T2+ Mundane Class: What would it take?

[NichG]

Let me see if I understand you corretly, NichG:
You are saying that the rogue cannot be T2+, since there are classes like bard that can do the same, but more easily (with less resources of their class abilities), plus more game-breaking stuff ("nukes" in JaronK's definition), and classes which are again 1-2 tiers higher, due to even more ease at nukes. Correct?

Three more objections:
1) I am not talking a rogue becoming pun-pun here, but just maxing some of his class skills and using his sneak attack ability (basics of playing that class, actually, no big optimization). If that is not considered in assessing a class, what is?
2) JaronK himself provided the definition for T2 that I quoted above. 10 nukes=T2. So is he wrong with this definition, and 10 nukes mean actually T4?

I'd argue you've provided only one 'nuke', which requires a fairly intense focus of resources. In order for Diplomacy to be a nuke at all, you have to hit the epic chart, which means reliably hitting DC 60 Diplomacy checks (50 to turn helpful into fanatic, and +10 to do it in a single round). Until you can do that, 'Helpful' isn't actually all that nukey if you look at the strict definition - its certainly insufficient to actually get people to do anything specific for you. In order to actually hit that DC 60 check with regularity as a Rogue, you're going to be pumping most of your WBL into it at least for the first 10 levels or so. Bluff-o-mancy and Diplomancy are the same nuke. Both are of the 'get people to take the action I want them to' variety. The Bluff nuke is weaker because it explicitly has a 1 round duration. So its basically a skill-based, non-repeatable Daze; nice, but not really deserving of a separate 'nuke' label.

The other potential 'nuke' is Sleight of Hand. The ability to take items from someone without a counter is pretty potent. But actually, there are many counters to it. The way to see this is, again, the epic table. A sheathed weapon is explicitly DC 50 rather than DC 20, and so storing things in the form of sheathed weapons helps protect them from being snatched by Sleight of Hand (e.g. wand chambers in a weapon, rather than keeping the wand separate). Similarly, there is no Sleight of Hand check DC for stealing things that are not 'small objects', so making objects part of larger objects protects them (e.g. build things into your clothing or armor). So that's not really a nuke.

UMD is not a 'nuke' by itself - its dependent on being able to afford various things with your WBL. So you can nova, but you can't use the powers you get from UMD with any sort of long-term reliability. E.g. Shapechange is going to cost you 4kgp each time you call on it using UMD, which is about 1/85th of your total WBL at the level at which a wizard would get it. That means, if a Wizard of that level, gets 2 9th level slots a day, you can pretend to be a very specialized wizard for about a month before the money - all of it, assuming you have no gear at all aside from scrolls of Shapechange - runs out. The Wizard also gets to do other things too, like Gate (9kgp per casting) or Wish (29kgp per casting) or all the rest of his spell levels worth of slots. In any event, this trick is exactly as expensive for an Expert, Commoner, etc. So in terms of comparative measures, this isn't specifically something the rogue gets over other classes.

Sneak attack certainly isn't a nuke whatsoever. As far as damage potential goes, its not very high up there, and has gaping holes in its applicability. A TWF/SA-focused Rogue will be shut down by things that avoid being Flatfooted/denied Dex, as well as things that are immune to SA/Crits (about half of the base Types in the game). There are some ways to reduce that immunity (feats and weapon crystals), but it still leaves you doing pretty mediocre damage (20d6 SA times 8 attacks per round is about 500 damage if they all hit, and requires the rogue to be Lv20 where boss-level things average about 300-400hp; that's a pretty close shave - 20% miss chance, a reasonable AC, etc means that the target survives. Compare with an ubercharger who, by those levels, can dish out multiple thousands of damage per round. So this isn't really a nuke).

So you've got a class that can specialize in a single trick which may or may not be game-breaking (really, it requires more than just 'max ranks in Bluff!' to be so, to the extent that we're talking about a specific build now instead of a general feature of the class). Compare this to, e.g., a Dungeoncrasher (a T3 Fighter ACF). A Dungeoncrasher can basically nuke anything that it can attack, but that's all it can do. This makes a Dungeoncrasher T3, as a specific ACF/Build that makes good use of extra materials. But this does not make the fighter itself T3, because again, the Tier system generalizes over many builds, not just the specifically optimized ones.

So, much in the same way, perhaps you can build a specific Rogue that can keep up with T3 characters by really focusing on Diplomancy (much like I could build a specific Commoner that would do the same), but in general the rogue will be a lower tier because aside from that one highly focused trick with a specific item loadout, most rogues actually won't be able to easily hit the epic Diplomacy chart DCs.

3) I still am not convinced that the bard really can do diplomancy/buffomancy so much better (see also below), and also even if he could - what about the other 3 nukes of the rogue that I presented? Can the bard, at the same time be better at diplomancy/buffomancy, AND also beat the rogue at sleight of hand tricks, sneak attack touch attacks and UMD? And then be able to do even more nukes?
(...3b) Why would the bard then even be considered only T3? ... but that is exceeding this thread, probably)

See above. I think you only have one 'nuke'.

Also, wouldn't using these nukes for the rogue also mean boosting various other aspects of his class?
For instance, skill mastery applies to various skills. The rogue can take it four times. Each time it applies to a minimum of 3 skills, so 12 in total. Devoting one of these skills to bluff is just 1/12th of just one of his many class abilities. I do not really see the big bard advantage here.

Skill mastery really doesn't do much for you. It lets you take 10. Which doesn't change the average result. So this is at best perturbative, and in no way, shape, or form really counts as a 'nuke'.

Another example is UMD. The rogue gets it for the glibness effect, costing just a fraction of his wbl, a fraction of his total skill points, and also boosting his overall (fake) casting ability.

In my view, the rogue so far makes a solid T2 mundane class.

Again, see above for the UMD analysis.

[Psyren]

Playing without any key items available in the campaign would lead to clerics having to get along without holy symbol, for instance, or wizards unable to get a familiar (for lack of 100gp worth of materials to summon it), which would be silly.

There is an order of magnitude between saying "you can't buy a holy symbol" and saying "you can't buy a wand of polymorph, three scrolls of locate object and a staff of greater teleport." The DM is under no obligation to provide these specific items to you just so your rogue can solve those particular problems. And s/he is definitely under no obligation to say things like "here is a CL 12 wand of polymorph so you can become a war troll."

Also, your examples don't work - even if a holy symbol isn't provided, the cleric starts with one, or can carve one out of wood, or just summon one with an orison. Similarly, the wizard starts with a spellbook, and can also scribe on his own body, and that's before we get into easy-bake wizard again. The DM has to actively deny these things, not provide them.

[Segev]

Technically, for T2, you only need one "nuke," as long as that "nuke" can be shown to provide game-breaking potential and be geared to solve arbitrary problems.

Heck, if that nuke can solve ANY problem through correct application, that's enough for T1! ("Can do anything, often better than specialists in the fields," being T1, if Diplomancy can achieve anything, it is, by itself, T1.)

[NichG]

I would say that Diplomancy cannot actually do everything. The 'Fanatic' state is explicitly subject to immunity to Mind-affecting, so you can't do the 'I get a god to do it for me' trick to borrow the abilities of more powerful characters. By the time you're hitting the big leagues, everyone relevant will be immune to Fanatic status. The language behind 'Helpful' is vague enough that its not actually useful in situations that are threatening or costly to the beings whose aid you want to request to the extent of leading to their probable demise or the destruction of their life (e.g. Helpful explicitly stops at the point where an NPC would be giving up their life for your benefit; that's explicitly the Fanatic status)

[Psyren]

In addition, optimizing your diplo check enough to achieve fanatic status will involve magic at some point. Pre-epic, it is simply not possible to hit a DC 150 check by purely mundane means.

[aleucard]

In general, unless if the class has skill-specific abilities that matter significantly or otherwise key off them (but then it's not the skill itself that matters), skills aren't counted for tier rankings. The reason being; having it as a class skill generally means the difference of 9 ranks maximum for the levels in consideration for tier ranking (5-15 is the range, last I checked). For the uses that make skills definable as a nuke, such a difference is small to the point of near-meaninglessness. The vast majority, if not the entirety of methods required to get skill checks that high bear no heed whatsoever to if the skill is a class skill or not. As such, provided that sufficient ranks are attained for whatever methods you are using (1 rank to count as trained seems to be enough for most relevant things, and some skills don't even need that much), it doesn't matter what your class is, you can do it anyway. Again, things that are universal to all PC's like this and WBL (save for the ability to apply either, which is determined by class abilities) don't count.

[Ivanhoe]

NichG, thanks, your new suggestions make way more sense to me. Still, I am not entirely convinced ...

I'd argue you've provided only one 'nuke', which requires a fairly intense focus of resources. In order for Diplomacy to be a nuke at all, you have to hit the epic chart, which means reliably hitting DC 60 Diplomacy checks (50 to turn helpful into fanatic, and +10 to do it in a single round). Until you can do that, 'Helpful' isn't actually all that nukey if you look at the strict definition - its certainly insufficient to actually get people to do anything specific for you. In order to actually hit that DC 60 check with regularity as a Rogue, you're going to be pumping most of your WBL into it at least for the first 10 levels or so.

Who needs epic diplomacy DC? Convert hostile creature to unfriendly (avoiding thus combat), is DC 40 with a rushed check. That ends whole encounters, pretty campaign-shaping to me. And a rogue gets to make that check reliably much faster than anyone else, thanks to class skill and skill mastery.

Bluff-o-mancy and Diplomancy are the same nuke. Both are of the 'get people to take the action I want them to' variety. The Bluff nuke is weaker because it explicitly has a 1 round duration. So its basically a skill-based, non-repeatable Daze; nice, but not really deserving of a separate 'nuke' label.

No. Convincing someone of wrong information is a completely different ballpark to even helpful attitude. You can change whole campaigns with a high enough bluff check (BBEG: "What? My main minion is a secret enemy just waiting to backstab? Thanks for that info" (kills 2nd BBEG).
A wrong information runs way deeper than a 1 round daze. No duration is given for bluff, to, so the bluff holds until something happens that contradicts the bluff sufficiently (DM fiat).

The other potential 'nuke' is Sleight of Hand. The ability to take items from someone without a counter is pretty potent. But actually, there are many counters to it. The way to see this is, again, the epic table. A sheathed weapon is explicitly DC 50 rather than DC 20, and so storing things in the form of sheathed weapons helps protect them from being snatched by Sleight of Hand (e.g. wand chambers in a weapon, rather than keeping the wand separate). Similarly, there is no Sleight of Hand check DC for stealing things that are not 'small objects', so making objects part of larger objects protects them (e.g. build things into your clothing or armor). So that's not really a nuke.

You are reading the epic sleight of hand wrongly. It is not to take a sheathed weapon from someone that is DC 50, but "Lift a sheathed weapon from another creature and hide it on the character’s person, if the weapon is no more than one size category larger than the character’s own size." The complete action is DC 50, it is not an enumeration of actions. That is a huge difference.
So the typical free action wealth amassing abuse appears still open. Sheathed weapons are not protected from stealing, and basically everything somebody wears or carries can be taken by the RAW text of the skill description (the table providing example DC for small objects, which is the max size of most of what medium-people have, excepting large weapons and armour). You just steal them with DC 40 without the other creature able to do anything about it (other than being incorporal, for instance). I'd of course houserule it to less silliness, but it remains a nuke.

UMD is not a 'nuke' by itself - its dependent on being able to afford various things with your WBL. So you can nova, but you can't use the powers you get from UMD with any sort of long-term reliability. E.g. Shapechange is going to cost you 4kgp each time you call on it using UMD, which is about 1/85th of your total WBL at the level at which a wizard would get it. That means, if a Wizard of that level, gets 2 9th level slots a day, you can pretend to be a very specialized wizard for about a month before the money - all of it, assuming you have no gear at all aside from scrolls of Shapechange - runs out. The Wizard also gets to do other things too, like Gate (9kgp per casting) or Wish (29kgp per casting) or all the rest of his spell levels worth of slots. In any event, this trick is exactly as expensive for an Expert, Commoner, etc. So in terms of comparative measures, this isn't specifically something the rogue gets over other classes.

Sure the rogue is worse at T1 stuff than the T1 classes. That's why I put it at T2. It can do the super shapechange, gate, etc. tricks but not as often. The raw power, though, will be there for the key encounters (if the often quoted 14 encounters/level is any guideline). And as said before, the rogue (as the bard) get it as class skill, but many others not. So it remains a rogue nuke.

Sneak attack certainly isn't a nuke whatsoever. As far as damage potential goes, its not very high up there, and has gaping holes in its applicability. A TWF/SA-focused Rogue will be shut down by things that avoid being Flatfooted/denied Dex, as well as things that are immune to SA/Crits (about half of the base Types in the game). There are some ways to reduce that immunity (feats and weapon crystals), but it still leaves you doing pretty mediocre damage (20d6 SA times 8 attacks per round is about 500 damage if they all hit, and requires the rogue to be Lv20 where boss-level things average about 300-400hp; that's a pretty close shave - 20% miss chance, a reasonable AC, etc means that the target survives. Compare with an ubercharger who, by those levels, can dish out multiple thousands of damage per round. So this isn't really a nuke).

I guess the SA-immune creatures are less than a third of the opponents in the game, and as you said, it is irrelevant since the rogue has so many possibilities to bypass them.
By level 20 a total 500 damage is not really such a big deal, but the touch attacks with sneak damage tend to end encounters quickly in low-mid levels. You can also get the special ability to add STR damage, ending encounters even more quickly. It is not as big of a nuke, as the diplomacy thing, but still big. And comparing doing 500 damage to 1000s of damage when only 300-400 damage is needed is not really useful. Also, uberchargers need to charge (straight line, visible targets), and I'd even doubt that they can dish out more damage per round than a similarly maxed rogue.

So you've got a class that can specialize in a single trick which may or may not be game-breaking (really, it requires more than just 'max ranks in Bluff!' to be so, to the extent that we're talking about a specific build now instead of a general feature of the class). Compare this to, e.g., a Dungeoncrasher (a T3 Fighter ACF). A Dungeoncrasher can basically nuke anything that it can attack, but that's all it can do. This makes a Dungeoncrasher T3, as a specific ACF/Build that makes good use of extra materials. But this does not make the fighter itself T3, because again, the Tier system generalizes over many builds, not just the specifically optimized ones.

This dungeoncrasher T3 comparison confuses me even more. It can do one nuke, I showed the rogue can do 5 (minimum). The sneak attack nuke is on par with the damage the dungeoncrasher does, but can do so melee and ranged, and vs more opponents (if he so chooses), so is even better at it.
And again, "max ranks in bluff" is a rogue-specific thing. Other classes can also do it, but not all.

So, much in the same way, perhaps you can build a specific Rogue that can keep up with T3 characters by really focusing on Diplomancy (much like I could build a specific Commoner that would do the same), but in general the rogue will be a lower tier because aside from that one highly focused trick with a specific item loadout, most rogues actually won't be able to easily hit the epic Diplomacy chart DCs.

Are you saying that a commoner could do the same 5 nukes I showed for the rogue? Meaning a T6 class able to fulfill JaronK's definition for T2 class? That would completely turn upside down the tier system as I understood it so far.

Skill mastery really doesn't do much for you. It lets you take 10. Which doesn't change the average result. So this is at best perturbative, and in no way, shape, or form really counts as a 'nuke'.

Skill mastery alone is not a nuke. But it contributes to the performance of one, say diplomancy.
That 40 DC from above for avoiding all hostile encounters? Well, with taking 10 and adding 13 ranks, plus synergies from bluff, sense motive and local history, plus 11 other bonuses from items and mixed stuff, you can reliably get that nuke at level 10. A commoner cannot do that.

[eggynack]

None of these things are really nukes, is the issue. None of these things can really stand up to the versatility or power provided by stuff like silent image at first, polymorph at 4th, control winds at 5th, or obviously shapechange at 9th. Even something like the rogue's use of WBL, as occasionally powerful as it is, is both more locked in and less consistent than what a sorcerer does. The rogue has some nice tricks, and as is the case with most classes, the ceiling is higher than the tier listing, but this isn't a sorcerer by any means, or even a bard.

You can keep redefining what nuke means until things fall into place for a given character (CW samurai is tier two. Look at its TWF nuke, and its intimidate nuke, and its improved initiative nuke. That's three nukes right there.), but we're still looking at a class that's generally worse than the classes a tier above it, generally better than the classes a tier below it, and much worse than classes two or three tiers above it. By the time you hit the point that these things can be tentatively be called "nukes", the sorcerer has upped its game to the point that these things are just a pale imitation of his full power.

[NichG]

NichG, thanks, your new suggestions make way more sense to me. Still, I am not entirely convinced ...

Who needs epic diplomacy DC? Convert hostile creature to unfriendly (avoiding thus combat), is DC 40 with a rushed check. That ends whole encounters, pretty campaign-shaping to me. And a rogue gets to make that check reliably much faster than anyone else, thanks to class skill and skill mastery.

This is actually very limited - more limited than an ubercharger if you really do the analysis. For example, here are the situations in which a DC 40 rushed Diplomacy check will not actually end the encounter:

- Unintelligent enemies
- Enemies who have no language you can communicate in (Diplomacy is language-dependent, and Tongues doesn't let you talk to e.g. plants)
- Any encounter with multiple enemies. You can perhaps neutralize one enemy per round with Diplomacy here, same as a Dungeoncrasher, but just making one guy unfriendly will not end the encounter. If there's a strong chain of command and you make the leader unfriendly, that doesn't mean he'll stop the others from attacking. And if you have an encounter without a strong chain of command, then it doesn't do anything aside from taking that single guy out of the mix.

Also, in no way does a rogue get to make the check reliably 'much faster than anyone else'. Just as a starter, Bards, Factotums, and Clerics all have it as a class skill and each get their own way of adding a big chunk to the skill check. Skill Mastery is a late-comer to this and doesn't really add anything.

No. Convincing someone of wrong information is a completely different ballpark to even helpful attitude. You can change whole campaigns with a high enough bluff check (BBEG: "What? My main minion is a secret enemy just waiting to backstab? Thanks for that info" (kills 2nd BBEG).[
A wrong information runs way deeper than a 1 round daze. No duration is given for bluff, to, so the bluff holds until something happens that contradicts the bluff sufficiently (DM fiat).

Actually, Bluff explicitly says the following:

A successful Bluff check indicates that the target reacts as you wish, at least for a short time (usually 1 round or less) or believes something that you want it to believe. Bluff, however, is not a suggestion spell.

So, one round or less. That might be enough to cause a bit of chaos when the BBEG begins to turn on his lieutenant, but then a round later he's free to say 'Oh, wait a minute, thats dumb, what are we doing? Kill the messenger!'.

You are reading the epic sleight of hand wrongly. It is not to take a sheathed weapon from someone that is DC 50, but "Lift a sheathed weapon from another creature and hide it on the character’s person, if the weapon is no more than one size category larger than the character’s own size." The complete action is DC 50, it is not an enumeration of actions. That is a huge difference.
So the typical free action wealth amassing abuse appears still open. Sheathed weapons are not protected from stealing, and basically everything somebody wears or carries can be taken by the RAW text of the skill description (the table providing example DC for small objects, which is the max size of most of what medium-people have, excepting large weapons and armour). You just steal them with DC 40 without the other creature able to do anything about it (other than being incorporal, for instance). I'd of course houserule it to less silliness, but it remains a nuke.

So, as I said, attaching everything in an inseparable or locked way to a medium object such as your clothing will make them immune to Sleight of Hand. A bit of rope and a 25gp lock basically makes this nuke go away entirely.

I guess the SA-immune creatures are less than a third of the opponents in the game, and as you said, it is irrelevant since the rogue has so many possibilities to bypass them.
By level 20 a total 500 damage is not really such a big deal, but the touch attacks with sneak damage tend to end encounters quickly in low-mid levels. You can also get the special ability to add STR damage, ending encounters even more quickly. It is not as big of a nuke, as the diplomacy thing, but still big. And comparing doing 500 damage to 1000s of damage when only 300-400 damage is needed is not really useful. Also, uberchargers need to charge (straight line, visible targets), and I'd even doubt that they can dish out more damage per round than a similarly maxed rogue.

The problem is that that 300-400 damage is in optimal circumstances - you hit on every attack, there's no miss chance, etc. 50% of 300-400 damage isn't enough to one-shot most CR 20 monsters, and most CR 20 monsters will have defenses that can avoid at least half of a TWF Rogue's attack sequence.

If you don't think Uberchargers can out-damage a maxed out rogue, then you haven't really looked at Ubercharger builds. Several of the gimmicks used are Barbarian class features/ACFs, so a Rogue can't go as far along that path. Others are in fact feat based, but without Pounce the Rogue is going to be doing about a factor of 5 less damage. For an Ubercharger build, 500 damage per round at Lv6 (Barbarian 1, Fighter 5) is possible: Leap Attack (power attack is at -1 for +3), Power Attack, Shock Trooper, Spirited Charge (x2 on total damage), Battle Jump (x2 -> x3) Valorous Weapon (x3 -> x4), Headlong Rush (x4 -> x5, requires Orc), Pounce from Lion Totem Barbarian, Whirling Frenzy ACF (+1 attack). So here you're getting 3 attacks where you get +15 damage per -1 to AC (so take -6 for +90 damage per attack). In addition, add 1.5*Str mod*5 to this (as a raging Orc Barbarian, Str mod is probably about +6 to +8, so this is another +45 per attack minimum). So this is about 405 damage per round plus 15 * base weapon damage (using a 2-handed sword, thats 15*7 = 95 damage on average, for a total of 500).

Are you saying that a commoner could do the same 5 nukes I showed for the rogue? Meaning a T6 class able to fulfill JaronK's definition for T2 class? That would completely turn upside down the tier system as I understood it so far.

A specific build, using things that have nothing to do with class features, could do the same 5 things you showed for the rogue, yes. This is irrelevant to the Tier of the Commoner because, as I mentioned:

- Its a single specific build. The Tier system does not measure the most extreme outcomes, it just measures the average outcomes.
- Its primarily reliant on things which are not class features of the class in question.

Skill mastery alone is not a nuke. But it contributes to the performance of one, say diplomancy.
That 40 DC from above for avoiding all hostile encounters? Well, with taking 10 and adding 13 ranks, plus synergies from bluff, sense motive and local history, plus 11 other bonuses from items and mixed stuff, you can reliably get that nuke at level 10. A commoner cannot do that.

Sure they can. Lv10?

WBL at Lv10 is 49000gp. A Human Commoner with 28 point buy invests in a 16 Cha and Int dumpstats everything else. At Lv1 they take Nymph's Kiss (+2 to Cha based checks, +1 skill point per level) and Noble Born (Diplomacy is a class skill for you). At Lv3 they take the feat Item Familiar.

Using Item Familiar and their 7 skill ranks per level, they invest in UMD, Diplomacy, Bluff, Sleight of Hand, and three more skills of their choice. This gets them up to half of the normal max rank, but then Item Familiar allows you to get bonuses by investing skill ranks in the Familiar (this doesn't burn up the ranks, it just makes them dependent on having the item at hand). This means that you can effectively double your skill investment, and so you have the normal max ranks for 2-3 of the skills you want to max out (based on the fact that its 3-for-1 to add a rank via Item Familiar).

So effectively you can keep up with the Sleight of Hand, UMD, and Diplomacy of a normal class-skilled character this way. Oh, and because of Noble Born, you don't even have to do this trick for Diplomacy so that surges ahead to be double the value of a normal character.

So by Lv10, you can have 13 base ranks of Diplomacy, 13 ranks from Item Familiar, +2 to the check from Nymph's Kiss, +4 from Synergies, and then whatever you toss on using your WBL to buy magic (probably a +15 from Divine Insight is easiest). So that's a base check of +47. At that point, you don't need Skill Mastery to guarantee the DC 40 check which you claim is campaign-breaking.

Similar math can apply for Sleight of Hand, as there are feats which let you add Sleight of Hand to your class skill list (Cosmopolitan, for example, lets you get a +2 to any one skill and make it a class skill). Able Learner is also a reasonable purchase for this build since it makes everything cost only half as much skill-rank-wise.

By the logic you present, once I've done the UMD and Sleight of Hand parts of this build, I have infinite wealth from Sleight of Hand to buy any magic I want, so I just load myself down with scrolls and pretend to be a wizard.

[Ivanhoe]

Thanks for the posts, I think I learn quite a few things about the tier system.

So, aleucard says that the tier system is intended for levels 5-15 (thus covering only a bit more than half of the 1-20 level range). And, that skills do not count for the tiers. Then, of course, the rogue has not much going for him. In fact, omitting levels 1-4 from tiers means a likely strong phase of mundanes is excluded. If that is the case, I understand the tier system much better now.

And eggynack says that diplomacy (nor the other four nukes I mentioned) are campaign-changers. But spells like silent image, polymorph, control winds and shapechange are - when used as spells cast, not as obtained by items and/or UMD.
I think that's not really true, but I'd be interested in opinions from other posters as well on these remarks.

@NichG: I think there are some mistakes in what you write.

1) ubercharger vs diplomancer at level 10
- unintelligent and non-common-speaking enemies are not really that frequent, and the rogue can use another of his nukes (sneak attack) if diplomacy can't. Does the ubercharger have another nuke?
- multiple enemies similarly can be addressed with sneak attack, while the ubercharger just charges one, then is in danger of many counterattacks with his low AC (shocktrooper)

2) Factotum, cleric, bard better at diplomacy than rogue at level 10
Probably not, I'd say.
The cleric has only 2 skill pts/lvl (and usually needs religion, spellcraft, concentration as well). The big buffs work only on one check iirc.
The bard can probably be equal or slightly better with his buffs to the rogue with skill mastery. I'd have to see the math with buffs like improvisation spell (again, only applicable to some, not all diplomacy checks). Probably the difference is not significant.
The factotum again only gets a boost to one check per day with cunning knowledge.

3) Bluff lasts max only one round
Please read again the skill description that you quoted. It says "A successful Bluff check indicates that the target reacts as you wish, at least for a short time (usually 1 round or less) or believes something that you want it to believe."
The example I used (lie to the BBEG about the loyalty of his 2nd BBEG) refers to the part after "or" (and it is without any duration given); also the first entry refers likely to the combat/diversion uses of bluff.

4) A rope and lock making immune vs Sleight of hand
I do not think so. First of all, there are no rules that this blocks thieving altogether (for instance, will this mean that the item gets bigger than small? Even pieces of clothing excepting a robe are not of medium size or bigger). Then, it would look extremely strange, and would necessitate that all npcs do several move actions to use what they have (weapons, spell components, magic items, whatever). Including all people in cities of low CR that the rogue could all rob fairly easily in my base campaign-changing scenario. Then, purses were usually fixed to belts and the basic thievery therefore assumes that you get it, anyhow, regardless of any ropes attached. In essence: I do not think that works.

5) 300-400 damage at level 20 being reduced by not hitting, miss chances etc.
Quite probably a level 20 rogue can be built to one-shot most CR 20 monsters with sneak attack with touch attacks. Do you really believe that is not possible? Damage-dealing has never been a problem for rogues.

6) Ubercharger does the same damage of a level 20 rogue already at level 6
Here I think your barbarian ubercharger build will not work.
- you have used way too many feats (orc barbarian has only 3 at level 6, plus maybe 2 for flaws). And multiclassing in the tier system is not measured - or is it?
- shocktropper needs +6 BAB and headlong rush +4 BAB, so you can only take one at level 6 (likely shocktrooper, or you won't hit anything). Down one multiplier.
- Spirited charge only works when mounted (and needs too many feat reqs for your build). You cannot be mounted and leap attack at the same time. Down another multiplier.
- Valorous weapon is minimum +2 (8,300+), too costly for a level 6 character with the DMG restriction of 50% max wealth going to one item. Down another multiplier.
(- battle jump is a regional, setting-specific feat. I doubt that orcs can take it, but I am not sure, so I guess it's OK)
- so you still have the feats power attack, shock trooper (plus improved bull rush as rereq), leap attack, battle jump, meaning you need 2 flaws.
So this yields in my view an elite array orc barbarian with 24 raging strength at level 6, damage 2d6 + 10 (strength and TH-Weapon) +18 (power attack), probably +1 enhancement from a magic weapon for a total of 36 damage average; multiplied x2 for 72 damage per hit. You attack three times with pounce at +6/+4(whirling frenzy)+1 BAB, +7 Strength, +1 enhancement, +2 charge for a total of +16/+14/+11. Likely against AC of around 20 at these levels, so probably per round around 140 damage. Not bad. But not as high as you thought.

7) Commoner vs rogue.
Here I do not quite get you. You say, specific builds should not be used for tier categories, and then you go build a commoner with wbl and feats all classes can take and say that it shows it can do it as well as the rogue? I mean, you use an item familiar to equate the rogue's class ability advantage of skill mastery. A rogue can get an item familiar as well, putting him forever out of reach of the commoner. Moreover, a rogue can play with just the core rules in all campaigns with that campaing-changer nuke, whereas the commoner needs such an item to even compete. Confusion.

[NichG]

@NichG: I think there are some mistakes in what you write.

1) ubercharger vs diplomancer at level 10
- unintelligent and non-common-speaking enemies are not really that frequent, and the rogue can use another of his nukes (sneak attack) if diplomacy can't. Does the ubercharger have another nuke?

I'm not arguing that Uberchargers are Tier 2, so they don't need to have a second nuke. Anyhow, sneak attack does not qualify as a nuke. Its pretty underwhelming really, maybe even a bit of a trap option.

Unintelligent enemies can be quite common; the same is true of non-common-speaking enemies. Non-sentient undead, for example. Also constructs, many abberations, most plants, animals, oozes, vermin, elementals, and some magical beasts. So thats about half of the Types in the game (7.5 out of 15 if you count undead, abberations, and magical beasts as 50% of a type).

2) Factotum, cleric, bard better at diplomacy than rogue at level 10
Probably not, I'd say.
The cleric has only 2 skill pts/lvl (and usually needs religion, spellcraft, concentration as well). The big buffs work only on one check iirc.

I'm not sure why you're so focused on skill points per level, since there are so many ways to get more that its pretty much irrelevant. And you only really need one check if Diplomacy is as much of a nuke as you say it is. You only need 1 skill point per level to max a skill, anyhow. The Cleric can max Diplomacy just as well as anyone else if they really want to.

3) Bluff lasts max only one round
Please read again the skill description that you quoted. It says "A successful Bluff check indicates that the target reacts as you wish, at least for a short time (usually 1 round or less) or believes something that you want it to believe."
The example I used (lie to the BBEG about the loyalty of his 2nd BBEG) refers to the part after "or" (and it is without any duration given); also the first entry refers likely to the combat/diversion uses of bluff.

It says 'usually 1 round or less'. If it works for more than 1 round, that's because the DM is specifically allowing it to do so, not because the rules demand it do so. So, assuming a completely neutral DM (who doesn't do any particular favors above and beyond the normal, but who also doesn't go out of his way to stop things from working), 1 round is all you can rely on.

4) A rope and lock making immune vs Sleight of hand
I do not think so. First of all, there are no rules that this blocks thieving altogether (for instance, will this mean that the item gets bigger than small? Even pieces of clothing excepting a robe are not of medium size or bigger). Then, it would look extremely strange, and would necessitate that all npcs do several move actions to use what they have (weapons, spell components, magic items, whatever). Including all people in cities of low CR that the rogue could all rob fairly easily in my base campaign-changing scenario. Then, purses were usually fixed to belts and the basic thievery therefore assumes that you get it, anyhow, regardless of any ropes attached. In essence: I do not think that works.

A 2-handed weapon is explicitly the same size category as its wielder:

In general, a light weapon is an object two size categories smaller than the wielder, a one-handed weapon is an object one size category smaller than the wielder, and a two-handed weapon is an object of the same size category as the wielder.

Full-body clothing should also fit the same sort of scaling (e.g. if it can contain the wielder, it should be the same size category as the wielder). So e.g. wearing your magic rings under your locked gauntlet which is attached to your full plate should protect them. Having your amulet use a chain that needs to be unlocked to get it over your head should end up doing the same.

5) 300-400 damage at level 20 being reduced by not hitting, miss chances etc.
Quite probably a level 20 rogue can be built to one-shot most CR 20 monsters with sneak attack with touch attacks. Do you really believe that is not possible? Damage-dealing has never been a problem for rogues.

Generally speaking, I think the rogue will have a much harder time pulling that off than one of the many other options available. Sneak Attack isn't actually very effective. Also, in my calculation I erred and gave the Rogue 20d6 sneak attack, but really they only have 10d6 sneak attack. So even their max damage won't kill a CR appropriate monster in one shot.

6) Ubercharger does the same damage of a level 20 rogue already at level 6
Here I think your barbarian ubercharger build will not work.
- you have used way too many feats (orc barbarian has only 3 at level 6, plus maybe 2 for flaws). And multiclassing in the tier system is not measured - or is it?

Specific builds are not measured by the tier system, so the Ubercharger is not given as an argument for the tier of a character, its given as an argument that what you're calling a 'nuke' doesn't qualify. Rogues are not, on an absolute scale, all that good at damage.

As for the feats, you get 3 from normal progression, 3 more from Fighter bonus feats, and 2 more from flaws. I think that the big list can be done with 8 feats, but lets see.

- shocktropper needs +6 BAB and headlong rush +4 BAB, so you can only take one at level 6 (likely shocktrooper, or you won't hit anything). Down one multiplier.

You get a feat at Lv6, so make it Shocktrooper. You get a Fighter Bonus Feat at Lv5 of this build, so use that for Headlong Rush.

- Spirited charge only works when mounted (and needs too many feat reqs for your build). You cannot be mounted and leap attack at the same time. Down another multiplier.

Being mounted isn't a problem, and you can certainly jump while mounted. You've got the WBL with this character to go buy a hippogriff if you like. Nothing stopping you from leaping from its back and landing back on its back when delivering the strike. There are other, cheesier ways to do this as well (e.g. co-locating with an incorporeal mount).

As far as feats, Spirited Charge costs you three, so lets tick them off. We've used 5 of 8.

- Valorous weapon is minimum +2 (8,300+), too costly for a level 6 character with the DMG restriction of 50% max wealth going to one item. Down another multiplier.

That DMG restriction is a suggestion rather than a rule, and also wouldn't apply to a character in play who amassed actual wealth and then used it to buy things. The character can afford this weapon and the hippogriff and still has a few thousand gp left over.

(- battle jump is a regional, setting-specific feat. I doubt that orcs can take it, but I am not sure, so I guess it's OK)
- so you still have the feats power attack, shock trooper (plus improved bull rush as rereq), leap attack, battle jump, meaning you need 2 flaws.

So this yields in my view an elite array orc barbarian with 24 raging strength at level 6, damage 2d6 + 10 (strength and TH-Weapon) +18 (power attack), probably +1 enhancement from a magic weapon for a total of 36 damage average; multiplied x2 for 72 damage per hit. You attack three times with pounce at +6/+4(whirling frenzy)+1 BAB, +7 Strength, +1 enhancement, +2 charge for a total of +16/+14/+11. Likely against AC of around 20 at these levels, so probably per round around 140 damage. Not bad. But not as high as you thought.

Battle Jump, Power Attack, Leap Attack are the remaining 3 feats. So it looks like we're one feat short (we'll get it at Lv7). So I lose one of the multipliers, which means that the damage per round is about 400 instead of 500.

7) Commoner vs rogue.
Here I do not quite get you. You say, specific builds should not be used for tier categories, and then you go build a commoner with wbl and feats all classes can take and say that it shows it can do it as well as the rogue? I mean, you use an item familiar to equate the rogue's class ability advantage of skill mastery. A rogue can get an item familiar as well, putting him forever out of reach of the commoner. Moreover, a rogue can play with just the core rules in all campaigns with that campaing-changer nuke, whereas the commoner needs such an item to even compete. Confusion.

I'm showing you why there's the rule that specific builds don't matter for Tier comparisons. If they did, the Tier system would be hopelessly useless because you can always make a specific build that relies primarily on non-class-features that can basically do anything. The game is pretty malleable in that way.

The point of the Tier system is to look at how a given class is going to play out 'on average', so that you can do a quick estimate as to whether a given party composition is going to give everyone equal spotlight time. Its probably most useful when dealing with players who play around medium op level and generally constrain themselves voluntarily to focusing on class archetypes - e.g. someone makes intelligent picks within a class, but doesn't do encyclopedic-knowledge cross-setting shenanigans or other such things. For characters whose abilities are primarily dominated by choice of class, the Tier system gives a way of guessing how things are going to play out. For characters whose abilities are dominated by feats, items, or peculiar/particular optimization choices (diplomancer commoner), 'Tier' of the component classes is less of a relevant factor.

The other major way in which Tier is a useful concept is in game design (e.g. of homebrew, or campaign rules, or things like that). Understanding that versatility is what ultimately breaks the game is important. Understanding T4/T3/T2 is pretty useful for designing classes to be fun to play without dominating the game (e.g. by looking at T3 examples, you can see how to build a class which is highly competent in focus, is not incompetent outside-of-focus, but which also doesn't overshadow others). From that point of view, the ability to do some particular build using things available to all characters (max ranks in a skill, feat chains, magic items) isn't relevant to the game design problem of designing a class to the desired level of spotlight control.

[Ivanhoe]

I have thought a bit, and come to realise that a T2 rogue is not possible with the assumptions posted here. So, another contender for mundane T2+ fails

Some answers for NichG;

I'm not arguing that Uberchargers are Tier 2, so they don't need to have a second nuke. Anyhow, sneak attack does not qualify as a nuke. Its pretty underwhelming really, maybe even a bit of a trap option.

Unintelligent enemies can be quite common; the same is true of non-common-speaking enemies. Non-sentient undead, for example. Also constructs, many abberations, most plants, animals, oozes, vermin, elementals, and some magical beasts. So thats about half of the Types in the game (7.5 out of 15 if you count undead, abberations, and magical beasts as 50% of a type).

Looked up the core CR 10 creatures and without humanoids npcs you are correct, I think it is roundabout only half of the creatures that would understand common. A diplomancer would have to at least learn draconic and abyssal, I think. Problably not that big effort, in particular for a skillmonkey, but still.

I'm not sure why you're so focused on skill points per level, since there are so many ways to get more that its pretty much irrelevant. And you only really need one check if Diplomacy is as much of a nuke as you say it is. You only need 1 skill point per level to max a skill, anyhow. The Cleric can max Diplomacy just as well as anyone else if they really want to.

Yes, if clerics really want to, but the opportunity cost is so much higher since they need their low skill points for other skills. But if skills do not count, anyhow, for tier discussions, it does not really matter.

It says 'usually 1 round or less'. If it works for more than 1 round, that's because the DM is specifically allowing it to do so, not because the rules demand it do so. So, assuming a completely neutral DM (who doesn't do any particular favors above and beyond the normal, but who also doesn't go out of his way to stop things from working), 1 round is all you can rely on.

Here, again, check the second part of the duration sentence (after "or"). The lie part has no set duration. That was what I was referring to (I mean, why should someone believe a lie for only max 6 seconds? Would be odd).

Full-body clothing should also fit the same sort of scaling (e.g. if it can contain the wielder, it should be the same size category as the wielder). So e.g. wearing your magic rings under your locked gauntlet which is attached to your full plate should protect them. Having your amulet use a chain that needs to be unlocked to get it over your head should end up doing the same.

Do you think that all npcs in the world will do that? I meant the wealth-amassing effect of a free action stealing rogue would not be targeted on the major npcs of the city, but the mid-low level npcs.

Generally speaking, I think the rogue will have a much harder time pulling that off than one of the many other options available. Sneak Attack isn't actually very effective. Also, in my calculation I erred and gave the Rogue 20d6 sneak attack, but really they only have 10d6 sneak attack. So even their max damage won't kill a CR appropriate monster in one shot.

A reasonably well maxed rogue could get 20d6 sneak at level 20, I think (with elements like assassin's stance and items). And again, the campaign-changing aspect I thought about was a level 10 rogue with touch sneak attacks killing way higher CRs. But I may have exaggerated here.

Specific builds are not measured by the tier system, so the Ubercharger is not given as an argument for the tier of a character, its given as an argument that what you're calling a 'nuke' doesn't qualify. Rogues are not, on an absolute scale, all that good at damage.

Rogues are, in my view, on an absolute scale, quite excellent at damage, throughout most levels (possibly a bit weaker at low levels). And they are in the same league as uberchargers, but difficult to be compared, in particular since sneak attack and ubercharging need different conditions to work. But our different opinions seem to also reflect different gaming experiences.
(and, btw, I still do not think a jump multiplyer while riding and doing spirited charge is possible. At least not with a running jump modifier and jumping off the mount.)

[NichG]

Here, again, check the second part of the duration sentence (after "or"). The lie part has no set duration. That was what I was referring to (I mean, why should someone believe a lie for only max 6 seconds? Would be odd).

Its kind of ill-defined then, if they 'believe the bluff' but 'do not react as you want'. That falls in the grey area that can easily be defended against by layers of bureaucracy - namely, you lie to someone (e.g. a guard or lower-tier official) and they make a report to their superior saying 'so and so said X'. The superior then is never directly exposed to your Bluff, and so can react to the information in a reasoned way. You may want the lower-tier official to make a snap decision because of the importance of the information, but at that point you stray into the 'react as you desire' territory which has a 1 round limit.

Bluff certainly has its uses, but its not really well suited to be a 'nuke' per se. To use it effectively, you have to arrange to be in a situation in which the target is relying on you for information (e.g. you have to get them to 'pull' information from you), as opposed to deciding that 'I want to push information on the target and force him to believe me'. The difference between that kind of 'push' mechanism and the 'pull' mechanism is that you can't just make whatever you want happen, you have to do a combination of larger-scale manipulation and waiting for the opportune moment.

Which differentiates it from what people are describing as campaign-shattering nukes, in that with a nuke the player can, at any time and at their whim, cause the campaign to implode. Compare 'I start casting Wall of Salt and sell the salt' with 'when the king asks me to investigate some murders in the kingdom, I'll use Bluff to pin it on the vizier as a power play to rise in the ranks of the king's service'. The second is certainly an awesome thing to pull off, but it requires a lot more set-up. So its just a nice ability rather than a 'nuke' - its actually playing along with the game rather than acting to disrupt the game.

What tends to separate T2 and T3 is whether or not the application of the class' powers make it impossible for some kind of extant storyline to persist. For T3 characters, they're still forced to deal with the context of the scenario. For T2 and up, in principle those characters can uproot the context and just force it to be different if they don't like how things are. T2 characters can do that for a subset of contexts (now that I'm here, we will auto-win all combats, but I can't really help in social encounters) whereas T1 characters can do that for all contexts (now that I'm here, we auto-win all combats, economics no longer applies to us, and I can go and Dominate the entire government during a downtime if we're having some political problems).

Do you think that all npcs in the world will do that? I meant the wealth-amassing effect of a free action stealing rogue would not be targeted on the major npcs of the city, but the mid-low level npcs.

Well, those NPCs who are carrying around a king's ransom in magical items, yeah. Considering a commoner can get enough Sleight of Hand to steal something, it seems like it'd be the natural thing to do if you're going to carry more than 50gp worth of stuff with you.

If you're targeting people who are not PC-level rich, you can make money at it but it will take a long time, especially if you end up with a lot of marks who are just carrying copper. I mean, lets say you steal from someone each round and your carrying capacity is 200lbs (which is 10000 coins). If you're stealing copper pieces, then every 100gp worth of copper you steal you have to go home and store it. Lets say every round for 8 hours you can steal 100 copper (1gp) from peasants, etc who are wandering around. So every 100 rounds (10 minutes) you get 100gp - not bad, but then you spend another 5-15 minutes lugging 200lbs of metal home with you and back. So effectively you get 100gp every half-hour, or 1600gp a day. Not bad, but not Wall of Salt or Craft shenanigans levels of income. Its awesome at Lv3, pretty sweet at Lv6, and a nice bonus up to Lv10 or so, but not really worth your time after that.

However, lets add to this that in a city with 2000 people in it, that means you go through the entire population of the city each day. So you are quickly going to re-encounter people you've stolen from before at that kind of rate. Not to mention that the guard will probably notice the currency black-hole hanging out at a particular point in town even if no one can see through your Sleight of Hand checks directly. So you probably can't really keep it up to even that kind of level.

A reasonably well maxed rogue could get 20d6 sneak at level 20, I think (with elements like assassin's stance and items). And again, the campaign-changing aspect I thought about was a level 10 rogue with touch sneak attacks killing way higher CRs. But I may have exaggerated here.

True. Craven is the big one, at +1 damage per level (so at Lv20 this is equivalent to an extra ~6d6), assassin stance is 2d6. So 18d6. (Of course, the Ubercharger can use Assassin Stance as well, though its probably stretching the rules to argue that Assassin Stance lets you qualify for Craven)

Rogues are, in my view, on an absolute scale, quite excellent at damage, throughout most levels (possibly a bit weaker at low levels). And they are in the same league as uberchargers, but difficult to be compared, in particular since sneak attack and ubercharging need different conditions to work. But our different opinions seem to also reflect different gaming experiences.

Mostly I'm arguing that 'doing a reasonable amount of damage' isn't a nuke. A nuke breaks things and makes part of the game irrelevant. Having a rogue in the party does not make enemy hitpoint totals irrelevant, but having a Ubercharger does (because what will determine if the enemy lives or dies will be whether the charge can be set up, not whether the damage is enough to drop them - if the Ubercharger manages to make contact, the damage will always drop anything in the books beyond Lv6 or Lv7).

(and, btw, I still do not think a jump multiplyer while riding and doing spirited charge is possible. At least not with a running jump modifier and jumping off the mount.)

Well, not getting the running modifier isn't such a big deal, because the Jump DC is just 10 (so DC 20 without the running modifier). Not too hard to hit reliably by Lv6 (max ranks is 9, barbarian fast movement gets you another +4, and +6 from Strength means you have a +19 modifier and never miss the check even on a 1). At Lv7 you also get a +2 synergy from (cross-class but totally worth it) Tumble, in case there are situational penalties from doing mid-air jumpy hijinks. Its also a good candidate for buffing via magical items or 'a masterwork item of jumping' e.g. springy boots or something like that.

[Ivanhoe]

Wow, what a fast reply!

One more thing: you say that casting a wall of salt is more game-breaking than getting in a lie to put the royal vizier into jail. This would be because there is a lie-detecting bureaucracy there. Similarly, people being robbed in a city will quickly notice it and focus on the rogue as the culprit, blocking this kind of income before reaching any campaign-disruption.
But you'll notice that you had to assume DM reactions to what the rogue did with bluff and sleight of hand.
I'd say there could be similar reactions of the DM to wealth-creating spells like the wall of salt, which appear to me also as "playing along with the game".
Here is what I'd do as a DM:
PC wizard casts wall of salt to set up a salt business. OK - who is going to buy the salt? Is the wizard going to hire people for his business and will they be honest? Is there going to be competition? What will market prices of salt do when the wizard dumps tons of it in the local 2,000 people town (that you used as an example to "counter" the sleight-of-hand stealing from getting out of hand).

The rogue abilities are basically assumed to be countered more easily/less fiat-like by the DM than powerful spells, thus meaning mundane T2 are never possible. What would you say? Does this reflect your game experience?
Or are basically both spell and skill nukes stopped by any sane DM quickly before they can disrupt a campaign?

[Segev]

This gets into economics, and likely is thus a topic for another thread, but I feel it necessary to point out that the actual effect of the wizard selling his salt can be seen in the real modern world. Time was, salt was literally worth its weight in gold. As salt has become easier to harvest and distribute, it has become a cheap household commodity. The wizard's Wall of Salt-based salt-selling business would have a similar effect. You can discuss guild protectionism and dishonesty and the like, but the truth is that the reason the wizard will be more successful with less ire than the rogue is because the wizard is actually creating wealth (much like most successful, truly private businesses which have to convince people to willingly part with their money in return for goods or services). The rogue is stealing, and thus at best resulting in a zero-sum game wherein he reduces others' wealth to increase his own. (In practice, he often destroys some amount of wealth by having to use up resources that could be put towards more productive ends in order to enact his fraud, burglary, or sleight of hand. Even if that's just necessitating more security. Sure, that's jobs...but those security guards COULD be doing labor instead of preventing theft if there weren't thieves around. And labor produces wealth, while all they're doing is trying to prevent its theft.)

That's why wizards are instinctively viewed as doing more with less risk when they create salt ex nihilo versus the rogue who is stealing a similar amount of wealth.

As an example of this taken to the extreme, a Tippyverse of rogues stealing on the industrial scale of Tippy's traps-of-create-whatever-is-needed would not produce the post-scarcity society that the Tippyverse's actual traps-of-create-whatever-is-needed on an industrial scale do. Just because the relative value of the created goods reduces compared to money doesn't mean the overall wealth-value hasn't gone up by virtue of having more of it. The result of such things is reduction of scarcity and advancement further and further beyond subsistence.

[NichG]

Wow, what a fast reply!

One more thing: you say that casting a wall of salt is more game-breaking than getting in a lie to put the royal vizier into jail. This would be because there is a lie-detecting bureaucracy there. Similarly, people being robbed in a city will quickly notice it and focus on the rogue as the culprit, blocking this kind of income before reaching any campaign-disruption.
But you'll notice that you had to assume DM reactions to what the rogue did with bluff and sleight of hand.
I'd say there could be similar reactions of the DM to wealth-creating spells like the wall of salt, which appear to me also as "playing along with the game".
Here is what I'd do as a DM:
PC wizard casts wall of salt to set up a salt business. OK - who is going to buy the salt? Is the wizard going to hire people for his business and will they be honest? Is there going to be competition? What will market prices of salt do when the wizard dumps tons of it in the local 2,000 people town (that you used as an example to "counter" the sleight-of-hand stealing from getting out of hand).

Generally the assumption when doing analyzes of this sort is a DM who is neither permissive nor restrictive. That is to say, the DM in such an analysis will never give you anything more than the rules require him to give you, nor will he bend the rules to stop you from doing something even if it seems ridiculous (that said, NPCs can also legitimately do anything the PCs can do in these analyses, so counters that can be implemented by the NPCs using the rules as written are also fair game for consideration - this is part of the line of thinking that gets you to the Tippyverse). This analysis, of course, looks nothing like any real game when you get to the fringe things like Wall of Salt or Sleight of Hand economics, Diplomancy, and the like. But its generally something that is taken as a bit of a shared point to prevent discursions of 'well in my campaign...'

The rogue abilities are basically assumed to be countered more easily/less fiat-like by the DM than powerful spells, thus meaning mundane T2 are never possible. What would you say? Does this reflect your game experience?
Or are basically both spell and skill nukes stopped by any sane DM quickly before they can disrupt a campaign?

This becomes more complex, and personally I think it gets into a more interesting area of discussion even if it means we're not on a firm ground of shared assumptions.

In my experience, what happens is that the outright broken interpretations of Diplomacy, Sleight of Hand, and Bluff are simply refused without making any particular in-game structures to adapt for them. House rules to prevent these broken things are easy and are usually the first thing that happens - the attitude chart is thrown out and Diplomacy is used in a go-fish manner where the DC is ad-hoc'd for the situation and how you're trying to use it; Bluff is limited to 'they believe that you believe what you're saying' and so you cannot go up to a guard and say things like 'I'm actually the king', period; Sleight of Hand fails to get the item if the victim makes their Spot check; Wall of Salt suffers quick local economic equalization, limiting the money you get (or, sometimes, some resource more valuable than gold is made the heart of the 'real' economy of things PCs would be interested in - xp for example - and that yellow stuff you get from merchants is kind of worthless in the arena of cosmic conflicts).

So yes, I can't say I've been in or run a campaign in the last 5 years where any of the base D&D stuff produced real 'nukes' of any sort. Generally I and the DMs I play with are experienced enough to recognize where most of the problem spots are going to be as far as nukes, and can take actions to soften those things out either by adapting the rules or adapting the world or adapting the underlying assumptions or even by adding new stuff in.

That said, in general I have still found that in anything like the base game, a Wizard is going to 'dominate the scene' more often than a Rogue or a Bard will, and the Rogue will, on the basis of his own powers, often have nothing to do that gets even close to what the Wizard, Archivist, or even Factotum can just pull out of nowhere. It hasn't been uncommon for me to play in or run games where, e.g. skill checks in the 60s are getting thrown around at Lv10, and 90% of that comes from direct application of magic (e.g. someone casts the skill buff spells and goes to town). Someone playing a Rogue in a game with e.g. the Factotum character that one of my players had in my last campaign would basically have nothing to do. The Factotum was using a bunch of tricks that relied on things like Arcane Dilettante and boosting skills via Inspiration and the like to be able to hit Sleight of Hand checks in the 80-90 range - which in that campaign was used to do things like 'steal' another PC from underneath a pile of rubble to save them. At the same time, he could hit a 50 in any skill check he wanted, and could also pull off some other tricks besides.

So I think if you blunt the tips of all the nukes, you still end up with situations in which some characters simply have more things they can do effectively than others. Its not Tippyverse-level rewrite the campaign setting stuff - T2 arguably disappears as a distinct Tier because of this - but you still end up with something like:

T1 - Can do most things that need doing on their own, and can do them very well
T3 - Can do one thing that needs doing very well (e.g. they are among the best at this), and can suffice with most other things but will be outclassed by a specialist
T4 - Can do one thing that needs doing very well (e.g. they are among the best at this), but can't really do other things. Alternately they suffice at many things but are a master of none.
T5 - Can suffice at one thing that needs doing, but will be outclassed by a specialist
T6 - Cannot suffice at any task

(All of these of course are 'by virtue of their class abilities'). Somewhere in there should also be a consideration of what the permanent resource expenditure is - e.g. whether a class has to specialize permanently or if they can specialize on the fly.

So a rogue would arguably end up at T4 or T3 at best in this hierarchy. They can suffice at several things, and even do a couple of them pretty well (especially if they specialize). But its hard to argue that they're among the best at all of them. They can do a decent job as a skill monkey, but they lack the magical oomph to really compete with a Bard (who gets nearly as many skill points, but magic to buff skills too) or Factotum. They can do a pretty good job for damage, but its fragile - some creature types are just outright immune to their tricks, so they often end up outclassed by Warblades, Swordsages, or even Barbarians at an equal level of optimization.

[georgie_leech]

-snip-

Tier 2 can still exist in this construction. While Tier 1's can do most things on their own and do them very well, Tier 2's can theoretically do the same but are limited in practise to a limited subset. In other words, a Wizard can change his spells to have the perfect answer to any problem; there is a Sorcerer that has the right set up for any given problem, but not every Sorcerer will have that solution.