uses of things like Tripping tend to fall off. The totals by chance happened to end up balancing out - some stuff went down, some went up. Still, perhaps a bad example for this.

I do feel for a number of classes that the level can make a vast difference, but I could see that this could be dropped from consideration. The question would then remain as to what level should be used - perhaps level 20, although the game is mostly over then (unless you are going Epic, which is a whole different discussion). Perhaps level 10, being half way through?

I tried to collapse the niches as I could, while retaining some adequate differentiation, but its merely a first pass and could change. Resourceful is meant to be a measure of versatility while the character is in play (so not a measure of possible different builds, which I don't feel warrants being rated). I'm not stuck on the name - it could change! Maybe just Versatility.

As for Crafter, hmm, I felt it sufficiently game-changing enough to be included since a lot of power for some classes is derived from their wealth and items. You may be right that Resourceful would make a better name for this category. Oh, it also made the number of categories equal 10, which made for a nice 0-100 scale.

[I think I'll make those name changes]


I think that there should not be a baseline level at all. If a class has the best healing in the game, but only gets it at level 10, it cannot be a healer 10. If a different class can cast CMW as an SLA from level 1 but that's it, they'll be an amazing early healer but still not good overall. High-scoring categories are those which the class can perform at that proficiency against appropriately escalating challenges with increasing level. Think of it a bit like class feature options: it's irrational to pretend a class always does or always doesn't have it, because it is simply true that some have and others don't. Going in-depth into the fact that tripping sucks at high level and crafting doesn't do anything until you hit a certain WBL is for handbook, not a quick class power ranking overview.

The clear definition of different rankings are indeed super important. But from what perspective should they be defined? Personally, I don't think "tank" should be a niche. Rename "elusive" to "survivability", put all the toughness from tank into it. Rename "manipulator" to BFC, put all the BFC from tank into it. Tank is gone. Then put the sneaky stuff from elusive into ninja.

Slightly simpler, but is it more clear and easy to understand, or more accurately form delineations between various niche uses? The categorization pushes the ranking somewhat, but less than normal tiers.

[A_S]

Tier 7: Truenamer
This is often joked about, but not true at all. Truenamer is actually a great showcase of the limitations of the tier system as originally envisioned:

At a medium-high level of optimization, it's comparable to the more versatile tier 4 classes like Rogue and Warlock for most of its career.
At lower optimization levels, it can't make its Truespeak checks and drops like a rock to tier 6.
It suddenly becomes tier 1 at level 20.
Several of its mechanics are broken (in the "don't work as written" sense, not the "too good" sense), and require houserules or subjective interpretation to do anything.

What tier are you supposed to assign something whose actual effectiveness in play is going to vary depending on all those things?

-----

The tier system attempts to assign a single number to effectiveness-in-play. In doing so, it glosses over a lot of things that can impact that effectiveness:

The power-vs.-versatility distinction
Degree of optimization, including the ideas of optimization floor/ceiling
Splatbook dependence
The impact of character level

This isn't really a bug of the tier system, it's an inevitable consequence of building a single composite variable out of several related-but-different measures of effectiveness. If you want more detail, you can break up your single variable into multiple dimensions (e.g., Nifft's "1A/6F" system above, or RFLS' In-Depth Tier System (http://www.giantitp.com/forums/showthread.php?308514-3-5-Tier-System-This-Time-with-Levels-and-Optimization)).

But you pay for it in clarity of communication. The shorter and simpler a measure is, the more people are willing to actually use it as a helpful shorthand. Everybody knows (to the limits of the precision of the system) what "tier 1" means. Does anybody refer to RFLS' divided-by-optimization-and-character-level when they're talking about character effectiveness?

[Florian]

This is a good point. Where is the line between competent Heroes and game breaking? Where is the line between game breaking and game changing?

I think it might be more helpful (in terms of tier theory) to think of it as power in relation to the DM's toolbox. If the only tools a DM has to beat a character is copying the PC and fighting fire with fire, or to use Rule 0 and kill him with Rocks Fall, then the game is broken because the DM has lost control to the mechanics. The DM is supposed to play and have fun, too. When teleportation kills the DM’S plans to set his Heroes on a world spanning trek, it's a form of game breaking (probably better viewed as game changing because in that case, a different game with its own balances can still be played).

I once did the work to analyze some 7 APs (and some modules) for each skill, skill check, item or spell necessary to solve each an every encounter or riddle with a brute force mechanical approach and score as high as possible using as few ressources as I could while still getting the maximum result possible.

That proved to be interesting, as it drastically shifted the apparent worth of some things around, as well as putting a heavy emphasis on the thing being availlable at the right time to actually be a "counter"/"solution" the moment it actually is a challenge and useful at solving it.

The really interesting part hereis, that each and every Encounter is designed to work with the optimal solution but gives hints to what happens as a second or third best solution.
It does not pose the question of what happens when a character can fly or planeshift, it poses the question if a character can fly or planeshift when expected to do to solve an encounter.

[Cosi]

II hear what you are saying; thirty categories would be overload, indeed. However, I don't think they are more complicated to evaluate, just more thorough. I wouldn't say the total score at the end is the most important part of them, either. If anything a tier system is worse for that - without a breakdown you can't separate out the parts that you consider important.

They're more complicated to evaluate, and they don't provide enough additional, useful information to justify the added complexity. You don't need an entire ranking system to explain that Wizards are good at BFC and Clerics are good at melee. That can, and should, go in the justifications section of the ranking, because you can add enough detail there to be useful.


@Cosi. Access to spells and access to feats is immensely different. When you take spells, you can change your selection tomorrow. When you take feats, you can change them next month if you're lucky. A spellcaster can equip themselves daily. A feat monger is pretty much stuck with their choices.

The poster I replied to was talking specifically about classes like Sorcerer that lock in their spells at selection.


This actually is the perfect demonstration of power vs versatility. A fighter with an optimized feat selection might be able to keep up with the wizard for a half a dozen levels (of they're lucky). Power is comparable at those levels. But the Fighter would have to undergo an intense process to change his feats to solve a different scenario. The wizard of the same power level just prepares a different set of spells and loses no power

lolwut? How is "comparable at low levels, but falls behind at high levels" a versatility issue?


I think that there should not be a baseline level at all. If a class has the best healing in the game, but only gets it at level 10, it cannot be a healer 10.

Yes, it can - in the rankings for 10th level and higher. If you don't have separate rankings by level, you get stupid arguments about whether the Healer is good because it gets gate at 17th or how to rank the Warblade, which is a valuable member of a 1st level party but largely useless to a 15th level party. Plus, having different levels for ranking solves the PrC problem handily. Just rank PrCs at the levels where they matter.


What tier are you supposed to assign something whose actual effectiveness in play is going to vary depending on all those things?

What you should do is hold power constant, then rank how difficult it is to get to that power level. So instead of arguing whether we should rank the noob Truenamer, the source-limited Truenamer, or the expert Truenamer, we should just look at "whatever Truenamer hits X level of power". Then it gets a ranking for how complicated that build is, and possibly what kind of sources that build requires.

[Florian]

lolwut? How is "comparable at low levels, but falls behind at high levels" a versatility issue?


Funny, as is the rest of your answers in your post, because you actually don't seem to answer anything, besides clearly having a certain picture in your mind what a class should be able to do or how a combat situation works out while no specifics have been mentioned.

[exelsisxax]

Cosi, trying to rank classes and PrC at exit levels is even worse than trying to rank them with 3 different level ranges. Different builds require different dip extents and stays in base classes at the same time. This is straight up "handbook of everything" levels of complexity and overload. It's not a reasonable method for evaluating the class itself.

[Cosi]

Funny, as is the rest of your answers in your post, because you actually don't seem to answer anything, besides clearly having a certain picture in your mind what a class should be able to do or how a combat situation works out while no specifics have been mentioned.

I don't know what you're saying. I haven't responded to anything, but I've clearly conveyed a vision, without mention anything specific? That's like three different contradictory statements, none of which particularly reflect my post's substance.

[Cosi]

Cosi, trying to rank classes and PrC at exit levels is even worse than trying to rank them with 3 different level ranges. Different builds require different dip extents and stays in base classes at the same time. This is straight up "handbook of everything" levels of complexity and overload. It's not a reasonable method for evaluating the class itself.

I don't know where I suggested ranking things at exit levels. Rank things at 5/10/15. Rank PrCs generally, unless there are specific cases that are importantly different (e.g. one entry for Sorcerer/Incantatrix and Wizard/Incantatrix, separate entries for Warmage/Rainbow Servant and Sorcerer/Rainbow Servant). There's more work that you can do, but you don't particularly have to do all of it. Doing it for all base classes is not more work than something like JaronK's system, it's more precise, it's easier to use, and it's less contentious. Then you can expand it if you want without dysfunctional things like the Tier System for PrCs.

[SaintRidley]

It's the classes what need revision rather than the tiers. Revising tiers without touching the classes won't do anything other than soothe egos.

[tsj]

It might be useful to have a list of types of wizard spells and what rolls they can handle/cover..
Since wizard can do everything the All rolls should be covered this way

It could be useful for Class fixes and tier could be Per level depending on stuff like nr of roles filled plus maybe other stuff

Ie.. Damage spells can fill the DPS role and summon spells fill the summon role... BFC spells for bfc etc

[Aimeryan]

I think what I want is being seen as too complicated by some people. If the tiering system was to be done by those people then my suggestions are going to be useless, but if they are done by others they might be quite useful.

"More complicated" is kind of wrong, kind of right, too. If you split white light up into a spectrum, say with a prism, is that light now more complicated? Not really, it is just you were ignoring the complexity before, which just makes you ignorant of what is going on. In some cases that might be enough, in others it pays to look a little more closely.

Another analogy may be that of a car; it seems pretty simple from the outside, but once you open the bonnet you see the complexity that actually exists. If you want to evaluate a car more accurately it helps to understand and measure that complexity more strongly. My suggestion is like having a mechanic look at your car; sure that is costly, but if you want something to be done well you might just have to pay up. What some people here seem to want is to just have your next-door-neighbour who only knows the model and general performance of that model to look at it and tell you what he thinks; it is better than nothing, but you may end up worse of for it.

[A_S]

What you should do is hold power constant, then rank how difficult it is to get to that power level. So instead of arguing whether we should rank the noob Truenamer, the source-limited Truenamer, or the expert Truenamer, we should just look at "whatever Truenamer hits X level of power". Then it gets a ranking for how complicated that build is, and possibly what kind of sources that build requires.
This sounds like a useful metric, but a totally different one from what the tier system is trying to measure. Like, if the constant power level you decide on is something reasonable like "able to do its fair share as part of a 4-person party against a wide variety of level-appropriate content," then Warblade has an easier time hitting it than Wizard does.

I think your suggested metric basically just reinvents what we already refer to as "optimization floor."

[Yukitsu]

@Cosi:

My problem with the Tier system as established by JaronK always was, that there was no baseline assumption on what performance a character (solo/team) should have and what the expectations to this actually are, so the terms "breaking the game" or "do a thing well" are pretty much meaningless.

I think those weren't included and probably can't be since those are almost entirely player side questions and not class side questions.

[Florian]

I think those weren't included and probably can't be since those are almost entirely player side questions and not class side questions.

We're talking d20 here, that means a good chunk of the system is designed using certain formulae and a lot of the mechanics are based on the underlying assumptions how these are going to interact.

Based on that, it should be possible to 1) create a benchmark list of things that should be doable and 2) define an upper limit of what should be doable before problems with the core mechanics start to crop up. Part of that is accepting that WBL and itemancy are a thing and how they influence those mechanics/are influenced by them.

It's been already said that it's not classes that are problematic, but certain specific abilities or spells and the arms race starts to become serious when these come into play and counters to them are (Pparently) needed.

[Cosi]

It's the classes what need revision rather than the tiers. Revising tiers without touching the classes won't do anything other than soothe egos.

The tiers are an incredibly bad rubric for designing new classes. If you want to revise the classes, using something other than the tiers is more necessary, not less.


This sounds like a useful metric, but a totally different one from what the tier system is trying to measure.

My point, or part of it, is that what the tier system is trying to measure is not a particularly useful thing. Asking "how much does this break the game" doesn't give any useful information.


I think your suggested metric basically just reinvents what we already refer to as "optimization floor."

Not really.

First, it's not looking at "what's the weakest this class can be" but "how much effort to be level appropriate", which is a different (and more useful) question.

Second, it's splitting that discussion into two areas - how complex the class is to build, and how complex the class is to play. The Beguiler is trivial to build (put your highest stat in INT, now you can no longer make choices that make your character weak), but complex to play (by 4th level you have more than twenty spells to choose from). Whereas the Sorcerer is more complex to build (you can choose spells, which allows you to choose bad spells), but simpler to play (because you have way less spells).

Third, it's looking at multiple levels, rather than the game as a whole, which makes rankings more accurate. Building a useful first level Fighter is easy. Between heavy armor and thumbs, you're most of the way there without doing anything. Building a useful 15th level Fighter requires you to know where all the various Ubercharger tricks are, and even then you don't have anything to do outside combat.

[johnbragg]

The tiers are an incredibly bad rubric for designing new classes. If you want to revise the classes, using something other than the tiers is more necessary, not less.

If the objective is to assist a class-revision project, then we have a concrete goal, and I'd argue a lot of the effort ("this is the schema") and projected effort ("now we have to apply the schema to all eleventy-seven base classes ever created") in this thread is wasted.

I submit that there are really TWO major axes that make classes un-fun for one or more players at the table.

1. Division of Labor.
1A. Does the class effectively do a useful job in an adventuring party?
1B. Does the class need other party members to do jobs they can't effectively do?
2. Fun gameplay.
2A. Is the useful job effectively done a fun role to play in an adventuring party?

The "Tier 5" classes fail test 1A. The god-wizard and CoDzilla fail 1B. The healbot cleric and the core-only bard and the marshal generally fail 2A.

Now that I've made everything really really simple, we can make it really complicated again. Run these tests for in-combat and out-of-combat. (My wife barely cares about in-combat utility if she can run out-of-combat shenanigans. She'd manage with an Expert. But not typical.) Run these tests at low- and high-levels.

[Pleh]

lolwut? How is "comparable at low levels, but falls behind at high levels" a versatility issue?

You misunderstood. My argument was holding level constant to demonstrate versatility. Let me lay it out more.

Let's suppose level 4 Fighter and level 4 wizard (it's fine that your other post was referencing sorcerers, my point is regarding the importance of versatility in general). At this level, disparity hasn't hit overdrive yet and the classes perform roughly comparable in Power for the purpose of defeating enemies in combat.

However, suppose these same characters, who perform with equal power in combat head into town and come into a social encounter with no combat involved. The wizard selects spells that help with social encounters and effectively operate at the same power level as they did in combat. The Fighter, however, is pretty SOL. His player goes into the other room to play smash bros while the party resolves the encounter.

Next, the party gets a quest to go dungeon diving, but before they get back into combat with the dungeon boss, they have to navigate puzzles, traps, and dangerous climbing/balancing obstacles. The Fighter can probably survive toughing it out, but it could make him too weak to face the boss. The wizard preps a few spells and bypasses the challenges with the same power as in combat.

Versatility represents how consistent your power is in the face of a variety of obstacles.


What you should do is hold power constant, then rank how difficult it is to get to that power level. So instead of arguing whether we should rank the noob Truenamer, the source-limited Truenamer, or the expert Truenamer, we should just look at "whatever Truenamer hits X level of power". Then it gets a ranking for how complicated that build is, and possibly what kind of sources that build requires.

I agree with holding power constant to measure how much work it takes to reach that level. You still seem to being trying to rank Builds rather than classes. That seems to be a mistake, because people don't play builds, they play characters that use Builds. They create Builds by selecting classes.

If you want to rank Builds, go ahead. I just don't think anyone else cares about that information. What people want to know is the effectiveness of classes to create Builds for their characters to employ. There's not really that much point to measuring a build, as a build is something you build into from a set of class levels. Builds are arbitrary to set.

[Grod_The_Giant]

I'm chiming in with the people saying "remember to be simple." Heck, I'll SHOUT "remember to be simple."

REMEMBER TO BE SIMPLE!

Because all the analysis in the world means nothing if you can't communicate it clearly and if people don't want to use it. Nifft's bi-axis system is, I think, the only thing proposed that works to communicate more information than jaronk's system without confusing thing issue. Give it a defined range (say, 4-12) and you're set.

EDIT: But heck, you don't even need that much. A three-step system should work: too good (can solo things that should take a party), good (is useful but still needs allies), or not good enough (needs heavy optimization or handicap to contribute). One scale for power, one for options. Runs from Wizard (A1, can win all encounters by himself with ease) to C3 (Samurai, can't really do much of anything).

There's no need to take heavy optimization into account, because that's not how most people play the game.

[Cosi]

The "Tier 5" classes fail test 1A. The god-wizard and CoDzilla fail 1B. The healbot cleric and the core-only bard and the marshal generally fail 2A.

Generally agree with the first two, though to some degree this is putting the cart before the horse. If you decide to balance to God Wizard, then the God Wizard is balanced. If you decide to balance to the Fighter, the Fighter is balanced.

The last few miss the mark, IMHO. There's nothing wrong per se with the Healbot or Buffbot or whatever. Some people enjoy playing that kind of character, and they should be allowed to do that. The problem is making those things mandatory. If someone wants to play a Healbot, the game should support choosing to do so. On the other hand, if no one wants to play a Healbot, it should be fine to not have a Healbot.


Now that I've made everything really really simple, we can make it really complicated again. Run these tests for in-combat and out-of-combat.

This is pretty much my view of how to balance things. Make a slate of encounters, pick a target win percentage, iterate until each class is within some error of that percentage. Basically, the Same Game Test, but with more out of combat stuff (which the SGT lacks mostly because there are no benchmarks for such abilities in 3e).


(My wife barely cares about in-combat utility if she can run out-of-combat shenanigans. She'd manage with an Expert. But not typical.)

The answer to that is the same as the answer to all the people who show up in Fighter threads and claim that their concept of a Fighter doesn't have non-combat abilities: don't use abilities you don't want. Also, have classes that are at varying complexity points in and out of combat so that someone who wants to be involved in combat heavily, but mostly ignore non-combat (or vice versa) can do that.


Versatility represents how consistent your power is in the face of a variety of obstacles.

That still sounds like two parts of one metric. The character with a lot of power and no versatility is bad (e.g. someone who has pawned ever ability for super-powered Turn Undead) the character with no power and a lot of versatility is also bad (e.g. someone with all 1st level spells at will at 15th level and no other abilities). I just don't see how you'd try and measure versatility without measuring power or vice versa.


If you want to rank Builds, go ahead. I just don't think anyone else cares about that information. What people want to know is the effectiveness of classes to create Builds for their characters to employ. There's not really that much point to measuring a build, as a build is something you build into from a set of class levels. Builds are arbitrary to set.

You're setting up a false dichotomy. "Dread Necromancer with list expansion options" is not a class, but it's also not a build. It (just like classes, really) is a slice of the space of possible builds. The difference between the Tier System and any other system is just how it slices that space. So how do you best slice up that space?

For the system to be useful, it needs to divide the space in a way that has two properties. First, divisions need to mimic divisions expected in the play space. I think the Tiers fail this by relying to much on "campaign smashing" as a divider. Second, divisions need to be precise enough to be usefully described. I think the Tiers fail this by conflating things like level (Fighters are probably better than Artificers at very low levels) and various feats/items/ACFs/PrCs that modify class value (Bard is much worse if it is not allowed options like Sublime Chord or Dragonfire Inspiration).

[Yukitsu]

We're talking d20 here, that means a good chunk of the system is designed using certain formulae and a lot of the mechanics are based on the underlying assumptions how these are going to interact.

Based on that, it should be possible to 1) create a benchmark list of things that should be doable and 2) define an upper limit of what should be doable before problems with the core mechanics start to crop up. Part of that is accepting that WBL and itemancy are a thing and how they influence those mechanics/are influenced by them.

It's been already said that it's not classes that are problematic, but certain specific abilities or spells and the arms race starts to become serious when these come into play and counters to them are (Pparently) needed.

That's the problem, checking all marks off that checklist is going to be player dependent, not based around the class. Class can indicate how cheaply those problems can be bypassed and class can indicate how many different problems can be bypassed but there isn't any class that can't bypass a single type of encounter if it was built to. You'd need to create an assumed build and keep it consistent across multiple theoretical encounters but that is now looking at the player skill of the person who made that specific model.

[Troacctid]

Is there something wrong with just ranking classes from best to worst? Why does it have to be complicated? Why do you need criteria other than "More powerful than the classes below it"?

[Aimeryan]

Alright, seems people want things to appear simple (even if they are not under the hood). I'll try to follow this line of thinking, although it is not my strong suit.


That still sounds like two parts of one metric. The character with a lot of power and no versatility is bad (e.g. someone who has pawned ever ability for super-powered Turn Undead) the character with no power and a lot of versatility is also bad (e.g. someone with all 1st level spells at will at 15th level and no other abilities). I just don't see how you'd try and measure versatility without measuring power or vice versa.

I think I understand what you are saying, but I think you didn't carry on the thought process far enough:
The first character can actually do something when the one thing they have power in is appropriate, while the latter can never do anything. The first character is bad, unless what they can do comes up often. The second character is always bad. However, the second character is not effectively versatile.

OldTrees1 actually explained all this a lot earlier in the thread, although I think he was passed over for the wall of text:


Parties(multiple characters) can be faced with countless varieties of challenges. In each of those challenges an individual character's ability to contribute is some real number(positive or negative). While negative values are possible, I am going to ignore them for simplicity. This ability to contribute can be near 0 (basically the character is sidelined or otherwise negated for the challenge), low (able to provide low skill contributions), high (able to provide high skill contributions), total (able to handle the challenge on their own), or excessive (their mere presence makes the challenge a waste of the player's time).

There are a few issues with "versatility" that come about from just having too many meanings or just not being strongly enough bounded:

1) A character that could potentially do many different things could be called versatile.

2) A character that can do many different things could be called versatile.

3 A character that can do many different things effectively could be called versatile.

All these have very different meanings. That said, I think for a tiering system only number 3 is useful to us.

Meaning number 1 could be used to call something like a Fighter versatile; they have a number of feats that could be used for many a different purpose. However, once built he is no longer versatile by meaning numbers 2 and 3 because he has to specialise to be able to actually do anything (heading towards the first character in Cosi's examples).

Meaning number 2 could be used to call the second character in Cosi's examples versatile. However, he doesn't get anything actually done; he would not be versatile under meaning number 3.

Meaning number 3 would require that a character also had a minimum amount of power to be considered effective (which I believe Cosi is what you were trying to put forward?). Where that line is drawn is a problem though; some people would say Call Item on a Psychic Warrior makes him effectively versatile (at low levels) - others would not. You also have to consider levels because something that doesn't scale may start out effectively versatile but become ineffectively versatile later.

[johnbragg]

Is there something wrong with just ranking classes from best to worst? Why does it have to be complicated?

Because it's complicated. You'd never see the end of arguments that Druid is better than Cleric because (specific build), but on the contrary Druid is better than Cleric because (specific example). On the other hand, Sorcerer is really better because (argument that is vociferously rejected by Druid and Cleric).

[quote]Why do you need criteria other than "More powerful than the classes below it"?

...because the idea is to create something that will be widely agreed on.

JaronK's Tier system caught on because it expressed something that we had dimly grasped, but did not have a terminology or an explanation for. It didn't just say "Wizard > Sorcerer > Bard > Rogue > Fighter", it gave an explanation as to why each one was in a category.

People on this thread seem to want to create something far more complex than JaronK's system. A more complicated system may or may not be a better reflection of the truth, but it will definitely hinder adoption of the system by anyone else.

[OldTrees1]

There are a few issues with "versatility" that come about from just having too many meanings or just not being strongly enough bounded:

1) A character that could potentially do many different things could be called versatile.

2) A character that can do many different things could be called versatile.

3 A character that can do many different things effectively could be called versatile.

All these have very different meanings. That said, I think for a tiering system only number 3 is useful to us.


These 3 definitions are a useful distinction. However I would argue both 2 & 3 are relevant to a useful tiering system.

To simplify via analogy:
At a particular task you can either be useless(don't bother rolling), able to assist(aid another), or proficient (able to attempt the main roll).


Somebody that can do 1 thing proficiently, but can't even assist at other tasks is a weak character.
Somebody that can do nothing proficiently, but can assist at lots of tasks is also a weak character.
However they are weak in qualitatively different ways.

Then you move onto the characters that have some mixture of things they are proficient at, and somethings they are able to assist at.

Somebody that can do many things proficiently, able to assist in a handful, but has noticeable gaps where they cannot even assist.
Somebody that has only a handful of things they are proficient in, but can assist at everything else with no gaps.

These two characters are qualitatively different. And notice I have not even touched on how strong they are at the things they are proficient in (beyond setting a minimum bar).

In the current Tier System(which already hides 2 variables in its Tier 1-6 listing), you cannot even begin to have a reasonable discussion about Tier 3 or 4 because they attempt to treat 2 different types of versatility as the same thing.

[Troacctid]

Because it's complicated. You'd never see the end of arguments that Druid is better than Cleric because (specific build), but on the contrary Druid is better than Cleric because (specific example). On the other hand, Sorcerer is really better because (argument that is vociferously rejected by Druid and Cleric).
No one said you couldn't still group them. Rank them in groups based on their power level. That's basically what JaronK did, except he also made descriptive observations about the things he noticed that classes in each group had in common.

[johnbragg]

No one said you couldn't still group them. Rank them in groups based on their power level. That's basically what JaronK did, except he also made descriptive observations about the things he noticed that classes in each group had in common.

Then you'll get bogged down in an endless series of arguments about the borderline cases. JaronK's classification system at least gave you a standard to evaluate what makes a Tier 4 Tier 4. "Better than/worse than" is just asking for endless arguments--ones even more pointless than usual.

You could argue endlessly, and with some merit, about whether Rogue or Barbarian is *stronger* than Bard. It's much harder to argue that there aren't situations where the Rogue or Barbarian doesn't have much to do, while the Bard can always at least Inspire Courage/Competence.

[Troacctid]

No, Bard is definitely stronger, not close.

[P.F.]

you'll get bogged down in an endless series of arguments about the borderline cases.

If the standard for a better tier system is that it won't be argued all the way to death on issues of outliers and in-betweeners, I would contend that such a ranking system does not exist.

[tsj]

Do we have Any replacements for All classes tiered in jaronks tier system, that are guaranteed to be of equal power level over all 20 levels both in and out of combat? If yes, do we have a list?

Maybe when a suiteable ranking system is found, then All the classes found in jaronks tier system can first be placed in the new system and secondly be redesigned to be equal (some specific power level in and out of combat etc must be selected)

[Pleh]

That still sounds like two parts of one metric. The character with a lot of power and no versatility is bad (e.g. someone who has pawned ever ability for super-powered Turn Undead) the character with no power and a lot of versatility is also bad (e.g. someone with all 1st level spells at will at 15th level and no other abilities). I just don't see how you'd try and measure versatility without measuring power or vice versa.

Ever study some Thermodynamics? PV=nRT.

What I mean is that Pressure, Volume, and Temperature for a gas are separate, but related. If you raise the temperature of a gas, it will try to expand its Volume to maintain equilibrium. If you constrain it to an air tight canister and raise the temperature, its Pressure will rise since the Volume isn't allowed to expand.

Likewise, if you put the gas in an airtight chamber with a pump, you can move the pump to change the volume of the space. If no air can enter or leave the vacuum you are creating by expanding the gas's Volume, the Pressure will drop, along with the Temperature.

But if you want to change just only exactly the temperature of the gas without changing the pressure or the volume, you can poke a tiny hole in the chamber and let little molecules out so the pressure and volume do not change while the temperature rises.

What I'm saying is that in D&D, Power and Versatility are related, but not the same. Having more power will make you more versatile and vice versa, but it is possible to hold one constant and change the other. You just have to understand how one affects the other. It's not that they are effectively the same unit. It is that there is a correlation between them that makes a change in one affect the other.

You are quite right that it is pointless to have a god-like magic hammer that only hammers nails really well and can't be used for anything else. And it's equally useless to have a Jack of All Trades that truly is a master of none and actually can't do anything at all. Both of these rank as Tier 5 or Tier 6 (maybe Tier 4 if your campaign involves a lot of nailing things down).






You're setting up a false dichotomy.

And you aren't defining your terms very well. Exactly where does "Playspace" appear in any of the books? Sure, "Build" doesn't really appear in the books either (not in the sense we're using beyond the introduction chapters about "how to build your character"), but I haven't heard of "playspace" much at all until you started using it. Is it an obscure reference from a DMG somewhere?

It can be handy to argue in a set of terms other people aren't familiar with. Makes it easy to win arguments without actually communicating anything sensible. [/drhousevoice]

For now, I'll assume I understand what you mean by "playspace" and you can correct my understanding in your inevitable rebuttal.


"Dread Necromancer with list expansion options" is not a class,

...that's... exactly what it is. A class is a set of character options that grant abilities and other benefits that allow a heroic character accomplish fantastic tasks. The fact that there is an option to increase their class options doesn't make it not a class.

In my understanding of "playspace," a Class marks a region of Playspace a character can occupy. Most classes cover more Playspace than any single build can cover, but highly Versatile builds can cover more of their class's available territory than one trick ponies can.

(as a side note, characters that specialize in their one shtick theoretically have a better chance of doing that one thing successfully far more often than their versatile counterparts at doing any of the many other things they are moderately successful at).


but it's also not a build. It (just like classes, really) is a slice of the space of possible builds. The difference between the Tier System and any other system is just how it slices that space.

So, because you don't have to (and/or don't get to) necessarily use all of the possible options made available to a class in a single build, that makes it not a class, but a slice of space in some abstract plane of character existence?


So how do you best slice up that space?

I'm guessing that your answer isn't simply, "With a class"?


For the system to be useful, it needs to divide the space in a way that has two properties.

Like power and versatility. Oh, wait. That can't be right.


First, divisions need to mimic divisions expected in the play space.

Like power and versatil-


I think the Tiers fail this by relying to much on "campaign smashing" as a divider.

Okay, fair point. I hear you saying that "campaign smashing" is not a good metric as it is not a set quantity. "Campaign Smashing" is a variable that is determined by where the DM sets the difficulty. No campaign can be smashed if the DM sets the difficulty to the proper level.

I think where most people would define "campaign smashing" is somewhere in the region of "playspace" where the RAW in question leaves the DM with little choice but to balance with Rule 0. Such as "you can't farm efreets for Wishes because NO."


Second, divisions need to be precise enough to be usefully described.

But they also need to be concise enough to be quickly and easily understood.

Honestly, if players want reliable and precise knowledge of the abilities and powers of a class, they can read the source material on the class and then followup with a handbook to avoid common pitfalls.

If there's going to be a Tier, it needs to be the "Global Map Overview" of your "Playspace" and leave the heavy work of precise power (your "local area maps" of the "playspace") levels to the handbooks.


I think the Tiers fail this by conflating things like level (Fighters are probably better than Artificers at very low levels) and various feats/items/ACFs/PrCs that modify class value (Bard is much worse if it is not allowed options like Sublime Chord or Dragonfire Inspiration).

This doesn't sound like a problem with the Tiers, so much as the "Why X is in Tier Y" thread that accompanies it. Any good tier thread is going to have an abstract explaining the tier list for a reader's digest version of what the handbooks will ultimately tell them in more detail.

I believe all those things (feats/items/ACFs/PrCs) deserve some attention in ranking classes, but I would argue that classes that rely on these things more heavily than others should be marked for their dependency compared to other classes. The fact that some Classes NEED help while others simply enjoy the fact that its there is absolutely a statement of relative power.

What I would want a Tier to tell me about Bard is this:

Bard
Poster Child T3 with an emphasis on Combat Support and Non-Combat Utility.
WARNING: Bard has a strong dependence on feats/items/ACFs/PrCs to get to T3 power level consistently throughout its level progression, so make sure you read the handbook on it so you don't end up building straight into a sub-par Bard.

---

That being said, I see a lot of people arguing that the real purpose for a solid Tier system isn't to help people build fun, balanced games, but to help rewrite the game and hardwire it to be balanced.

I would have thought 4e taught us better than to think that way by now. 3.5 is plenty flexible enough to let DM's craft balanced adventures with parties of any Tier as it is. As nice as it would be if the Fighter didn't suck, I'm not sure going back and rewriting it will do much more than soothe people's egos any more than "fixing" the tier system will.

[Cosi]

People on this thread seem to want to create something far more complex than JaronK's system. A more complicated system may or may not be a better reflection of the truth, but it will definitely hinder adoption of the system by anyone else.

I don't think the system is necessarily more complicated. There's more stuff, but that doesn't make it harder to use. You go to the level range you care about, then look at classes or rankings your interested in. Plus, categories like "complexity to build" are more descriptive than "tier".


Do we have Any replacements for All classes tiered in jaronks tier system, that are guaranteed to be of equal power level over all 20 levels both in and out of combat? If yes, do we have a list?

Maybe when a suiteable ranking system is found, then All the classes found in jaronks tier system can first be placed in the new system and secondly be redesigned to be equal (some specific power level in and out of combat etc must be selected)

I'm not sure what you're trying to say here.


What I'm saying is that in D&D, Power and Versatility are related, but not the same.

Sure, but they aren't usefully different.

Power is your ability to overcome encounters.

Versatility is what encounters you can overcome.

You could totally rank those separately, but why would you?


And you aren't defining your terms very well. Exactly where does "Playspace" appear in any of the books?

It's the possible characters you can play (formally, you could say "the enumeration of all possible Builds" or "the phase space of possible Builds"). I was not aware that this was an obscure term.


...that's... exactly what it is. A class is a set of character options that grant abilities and other benefits that allow a heroic character accomplish fantastic tasks. The fact that there is an option to increase their class options doesn't make it not a class.

Does your definition of Class include Feats and Items? If so, that seems like bad terminology. If not, I'm not sure how a Class plus additional options is equal to a class.


So, because you don't have to (and/or don't get to) necessarily use all of the possible options made available to a class in a single build, that makes it not a class, but a slice of space in some abstract plane of character existence?

Didn't you just agree that a Class represents a set of possible Builds?

Any property characters have (ability scores, number of class abilities whose names have a prime number of vowels, feats, etc) can be used to divide up Builds. That doesn't make it stop being whatever thing it nominally is.


Okay, fair point. I hear you saying that "campaign smashing" is not a good metric as it is not a set quantity. "Campaign Smashing" is a variable that is determined by where the DM sets the difficulty. No campaign can be smashed if the DM sets the difficulty to the proper level.

No, I'm saying that's a bad metric because it doesn't matter. People don't play broken games. When someone implements a power loop like More Wishes or Chain Binding that breaks the game, one of three things happens:

1. The game ends. Result: people are not playing that game, and are necessarily not playing a broken game.
2. The ability is banned. Result: people are not playing a game with the game-breaking ability, and are not playing a broken game.
3. The game and/or ability are modified to allow the game to continue. Result: the game is not broken.

None of those result in you getting to use your ability to break the game.


I believe all those things (feats/items/ACFs/PrCs) deserve some attention in ranking classes, but I would argue that classes that rely on these things more heavily than others should be marked for their dependency compared to other classes. The fact that some Classes NEED help while others simply enjoy the fact that its there is absolutely a statement of relative power.

No, it isn't. If I can build a guy with a pile of Beguiler levels at the same power as a guy with a similar pile of Wizard levels, there is no practical power difference between Beguiler and Wizard.


I would have thought 4e taught us better than to think that way by now.

No, it didn't. 4e failed because Mearls et al are hacks who made a terrible game, not because it tried to balance the game. Plus, the Yogi Hat Ranger, the Orbizard, and plenty of other cheese builds exist in 4e.

[tsj]

Cosi:

What I suggest was that if or when jaronks tier system is replaced by something that is hopefully better

Then it should be followed by placing All classes that were tiered by jaronk in to the new tiering system.

After this All the classes might need redesigned based on the wisdom gained from the new tiering system such that we can get versions of the classes that are All equal in both power and versatility across All 20 levels

That might Mean merging of some classes and splitting of others

Later prc should be looked at and given the same treatment

[Grod_The_Giant]

I don't think the system is necessarily more complicated. There's more stuff, but that doesn't make it harder to use. You go to the level range you care about, then look at classes or rankings your interested in. Plus, categories like "complexity to build" are more descriptive than "tier".
You're referring to the one Aimeryan proposed that gives thirty different numbers per class? The Tier system caught on and continues to be used because it's simple and descriptive. The single axis is an oversimplification at times, and Tiers 3 and 4 can be a mess of power-vs-versatility debates, but it's easy. And it's much, much more useful to be able to group classes into broad categories

If we want a way to include more information in the Tier system that will actually be useful? How about this:

We write in a specific condition to the base list, to prevent endless arguments about what is and isn't kosher. Say, "levels 4-14, no ACFs, feats, or multiclassing that makes significant changes to the basic structure of the class." No spell list expansion, no trading away crap abilities for Wild Shape, and so on, just the base class played about how (late edition) WotC expected.
Add a "+" to any class that significantly improves in a high optimization environment, when you start adding in those significant changes. An "A-game____" is possible, basically.
Add a "-" to any class that significantly falters in a high optimization environment, due to the lack of "A-game" options.
Options that add purely external power, (say, Ur Priest or dabbling in Incarnum) should be totally ignored, as should those that are broadly available (say, Incantrix or Shock Trooper combos). Only ones that are strongly linked to specific classes (say, Sublime Chord or Frenzied Berserker) can be considered as part of the class's Tier, and then only in the +/- side of things.


A Ranger, say, would be a Tier 4+. The base version is a mediocre generalist, but you can start adding stuff like Mystic Ranger, Wild Shape Ranger, and Sword of the Arcane Order to build something fantastic. A Barbarian, on the other hand, might be Tier 4-. The base version is a decent warrior, and no matter how hard you try you can't really make it more than that. Beguiler is likely to be Tier 3+, while the Death Master might be Tier 1-.


@ tsj-- a lovely idea, but historically community-based rewrite projects fare poorly. Regardless of Tier system, class flaws tend to be widely known, but the problem comes in agreeing how to fix them-- look at the recent thread about the Fighter. The homebrew board as hundreds of class/rewrites, if you want to pick your favorites.

[Pleh]

I don't think the system is necessarily more complicated. There's more stuff, but that doesn't make it harder to use. You go to the level range you care about, then look at classes or rankings your interested in. Plus, categories like "complexity to build" are more descriptive than "tier".

I'm just warning you that you're going to find the public at large to disagree with you. More "stuff" = harder to use. It's not that people mind using systems that are hard to use (we play the stupid game, anyway and the game itself has a tremendous learning curve).

But information has to be worth the effort to acquire. If someone wants to deal with an abundance of information to learn about character option nuances, they can just read a handbook. No need for tiers or ranking anywhere.

People ask for a Tier because they want a simple version that doesn't have a lot of "stuff." Cut it down to a sizable chunk to where we can see a whole picture at once without having to juggle 30 different numbers in our heads.

Just because that's how the game is actually played doesn't mean that's how the game is learned, studied, or understood. Physicists still haven't solved the multi-body problem.

People do not think in terms of multiple variables at once. That's why we invented Algebra to let us tackle multiple variables with one variable at a time. One thought. One thing. Not "stuff."

And if you want "playspace" to be the thing that gets ranked and talked about rather than classes, fine. But I don't think you'll get many players and DM's using Google to search for relative "playspace strength."

People want to know how much a Class, with its various options, can help them play their games. Even if a Class isn't played by itself in a vacuum, people still want to know what it can do, so they need it to be simplified so they understand the limitations of the class. If the class NEEDS feats and items and PrCs to work, just tell them that the class doesn't work on its own and rank it as a weaker class than other classes that don't actually need these things as desperately.


Sure, but they aren't usefully different.

Power is your ability to overcome encounters.

Versatility is what encounters you can overcome.

You could totally rank those separately, but why would you?

You are conflating terms. I know you think these terms SHOULD be conflated, but you seem to take issue when other people conflate terms elsewhere.

Power is your ultimate effectiveness when you are told to make a roll.

Versatility is the number of rolls that you qualify to even attempt.

A Ranger with DEX maxed out to heck can't just roll a Nat 20 on an Open Lock check if he doesn't have any ranks in Open Lock (or some OTHER class feature that otherwise allows him to make the attempt). No matter how Powerful he should be at making the attempt, he doesn't get the chance.

Meanwhile, a Wizard can Polymorph and make more Attacks per Round than the Fighter using a monster's Natural Attacks. Even if he misses all his rolls, he had the opportunity to make the attempt.

Versatility and Power.


Does your definition of Class include Feats and Items? If so, that seems like bad terminology. If not, I'm not sure how a Class plus additional options is equal to a class.

Your original post said,
"Dread Necromancer with list expansion options" is not a class, so are the list expansion options coming from feats and items then? Because the Spell List is a Class Feature, so if they're gaining List Expansions as part of their level progression, it's all part of the class



Didn't you just agree that a Class represents a set of possible Builds?

Any property characters have (ability scores, number of class abilities whose names have a prime number of vowels, feats, etc) can be used to divide up Builds. That doesn't make it stop being whatever thing it nominally is.

Sure, but builds are not the only thing that people care about and want information on. A Tiering system should be focused on places that actually have problems with distinction. Builds have much less problem on that then classes.

You want Ubercharger? Look up a handbook and weigh out your options. People have already pretty much got the science to your regular builds down to a science.

What needs more explanation are the classes. Builds are the final product and are easy to understand and evaluate. Classes are building blocks and Tier theory suggests that not all of them possess the same potential for making certain builds. Some lack Power (effectiveness) and some lack Versatility (applicability). But understanding the limits of how much playspace a class gives you is useful in making builds, so you know how much a class helps you crafting the build and how much its hurting your resources.

There's not much point in ranking Builds because there is no theoretical limit to builds. You put enough TO and Cheese into any build and everything hyperbolically approaches Pun Pun.



No, I'm saying that's a bad metric because it doesn't matter. People don't play broken games. When someone implements a power loop like More Wishes or Chain Binding that breaks the game, one of three things happens:

1. The game ends. Result: people are not playing that game, and are necessarily not playing a broken game.

This still is a problem and it still matters. A game, possibly years of hard work by a dedicated group, just went rotten and was thrown out because balance was lost.

The Game was Broken. Then, the game was lost. How can this not matter? At least, not any more than any of this matters.


No, it isn't. If I can build a guy with a pile of Beguiler levels at the same power as a guy with a similar pile of Wizard levels, there is no practical power difference between Beguiler and Wizard.

But if you have only one or two possible builds in your Playspace to create that Beguiler powerful enough to compete with the wizard, while the guy with the Wizard has about a dozen possible builds in his playspace around that same level, the fact that they turned out the same power doesn't exactly change the fact that the next person playing a beguiler really doesn't have much room to stray from your build if they want to keep up with their own table's wizard, who looks nothing like your wizard, but dominates the table anyway.

[Cosi]

You're referring to the one Aimeryan proposed that gives thirty different numbers per class?

Not at all. My idea system would probably be:

1. Complexity to Build
2. Complexity to Play
3. Likelihood of Breaking the Game (not in the way JaronK uses it, this would also include characters who are dramatically underpowered)
4. Source Dependency/Need to Fast Talk the GM

That's four things, but they're all (potentially) independent. Then iterate at 5, 10, and 15, holding people to however much power is necessary to bat 50% on the Same Game Test. Pretty heavily based on this (http://www.tgdmb.com/viewtopic.php?t=50101) discussion and the similar thread here a few weeks ago.


But information has to be worth the effort to acquire. If someone wants to deal with an abundance of information to learn about character option nuances, they can just read a handbook. No need for tiers or ranking anywhere.

That information exists in either system, and is not user facing in either. Whatever system you use has to consider various things like feats and ACFs at some point, regardless of where it draws the line on categories. And if you're figuring out that information, slicing at the level of "class" (particularly without multiple level break points) ends up discarding the majority of what is important.


And if you want "playspace" to be the thing that gets ranked and talked about rather than classes, fine. But I don't think you'll get many players and DM's using Google to search for relative "playspace strength."

You are always ranking the playspace, because it is defined as the set of things it is possible to be. The question is just how we should define the sections that get write-ups.


You are conflating terms. I know you think these terms SHOULD be conflated, but you seem to take issue when other people conflate terms elsewhere.

Power is your ultimate effectiveness when you are told to make a roll.

Versatility is the number of rolls that you qualify to even attempt.'

Yes, you can divide those things. You can also divide classes by starting age. The point is that "things you can do poorly enough to not matter" are functionally indistinguishable from "things you can't do".


Your original post said, so are the list expansion options coming from feats and items then? Because the Spell List is a Class Feature, so if they're gaining List Expansions as part of their level progression, it's all part of the class

How could they be anything else? If you're just talking about options you get for being a Dread Necromancer, that's just "Dread Necromancer".


There's not much point in ranking Builds because there is no theoretical limit to builds. You put enough TO and Cheese into any build and everything hyperbolically approaches Pun Pun.

This is exactly as true of Classes. Like, literally exactly. Candles of Invocation can be bought for money.


This still is a problem and it still matters. A game, possibly years of hard work by a dedicated group, just went rotten and was thrown out because balance was lost.

The Game was Broken. Then, the game was lost. How can this not matter? At least, not any more than any of this matters.

It matters, it just doesn't matter to class ranking. A list of Genuinely Broken Things (http://tgdmb.com/viewtopic.php?t=1156&postdays=0&postorder=asc&start=0) (ideally coupled with fixes) would be helpful for people running the game. But it wouldn't effect how we evaluate classes.


But if you have only one or two possible builds in your Playspace to create that Beguiler powerful enough to compete with the wizard, while the guy with the Wizard has about a dozen possible builds in his playspace around that same level, the fact that they turned out the same power doesn't exactly change the fact that the next person playing a beguiler really doesn't have much room to stray from your build if they want to keep up with their own table's wizard, who looks nothing like your wizard, but dominates the table anyway.

No. The playspace your build does not occupy is constant across all builds. Specifically, it's locked at the number of possible builds minus one. It is never larger or smaller than that number, regardless of how much of the playspace is similar to your build on some axis, and the size of those zones of similarity does not make your character any better.

[bekeleven]

I've always found it more useful to think of the tier system this way, even though it disagrees with JaronK's descriptions of the tiers, especially within tiers 2, 4, and 5.



Versatiliy
High
Low


High Power
1
2


Medium Power
3
4


Low Power
5
6


This has other pitfalls. While it's roughly as accurate as his original tier system at describing most extant classes, it's slightly inaccurate at describing the power level of a few, such as the Monk. However, it is also better than JaronK's system at describing others, like tier 2, where his definition locks out a ton of homebrew classes that fit within the same power/versatility band as sorcerers and standard-op psions.

That said, it's similar enough to JaronK's system, and so simple that I think it's worth any extra issues.

[rrwoods]

I said it in the community tier thread and I'll say it here:

I think the use of the words "campaign smasher" is unfortunate. No one* actually plays broken games, to be sure. But to me the 2/3 divide is between characters that trivialize encounters so easily that it's almost accidental and characters that are strong in a more expected sense. Glitterdust is a good example; its existence makes it very easy for a character to "oops" win an encounter. Campaign derailing? No. Enough to make a T3-4 character feel useless if expectations are shattered? Yeah.

*Tippyverse notwithstanding

[tsj]

@grod:
That means that it might be better to list the broken things and various suggested fixes instead such that a DM can select what he wants to fix and what fixes to use?

I just read the thread that was linked regarding broken stuff in 3.0 and 3.5. However I don't know if that list has everything.

It did give me a desire to make a vampire fix though... a vamp SHOULD be stakeable with a wooden stake and should be extra vulnerable to light, silver, holy water and holy symbols... Buffy style :-)

[Aimeryan]

You're referring to the one Aimeryan proposed that gives thirty different numbers per class?

Wait, what?! I proposed a systems that gives ONE total each for THREE different level ranges (Low, Mid, High). That isn't 30! And, I even said we could get rid of the level ranges if people found that three numbers were too many for them to comprehend.

If you are talking about the categories (of which there was ten), that was explanatory. JaronK's system isn't just one number - it has explanations as well, and more than 30 words per explanation, I am sure.

Is this really why people thought my proposal was too complicated? You still have to think about all that stuff when coming up with a tier number; the only difference is you are not showing your working out.

[VisitingDaGulag]

The only problem with how people use class tiering is that it isn't a complete picture on character tiers. This isn't a problem with class tiering itself. PrC tiers help, but the ranking system for items isn't a fully fledged tier system and it isn't finished. Plus there are other sources of PC power like races, templates, feats, rituals, etc.


On a related note, it's not clear which and how many splatbooks should be evaluated. Core + Class Book? Everything?Everything. It's not fair to rate a barbarian without madfoam rager because its a disallowed source. You can do it, but as usual t1 druids will laugh because natural spell is core. Without totally playing favorites, lower tier (less powerful) classes need more sources because they are the classes that most need to supplement their class features.


Simply saying "class X is in tier Y" is not very helpful for evaluating class power.Why not? If you say "I made a class that is tier x" I'll know exactly what level of power to expect from the class. Am I special for understanding this?


My proposal looks something like the proposal to rank classes by complexityHave you read this thread (http://www.giantitp.com/forums/showthread.php?509515-Is-there-an-ACTUAL-tier-list-for-3-5)? It links to someone who uses a coloring scheme for high/medium/low complexity.