Retiering the Classes: Home Base

[Troacctid]

There's no particular reason you can't have "endurance" fights in the SGT (indeed, the various "traps" encounters fall fairly close to that, as are the "navigate an environment" ones). You could plausibly have something like "beat up a CR 3 opponent every 15 minutes for an hour" as a 5th level SGT encounter, tuning the frequency or CR up or down to reach whatever level of difficulty you think is parity with the rest of the test.

Really, a lot of objections to the SGT are resolved by point out that it's a framework and you can slot in encounters that happen to test whatever you're looking for, whether that's "endurance" or "versatility" or "utility".

But you weren't going to do that.

[Beheld]

The solo aspect is to some degree necessary, because a group test would require an otherwise balanced party to slot characters into, which presents something of a bootstrap problem. Still that's a least approachable by running two Level - 2 characters at a time (a buffbot and a pile of stats), or adding encounters which include allies.

I think I actually have a pretty decent "fix" to that problem, but alas, making a thread to run SGTests on this forum would be a bad idea.

But you weren't going to do that.

Some of the encounters I have in the addendum to the SGT are in fact very endurance.

[eggynack]

I think that's probably an argument that power tiering is not very useful. If you're concerned that variations in optimization overwhelm other concerns, you should be ranking by optimization itself.

I don't think that optimization variance ruins everything. I simply makes constructing useful definitions really difficult. I think there's a good amount of meaning to tier threeness. It's just really difficult to convey without writing a huge essay about the various ways to be tier three, and without winding up unfortunately prescriptive. "Worse than the worst tier four, but better than the best tier six," has an insane amount of complexity bound within it. But I think it means something to people in spite of that. It's the classic math problem. You have this idea that's really straightforward, and which everyone intuitively gets more or less immediately, but stating and proving that idea rigorously takes a massive amount of effort.

[jywu98]

Eggy, you have yet to explain of what use a tier system is as a resource for 14 pages. It doesn't matter if it is of value to people, as people are often wrong.

[ryu]

Eggy, you have yet to explain of what use a tier system is as a resource for 14 pages. It doesn't matter if it is of value to people, as people are often wrong.

Do... Do you not understand that use is a term derived from people, that's right PEOPLE, usING something for purpose. Throw people out of the equation and the term loses all meaning.

[eggynack]

Yeah, I have no idea what you're talking about there, jywu98. I explicitly said the following:

I didn't agree with them. Obviously. I think there's utility to being able to identify how strong a class is, both in that you can use that to pick classes closer in power to classes already in the party, and so that the DM can make decisions about what to allow or ban on the basis of how good a class is. It helps balance. You might not agree with me either, but I think people here see the utility to this.

Those are clear sources of value that aren't simply, "I have no idea what the tier system does, but folks seem to like it." Whether you agree with my explanation is irrelevant. An explanation has absolutely been given.

[jywu98]

No, because your explanation is absolutely wrong. If you asked me to explain why the sky is blue, and I say it's because space is blue, yeah sure it's an explanation, but not a valid one because it's wrong.

I think there's utility to being able to identify how strong a class is, both in that you can use that to pick classes closer in power to classes already in the party, and so that the DM can make decisions about what to allow or ban on the basis of how good a class is. It helps balance. You might not agree with me either, but I think people here see the utility to this.

both in that you can use that to pick classes closer in power to classes already in the party

There is nothing here that is in any form correct, especially because retards keep thinking more versatile = more power. Kaelik already tried to address this when you last tried to pull out this bull**** on minmax.

So it sounds like what the Tier system is doing is making sure the party is imbalanced because of it's own dumb failures. If the same player could play either class, the same player would be more balanced in both parties of:

Beguiler/Wizard/Cleric/Druid and Archivist/Warblade/Totemist/Dread Necromancer

Than with Beguiler and Archivist reversed.

If they player isn't going to jump through hoops to spell acquisition, then the Beguiler bring trap finding, diplomacy, and spontaneous casting of pretty good spells, and maybe a minion army. Where the Archivist casts like a ****ty Cleric who wakes up every day with fewer spells per day or worse save DCs, pick one.

If the player is going to jump through hoops to spell acquisition, then the Beguiler spontaneously casts from a huge list of really good spells and can easily keep up with Wizards and Clerics and Druid, and the Archivist is like a ****ty Wizard who has fewer spells per day or lower DCs and also knows a couple good Druid spells, and an occasional other spell at a low level.

In both cases the Archivist is the odd man out in the Wizard party, where he doesn't contribute much, where the Beguiler always has something to do. In both cases the lower level party is a perfect fit for the Archivist, where the Beguiler would, if anything, remind everyone else how little he needs them.*

*Dread Necromancer excepted, since it's just like the Beguiler, but if the Dread Necro is a different player, it could be played at a lower optimization level.

And that's the point, the Archivist is just not a "higher tier" class than the Beguiler, it's a special snowflake class that JaronK liked, so he said "well obviously the DM is going to let you contract a 12th level Warlock make Divine Scrolls of Divine Bard Spells and Alternate Spell Source Assassin and Trapsmith Spells because it's a special Tier 1 class, and Tier 1 classes are VERSATILE!" But then he didn't like Beguilers so he turned around and said "What of course not, no Beguiler every spends feats or PrCs for more spells to cast, they definitely don't get their first Prestige Domain at level 2 and then use the Substitute Domain Spell to repick domains whenever they want and then spontaneously cast off the entire Cleric list! That's optimization, and filthy bad Tier 3 classes don't do optimization to expand their spell list!"

Which then leads to people, 8 years later, claiming with a straight face that if a single player has a choice between Beguiler and Archivist, that the exact same player will someone optimize the Archivist a different amount than the Beguiler, because the Beguiler expanded spell list tricks are "Just totally out there man" and the Archivist tricks are "Just something you do if you are an Archivist."

So how do you actually build a good party? Surely it's not slapping a bunch of classes of the same tier and calling it a day. Does having a Barbarian or a Rogue really hurt the Barb or Rogue in a party that also contains a Cleric and Wizard? No, as long as the class has a role to fulfil. The tier system creates this false dichotomy that implies that it is nigh-impossible for lower tier classes to be effective in the party containing higher tier classes - this is a blatantly WRONG statement that only retards believe. Last time I checked, D&D wasn't a competition amongst players to see how many EPIC MONSTERS I CAN PWN ALONE, so naturally something like a Wizard/Rogue duo won't be a cutthroat competition to see how many encounters they can solo, but will lead to synergy where each will fulfil their own party role, with the Rogue as the DPR, and Wizard as the BFC with utility split between them both.

so that the DM can make decisions about what to allow or ban on the basis of how good a class is

Too bad this statement is a crock of lies since the way the tier system determines power is inaccurate, false, misleading, and pointless, unless you're playing some sort of stupid solo dungeoncrawler akin to Test of Spite to show off your optimization-fu obviously to compensate for the lack of something else (probably social skills). Does being able to do a lot of things really matter that much in a D&D game, to the point where versatility is treated as the most important thing to have in a character? No, because you have a ****ING PARTY WITH YOU. All you need is to perform your role well and BAM you can be a valuable contributing member of your party. Yet everyday here some smartass says X class is better than Y because Y is just a one trick pony.

As a DM, if your party consists of a Rogue/Barb/Ranger/Wizard, your solution is not to nerf the Wizard to the level of the rest - the Wizard player won't have any fun, and the party will struggle even more and the other players might also have less fun. Your solution is to build better encounters that enable the rest of the party to contribute. YES you may need to ban some problem spells, but honestly if the Wizard player isn't a jackass he won't use them in the first place, and if the Wizard player is a jackass he'll still find a way to break the game. D&D has always been a game where the DM and the players aren't antagonistic forces, as they cooperate together to tell a story. Unless, of course, you're playing something like a dungeoncrawl, which I won't wrong (it isn't) but also isn't what most people play the game for.

Aside from that, the common complaints of class bias, lack of true determining factors, vague descriptions for each tier, and unequal optimization all still stand true.

It helps balance.

No it doesn't. Would you rather be the rogue doing massive damage in a Cleric+Wizard party, or the archivist in that party, forever being overshadowed by both?

I think people here see the utility to this

And this is probably the funniest thing I've read so far out of you. Conveniently forgetting to mention that you proposed the same idea in another tabletop forum, one with more experience, mind you, and got completely blown the **** out. """""People here"""""" see the utility in the tier system, because these are the same people that can't stand being told they're wrong to their faces and hence gathered in this hugbox to circlejerk about their own beliefs.

[eggynack]

I argued against these points back when they were made the first time. I don't see the point in arguing against them again. As for your specific and different complaints about versatility versus power, the tier system I've put together values both. If you're just powerful, you can be good. If you're just versatile, you can also be good. If you're both, you're obviously better. Also of note is the fact that Kaelik's specific example using fixed list casters no longer applies. Both classes are now listed in tier two.

[jywu98]

I argued against these points back when they were made against me the first time. I don't see the point in arguing against them again. As for your specific and different complaints about versatility versus power, the tier system I've put together values both. If you're just powerful, you can be good. If you're just versatile, you can also be good. If you're both, you're obviously better. Also of note is the fact that Kaelik's specific example using fixed list casters no longer applies. Both classes are now listed in tier two.

You don't see the point in arguing against them again because you have no good point to argue against them. And no, it is apparent that it doesn't. Because you don't grade the classes yourself, you outsource it to others, and expect them to follow your criteria. What the ****????? Cases in point: factotum is still tier 3, despite being not powerful in any way; artificer is tier 1, despite not being versatile in any way (you won't have anywhere enough time to pull off your crafting shenanigans in most campaigns don't even try to argue otherwise).

And once again how does being """"good""" even affect party dynamics (what actually matters)? It doesn't.

[eggynack]

You don't see the point in arguing against them again because you have no good point to argue against them.

That doesn't make any sense. If I had no points, I couldn't possibly have made them the first time such that I'd now be making them again.

And no, it is apparent that it doesn't. Because you don't grade the classes yourself, you outsource it to others, and expect them to follow your criteria. What the ****????? Cases in point: factotum is still tier 3, despite being not powerful in any way; artificer is tier 1, despite not being versatile in any way (you won't have anywhere enough time to pull off your crafting shenanigans in most campaigns don't even try to argue otherwise).

If I were putting the tier system together all on my lonesome, without input, I would have given those two classes those same scores. That other people, with differing opinions, have the ability to counteract my opinion, and convince other people (including me) of their opinions, is thus a thing I would think you would prize. If you think people are wrong, go out and convince them. You may succeed, or you may fail, but you're getting nowhere complaining to me about how a bunch of people disagree with you.

Anyway, this conversation is clearly not going anywhere positive. I'm not going to participate in it further. I'd advise others to do similar, because these kinds of crazy arguments already threatened to lock down this thread once.

[jywu98]

That doesn't make any sense. If I had no points, I couldn't possibly have made them the first time such that I'd now be making them again.


If I were putting the tier system together all on my lonesome, without input, I would have given those two classes those same scores. That other people, with differing opinions, have the ability to counteract my opinion, and convince other people (including me) of their opinions, is thus a thing I would think you would prize. If you think people are wrong, go out and convince them. You may succeed, or you may fail, but you're getting nowhere complaining to me about how a bunch of people disagree with you.

Anyway, this conversation is clearly not going anywhere positive. I'm not going to participate in it further. I'd advise others to do similar, because these kinds of crazy arguments already threatened to lock down this thread once.

Well, I guess you do truly belong to the GitP forums. What a disappointment, I thought you were smarter than the rest. How can I convince you when you ignore portions of my arguments? I've made a pretty in-depth post about the artificer in one of these stupid threads, and you basically never replied to any of my criticisms. What sort of discussion can we possibly have if you just ignore whatever you want?

[eggynack]

Well, I guess you do truly belong to the GitP forums. What a disappointment, I thought you were smarter than the rest. How can I convince you when you ignore portions of my arguments?

Convince me of what? That the tier system is fundamentally pointless? You probably can't convince me of that, or at the very least trying to do so here is rather off-topic. We're trying to build a tier system, not analyze the premises that would lead us to one in the first place. It would help your ability to convince me if you didn't act like points I've already responded to in the past are somehow iron clad and impossible for me to refute.

I've made a pretty in-depth post about the artificer in one of these stupid threads, and you basically never replied to any of my criticisms. What sort of discussion can we possibly have if you just ignore whatever you want?

Not every discussion has to happen with me. I'm not personally that invested in the artificer debate, because it's not a class I know a ton about. I'm not always going to be the most informed and open to discussion person in the room. These threads aren't, "Convince eggynack that he's wrong about various tierings." It's an open discussion between anyone in the community that wants to join in, and if you want to convince people, maybe try to convince them instead of assuming my response is somehow a ubiquitous and mandatory presence in every single conversation we're having. Or, to put it another way, in the context of the artificer debate, I am in no way special. I'm just another guy submitting his vote and making his arguments. Sometimes I do respond to stuff, more than most perhaps, just like everyone else, but it's not a thing you should necessarily expect.

[jywu98]

Convince me of what? That the tier system is fundamentally pointless? You probably can't convince me of that, or at the very least trying to do so here is rather off-topic. We're trying to build a tier system, not analyze the premises that would lead us to one in the first place. It would help your ability to convince me if you didn't act like points I've already responded to in the past are somehow iron clad and impossible for me to refute.

Honestly even if I didn't assume it I'm sure we'll just be going around in circles arguing X and Y over and over again. Whatever, I'm probably taking this too personally and that's my bad.

Not every discussion has to happen with me. I'm not personally that invested in the artificer debate, because it's not a class I know a ton about. I'm not always going to be the most informed and open to discussion person in the room. These threads aren't, "Convince eggynack that he's wrong about various tierings." It's an open discussion between anyone in the community that wants to join in, and if you want to convince people, maybe try to convince them instead of assuming my response is somehow a ubiquitous and mandatory presence in every single conversation we're having. Or, to put it another way, in the context of the artificer debate, I am in no way special. I'm just another guy submitting his vote and making his arguments. Sometimes I do respond to stuff, more than most perhaps, just like everyone else, but it's not a thing you should necessarily expect.

Man if this is another community retiering thread then why create a new one instead of adding to the already existing one. Do correct me if I'm wrong on this, I'm far too lazy to check post date but I'm pretty sure the other community tiering thread came first.

[eggynack]

Man if this is another community retiering thread then why create a new one instead of adding to the already existing one. Do correct me if I'm wrong on this, I'm far too lazy to check post date but I'm pretty sure the other community tiering thread came first.

It is a newer one, and was intended to fill a roughly similar role. I generated this for a few reasons. First, the original completely shut down vote changing or addition of any kind after the deadline (the first being especially problematic, because that makes it a poor survey of even those who were around within the time span), which was really hurting accuracy. Second, it was utterly discounting any ACF that wasn't tier impactful. Like, you wouldn't necessarily grant a tier increase on the basis of underdark knight paladin, but you might when also considering mystic fire knight and harmonious knight, especially all three on a single paladin. Thus, that ACF wouldn't get accounted for later on, when going through tier shifting ACFs (and maybe feats), and it wouldn't be considered in the original tiering, so it just gets arbitrarily wiped from the game. Third, Jormengand was weirdly attacking folks that argued against their procedure, particularly along the lines of those first two points, and never addressed any of those concerns. At this point, they've apparently decided to toss anyone that argues about procedural issues on their ignore list, which makes the voted on tierings partially a matter of who Jormengand doesn't take personal issue with. That I was one of those seemingly placed on said ignore list (along with maybe Beheld, eventually, I think) made the issue a bit personal, and was my final motivating factor, but all that other stuff had been broiling around the thread for awhile. I would have preferred to just make the existing thread, with all of its existing argument, more suited to the actual reality of the community and such, but that proved impossible.

Anyway, those were my original motivating factors. However, they weren't ultimately the only purposes that this newer thread served. If you look through the community tiering project, you'll see a lot of the arguments that have plagued the original tier system for awhile based on some issues with its construction. You get people thinking that we're primarily or only considering 20th level, or saying that beguilers are better than sorcerers power-wise but worse in terms of tier on the basis of the kinda wonky original definitions, or thinking that the capacity of two sorcerers to have different spells is somehow incredibly relevant to tiering. So, I fixed up the definitions to not include problematic prescriptive elements, aligned the underlying nature of the system to only the more logical reading, leaving aside the bits that contradict those bits, and generally set things up in a way that's not intrinsically problematic, assuming tiering is a thing you'd want to do. We're not tripping over ourselves from square one anymore, y'know? We get to trip over normal stuff, like how good classes actually are. Finally, I revamped the structure of the thread such that we get all these smaller thematically and/or mechanically linked threads that are way easier to parse than a 60-70 page thread done in alphabetical order.

What I have isn't necessarily perfect. I don't think I could possibly please everyone with any formulation of the tier system. But I think it's better. I think that, at the end of the day, we'll have a set of pretty close to accurate tiers in a reasonable structure with few to no blatantly contradictory or problematic elements beyond those intrinsic to tiering. It's better than some arbitrarily biased guy telling everyone what he thinks things look like, and better than a thread that cares more about fun voting times than about what's best for the tiers themselves (putting the notion of deadlines ahead of what people actually think and such).

[jywu98]

Thanks for giving a detailed response, but here are my main issues regarding what you're doing:

Obviously, the first one is that what you write, you really can't expect others to follow. Example: Rogue is great at DPR. Combat is supposed to have a heavy weightage in your new system. It still sits at tier 4, while the factotum which is good at nothing, decent at some stuff sits at tier 3. Maybe it's just me, but this highlights a sort of disconnect between what you want and what the forum thinks you want.

Secondly, you state that the 3 core problems are exploration, combat, and social interaction, without providing any level-appropriate examples of what this entails. You can use the examples listed in an SGT to make this point clearer, as the categories are broad and vague, which makes the tier classifications kind of moot (you basically listed 3 types of problems but down in tier classification talk about many problems). You don't even need to actually run the tests.

Thirdly, what does over all optimization levels mean? This statement is really useless when you don't give a baseline on what each op level entails. Once again, everyone has different perceptions. This problem has occured in previous tier arguments where people have argued Druids don't belong with the rest of the Big 3 due to the lack of gamebreaking spells. I'm not sure what one vague sentence is going to do to stop that.

Also I am quite curious to what you consider versatility. I know you admitted that you aren't familiar with artificers, but I'm pretty sure that outside of TO it is ridiculous to consider it versatile in any sense given the crafting times involved - if his starting equipment cannot solve the problem at hand, the campaign world isn't gonna wait while the artificer takes his own sweet time to craft his whatever. So why is he in a tier where the rest of the tier spends a maximum of 1 day to rectify this issue? Furthermore, no matter how hard you try, it is impossible to rank classes in such a broad matter. I'm probably beating a dead horse, but I'm going to use the example of the rogue and the factotum again. The former can do one thing well, but the latter can do a lot more things, but nowhere as effective as the rogue is for combat. How are you going to determine which is better?

Which comes to these points: Everyone plays each class differently and everyone DMs differently. There's nothing stopping a fighter from attacking the walls of a dungeon to bypass troublesome traps. Most people don't play like that, but that doesn't mean it's not a valid way to play a fighter. In fact, playing a fighter like that will probably give it a pretty decent boost to solving exploration problems, at least. Similarly, DMs run monsters differently. Shivering touch as long been considered broken because it can reliably disable a dragon in 2 casts. This is definitely a problem if the DM makes the dragon engage the party in close-range combat. However, a more unforgiving DM will just do strafing runs with its breath weapon, in which the encounter increases several levels in terms of difficulty. What I am basically trying to get at here is how are you able to make a tier system for D&D considering all these factors come together to determine the end result, which is party effectiveness. Like in economic models, you have to make a few of these factors constant, yet you don't state what should and should not be kept constant.

Also what does DM restrictiveness even mean? The statement is so open-ended and everyone has their own experiences. Again, at least provide some sort of baseline/example. If not typing that was a waste of space.

[Beheld]

Secondly, you state that the 3 core problems are exploration, combat, and social interaction, without providing any level-appropriate examples of what this entails. You can use the examples listed in an SGT to make this point clearer, as the categories are broad and vague, which makes the tier classifications kind of moot (you basically listed 3 types of problems but down in tier classification talk about many problems). You don't even need to actually run the tests.

That would be counter to his goals, since his goals are to rank classes based on the assumption that level appropriate doesn't exist and that being able to beat up two commoners and once and smell someone are meaningful abilities for level 10 characters. After all, his tiers are about "solving situations" not you know, level appropriate ones. Those are evil and bad.

[eggynack]

Obviously, the first one is that what you write, you really can't expect others to follow. Example: Rogue is great at DPR. Combat is supposed to have a heavy weightage in your new system. It still sits at tier 4, while the factotum which is good at nothing, decent at some stuff sits at tier 3. Maybe it's just me, but this highlights a sort of disconnect between what you want and what the forum thinks you want.

Not everyone necessarily agrees with that evaluation of either class. The tier three voters may be right or may be wrong (myself included), but I don't think it's really an issue with the form of the tier system. A differently written tier system would probably still have a lot of tier three voters. Maybe if I wrote, "Versatility is nearly irrelevant for the purposes of tiering, with power as the only truly pertinent metric," people would vote tier four or even five, but I don't really think that's a good measure of class quality.

Secondly, you state that the 3 core problems are exploration, combat, and social interaction, without providing any level-appropriate examples of what this entails. You can use the examples listed in an SGT to make this point clearer, as the categories are broad and vague, which makes the tier classifications kind of moot (you basically listed 3 types of problems but down in tier classification talk about many problems). You don't even need to actually run the tests.

My goal there was more to clarify that we are, in fact, considering non-combat situations as well as combat ones, than it was to strictly lay out the problem space. I could see adding in some examples or something, but I think that the idea is pretty well understood.

Thirdly, what does over all optimization levels mean? This statement is really useless when you don't give a baseline on what each op level entails. Once again, everyone has different perceptions. This problem has occured in previous tier arguments where people have argued Druids don't belong with the rest of the Big 3 due to the lack of gamebreaking spells. I'm not sure what one vague sentence is going to do to stop that.

The notion of a thing being optimized or not optimized is somewhat intrinsically defined. Less optimized means you make fewer good choices, put together the character less deliberately, and, if we consider the overall power level possibility space for a class, they'll land further to the left of that space. More optimized means the opposite, of course. What an exact optimization level means in rigorous fashion is incredibly tricky, but if you see one presented version of a class use their feats really well, or choose spells efficiently, and the presented version of the other class uses their feats and chooses spells poorly, then the analysis is not in accordance with the notion of fixed optimization. Obscurity of used game objects is a factor too, but it's one I presented somewhat separately. Should get accounted for though.

Also I am quite curious to what you consider versatility. I know you admitted that you aren't familiar with artificers, but I'm pretty sure that outside of TO it is ridiculous to consider it versatile in any sense given the crafting times involved - if his starting equipment cannot solve the problem at hand, the campaign world isn't gonna wait while the artificer takes his own sweet time to craft his whatever. So why is he in a tier where the rest of the tier spends a maximum of 1 day to rectify this issue? Furthermore, no matter how hard you try, it is impossible to rank classes in such a broad matter. I'm probably beating a dead horse, but I'm going to use the example of the rogue and the factotum again. The former can do one thing well, but the latter can do a lot more things, but nowhere as effective as the rogue is for combat. How are you going to determine which is better?

Versatility is a pretty straightforward idea, I think. All else being equal, given a set of problems, a more versatile class can apply themselves at a given level of power to more of that problem space. I don't think your disagreement here is one with the way I constructed the tier list. Instead, I think you're finding yourself disagreeing with the way people evaluate these classes, and assuming that difference in evaluation is coming from some definitional thing. It seems unlikely that it does come from definitions. The primary definition of tier three is, "Better than classes in tier four, roughly equal to some classes in tier three, and worse than classes in tier two." If people think factotums are better in the context of my system, then they also think they're better outside that context.

Which comes to these points: Everyone plays each class differently and everyone DMs differently. There's nothing stopping a fighter from attacking the walls of a dungeon to bypass troublesome traps. Most people don't play like that, but that doesn't mean it's not a valid way to play a fighter. In fact, playing a fighter like that will probably give it a pretty decent boost to solving exploration problems, at least. Similarly, DMs run monsters differently. Shivering touch as long been considered broken because it can reliably disable a dragon in 2 casts. This is definitely a problem if the DM makes the dragon engage the party in close-range combat. However, a more unforgiving DM will just do strafing runs with its breath weapon, in which the encounter increases several levels in terms of difficulty. What I am basically trying to get at here is how are you able to make a tier system for D&D considering all these factors come together to determine the end result, which is party effectiveness. Like in economic models, you have to make a few of these factors constant, yet you don't state what should and should not be kept constant.

Well, we're essentially fully discounting things that don't come from class. Fighters can stab dungeon walls, but commoners aren't that much worse at doing so, so it's not a big factor. things are relevant to the fighter within the context of a comparison if and only if the fighter uses those things better than the class being compared to. If a specific mode of player skill can give the fighter an advantage that a commoner won't get, then that could maybe be relevant. As for how monsters are run, I dunno that it's that big of an issue. You could always just consider DM toughness as a modifier on encounter difficulty. Also, I don't think these really swingy shivering touch or no shivering touch situations come up that often.

Also what does DM restrictiveness even mean? The statement is so open-ended and everyone has their own experiences. Again, at least provide some sort of baseline/example. If not typing that was a waste of space.

I wanted to allow in a broader notion of optimal or suboptimal than was being conveyed. Like, you could not be using greenbound summoning because of its obscurity or because you aren't optimizing that much. You could also not be using it because the DM doesn't allow it in, which isn't necessarily an optimization thing. I'm saying it is more or less an optimization thing, because the end state of the character is the same. More clarification than a completely new rule. It was one of several suggested additions awhile back, and that one seemed reasonable. I don't like to set things up as overly structured, because precise definitions lead to the problems that plagued the original tier system, but when I can clarify without generating prescription, doing so seems worthwhile.

[jywu98]

Not everyone necessarily agrees with that evaluation of either class. The tier three voters may be right or may be wrong (myself included), but I don't think it's really an issue with the form of the tier system. A differently written tier system would probably still have a lot of tier three voters. Maybe if I wrote, "Versatility is nearly irrelevant for the purposes of tiering, with power as the only truly pertinent metric," people would vote tier four or even five, but I don't really think that's a good measure of class quality.

Not much to say here, except for theoretical versatility and practical versatility are different. An artificer is a glaring example of this.

[QUOTE[My goal there was more to clarify that we are, in fact, considering non-combat situations as well as combat ones, than it was to strictly lay out the problem space. I could see adding in some examples or something, but I think that the idea is pretty well understood. [/QUOTE]
Fair enough.

The notion of a thing being optimized or not optimized is somewhat intrinsically defined. Less optimized means you make fewer good choices, put together the character less deliberately, and, if we consider the overall power level possibility space for a class, they'll land further to the left of that space. More optimized means the opposite, of course. What an exact optimization level means in rigorous fashion is incredibly tricky, but if you see one presented version of a class use their feats really well, or choose spells efficiently, and the presented version of the other class uses their feats and chooses spells poorly, then the analysis is not in accordance with the notion of fixed optimization. Obscurity of used game objects is a factor too, but it's one I presented somewhat separately. Should get accounted for though.

Do you realize you've spent a lot of words saying nothing? What constitutes really well? What is your basis of comparison to see if a class is high-op or low-op?

Versatility is a pretty straightforward idea, I think. All else being equal, given a set of problems, a more versatile class can apply themselves at a given level of power to more of that problem space.

Yes, except all else being equal doesn't hold ever, certainly not for a comparison between a factotum and a rogue.

Well, we're essentially fully discounting things that don't come from class. Fighters can stab dungeon walls, but commoners aren't that much worse at doing so, so it's not a big factor. things are relevant to the fighter within the context of a comparison if and only if the fighter uses those things better than the class being compared to. If a specific mode of player skill can give the fighter an advantage that a commoner won't get, then that could maybe be relevant.

Isn't this you admitting the artificer is wrongly tiered? Nothing that they do can't be done by other classes (infusions are negligible).

As for how monsters are run, I dunno that it's that big of an issue. You could always just consider DM toughness as a modifier on encounter difficulty. Also, I don't think these really swingy shivering touch or no shivering touch situations come up that often.

This actually matters quite a lot, as the difficulty swing that comes up can be quite significant. Tucker's Kobolds and Dragons are pretty easy examples to show what a bit of tactics on the DM's part can screw up a party. It also changes how classes fare quite a bit, as fights get more difficult the risk/reward of gimmicks diminishes significantly, easily separating classes that seem to be versatile from classes that actually are versatile.

I wanted to allow in a broader notion of optimal or suboptimal than was being conveyed. Like, you could not be using greenbound summoning because of its obscurity or because you aren't optimizing that much. You could also not be using it because the DM doesn't allow it in, which isn't necessarily an optimization thing. I'm saying it is more or less an optimization thing, because the end state of the character is the same. More clarification than a completely new rule. It was one of several suggested additions awhile back, and that one seemed reasonable.

This seems awfully subjective to me. All based on individual perception.

I don't like to set things up as overly structured, because precise definitions lead to the problems that plagued the original tier system, but when I can clarify without generating prescription, doing so seems worthwhile.

No, the reason the original tier system failed because it had no set criteria. It was vague and self-contradictory, and therefore is pointless. There was nothing precise about how JaronK explained the tiers, as what he said allowed for an open-ended interpretation that allowed him to slot whatever classes he desired to whatever tier. And so far, it seems like this hasn't changed.

[eggynack]

Do you realize you've spent a lot of words saying nothing? What constitutes really well? What is your basis of comparison to see if a class is high-op or low-op?

There's not necessarily a precise definition. It's the difference between, "I'm going to deliberately pick a pile of very powerful and synergistic feats from multiple source books," and, "I'm gonna take anything that looks cool, without much care for power." I think we are reasonable at assessing such things on a game object by game object basis, though it can be tricky. I suppose it could be considered in a somewhat mathematical fashion. Meaning that, while accounting for obscurity as a factor, you stick every possible build for a class onto a bell curve, where each build has some measure of power that we can theoretically determine, and then the optimization level is determined by how many standard deviations to the right or left you are. It's a rather rigorous definition. It's just difficult to apply in that rigorous form in the same way that we might apply the more intuitive definition.

Yes, except all else being equal doesn't hold ever, certainly not for a comparison between a factotum and a rogue.

Didn't say it does hold in this case (though I'd assert that some classes are simultaneously more powerful and more versatile than others). The factotum is arguably less powerful, which means that the power it applies to that greater number of challenges is less than what the rogue applies to each of its few challenges. The core questions, then, are how much less powerful and how much more versatile.

Isn't this you admitting the artificer is wrongly tiered? Nothing that they do can't be done by other classes (infusions are negligible).

It's not about doing things differently. When you assess tools that come directly from class, ones other classes aren't getting because they aren't the same class, you look at those tools directly. You don't need to fix anything, or take anything away from the game object. Where this sort of thing becomes important is in looking at things from outside the class. So, we could look at something like crafting. While a number of classes can craft, artificers have increase utility in this area. When comparing the artificer to, say, a wizard, some, but not all, of the artificer's crafting ability gets cancelled out. The neat class features based on crafting are still relevant because they represent additional utility in the comparison.

This actually matters quite a lot, as the difficulty swing that comes up can be quite significant. Tucker's Kobolds and Dragons are pretty easy examples to show what a bit of tactics on the DM's part can screw up a party. It also changes how classes fare quite a bit, as fights get more difficult the risk/reward of gimmicks diminishes significantly, easily separating classes that seem to be versatile from classes that actually are versatile.

It matters to how hard the challenges are, but I feel like we'd be able to call those tactics using situations somehow higher CR than the versions that don't use tactics, as it were.

This seems awfully subjective to me. All based on individual perception.

Sure. My point in the inclusion was that, if one character is crappy because the player is bad at optimizing, and the other is equally bad because of DM permissiveness, it doesn't matter all that much. Optimization here is more of an end-state than process oriented metric. This is in line with what I was talking about above. You imagine 10,000 artificers, without much care for why they are the way they are, stick them all on a big ol' bell curve, and you have a reasonable definition for what the class looks like at various optimization levels.

No, the reason the original tier system failed because it had no set criteria. It was vague and self-contradictory, and therefore is pointless. There was nothing precise about how JaronK explained the tiers, as what he said allowed for an open-ended interpretation that allowed him to slot whatever classes he desired to whatever tier.

I don't really agree, aside from the self-contradictory part. How the tiers are explained shouldn't matter that much. The main mode of evaluation should be whether a class is better than another class. Bit vague, of course, but what I mean is that we shouldn't wind up spending all of our time figuring out what the hell a "nuke" is, or how many of them a given class has.

[eggynack]

Just created the new thread hereabouts for battle dancer, monk, and mountebank. Also, because I haven't noted this in the home base thread yet, the mode calculation is currently rounding everything to the nearest .5, such that the wide variety of fractional votes don't get lost for the purposes of that calculation. Don't want them 4.3's rendered completely meaningless, y'know? That doesn't apply to anything besides mode, so if you want your vote to get counted in the count statistic, it's currently gotta be integer. Don't much want to change that, cause doing so would either mean pretending a 3.4 is a 3, or putting in five extra count rows, and it's not really an important statistic for any tiering purpose. Dunno if anyone actually cares about any of this information, but it was super annoying to implement, so it is information you shall receive. Anyway, I'ma update the various threads to reflect the existence of the new thread now.

Edit: Just added soulknife to the new thread Seemed to fit my overall theming, and while I've talked about how I'm mostly grouping psionics together, this makes a lot of sense as one of the exceptions to that rule.

[Troacctid]

I'm starting to think it might be a good idea to do a greater number of classes in each thread. With a separate thread for each group, I worry that someone who wants to vote in all of them will end up bumping a dozen threads to the front page at once, which seems not great?

[eggynack]

I'm starting to think it might be a good idea to do a greater number of classes in each thread. With a separate thread for each group, I worry that someone who wants to vote in all of them will end up bumping a dozen threads to the front page at once, which seems not great?

It's a thing that's happened a few times, and it's not ideal. Might make sense to trend closer to the four and five side of things than the three side, though some of the threads really do make a lot of sense in three form. There are problems attendant to going above six, however, so pushing it too high could be a problem. Thread title length is something of a limitation, though I suppose not a total one. Information compression exists. Not my preferred outcome though. Also, one or two conversations tend to dominate each thread, so minimizing class quantity means that that hits less focusing on classes less. Like, the very possible crazy monk discussion I may have unleashed just now simply by putting monk and tier in the same thread title may wash away some possible discussion for the other three classes, but at least it's not killing the discussion of six or seven classes. Also, again, I've generally been structuring these in a way that comports with some sorta theming, which I think makes sense from a discussion perspective. Sometimes really limits the number of classes I can include. Like, what am I gonna add to the ToB thread? I'm just glad I came up with soulknife for this one, given this issue. I think it's really suited.

[Troacctid]

You could roll ToB in with ToM. Tome-a-Palooza.

[eggynack]

You could roll ToB in with ToM. Tome-a-Palooza.

Not a terrible plan. Might consider something along those lines when running future threads, linking together reasonably related threads where I would have kept them apart otherwise. For example, it was suggested that I do a big warlock/dragon shaman type thread, which was four classes, but then I noted that that could be split into a pair of two threes by incorporating two other classes. Might be worth putting that all together for a six class majig. Not sure if that's the direction I want to go, cause the other issues are still a thing, but it's worth thinking about.

[Kurald Galain]

The Rankings

It would really help if you could sort this, either alphabetically or by tier.

Aside from that, for the (admittedly few) classes where there really isn't a clear consensus on the tier (like the Fighter and the CW Samurai), I don't think it's such a good idea to just list them as whichever option gets 51% of the vote. It would be a better representation of the threads to list them as (e.g.) "four or five, contested" or "between four and five" or something like that; $.02

[eggynack]

It would really help if you could sort this, either alphabetically or by tier.

Yeah, maybe.

Aside from that, for the (admittedly few) classes where there really isn't a clear consensus on the tier (like the Fighter and the CW Samurai), I don't think it's such a good idea to just list them as whichever option gets 51% of the vote. It would be a better representation of the threads to list them as (e.g.) "four or five, contested" or "between four and five" or something like that; $.02

I was planning to do stuff along these lines more towards the end. In particular, I'm planning to have an ordering by tier that specifically calls out the classes closest to the adjacent tiers as the line between those tiers. It's a bit premature to do something like that, or indeed any ordering by tier, because the ordering could change at any time. Dunno about other folks, but I tend to put a lot of stock in the specific values given by the spreadsheet, over and above the basic stuff I put in the main posts.

[GilesTheCleric]

Is it worth having a range field along with median, mode, and mean? It might be useful to see which classes have the largest range of voted/ potential tiers, particularly for things like the Monk or Soulknife like we're voting on now.

[eggynack]

Is it worth having a range field along with median, mode, and mean? It might be useful to see which classes have the largest range of voted/ potential tiers, particularly for things like the Monk or Soulknife like we're voting on now.

Min/Max might make more sense. You get range data, but also, y'know, where the edges of that range are. Also, range is weirdly not a function, far as I can tell.

Edit: Just realized that the min range will have to use the max function and vice versa. That's vaguely amusing.

[Jopustopin]

Min/Max might make more sense. You get range data, but also, y'know, where the edges of that range are. Also, range is weirdly not a function, far as I can tell.

Edit: Just realized that the min range will have to use the max function and vice versa. That's vaguely amusing.

We wouldn't want "range." Standard Deviation is more useful.

[eggynack]

We wouldn't want "range." Standard Deviation is more useful.

I dunno. I kinda like knowing how high or low people were willing to go on a given class. I do love standard deviation though. Was planning to add stuff like that later, that, variance, maybe some visualizations, whatever else we come up with, in some variety of conclusion thread when everything's over and done with, but it might be worth adding now. I'ma toss it in, but I gotta say, the section is getting a bit dense. Below the median and over the mode makes sense. Mode is probably the least useful of the stats, and while it'd also work to stick standard deviation right under mean, cause that's what it's based on, median is probably more important. Might also make sense to swap mean and median, such that I can get deviation right under mean, but mean has been our major statistical object, so that could be less than ideal.