Designing a Tier test

[maysarahs]

For a while I have toyed around with the idea of designing a series of encounters with the intent to demonstrate the overal utility of a character, but have no idea where to start. All I can really offer is a brainstorm, but hopefully this can be a resource we refer to to demonstrate the relative power of builds.

So, for starters, I figure it could take place as a series of tests and the idea is that builds would run through a standard 4 different encounters per day randomized (so a wizard can't breeze through it with divination tricks without sacrificing a few spell slots for that day) for a few days until all of the possible scenarios had been played through. The breadth of encounters would span what I assume to be anything an adventurer could possibly be required to do.

My brainstorm
Combat: this is is commonly one of the few places people look at for determining a class's power, but it is also one of the hardest things to test. Varied terrain (underwater, in driving snow, standard arena, difficult terrain) is a factor, as is awareness, and distance. I figure a bunch of different encounters is needed to give a fair assessment here.
Combat versus one main enemy
Combat versus mooks/minions

Travel: I figure long distance travel, (a race?) as well as say short term goals such as traversing a gorge, or getting to the other side of a well guarded room/building

Social encounters: say convincing a king for funds, or an innkeeper who
just saw you with his daughter/son to let you stay at the inn for the night, stopping a riot/stampede, finding the culprit in a crime

Stealth, traps, various monsters, also present obstacles

One way to test a few of these at once could be an obstacle course.

I have no idea what one can do to apply a grading system, but I guess the number of succeeded encounters and also the simplicity with which a particular obstacle is traversed is enough as a demonstration (Eg. for a long distance race instead of a series of survival checks, and random encounters and stuff for the fighter to ride a horse for 100 miles (completion: several days), the wizard consults the local libraries map to make sure hes aiming for the right area, sleeps, prepares teleport, and beats the fighter to the finish line just in time for prestidigitation to finish heating his cup of tea (completion: one day, two rounds)) to describe a classes power. I also figure a group dynamic is sometimes important, (in combat a DFI bard would show off his power best with a few beatsticks to help out) so a few NPC's should be made available for hire to every participant in the test.

For closing comments, I think that this could not only be used for demonstration purposes, but also as a gauntlet to run any future builds through to see if it succeeds in what it wants to do (someone who wants to build a character with the ability for tremendous burst damage, but who may not be a caster, could pit himself against the "one boss arena" challenge and make sure he at least can pull through in his preferred area of expertise.)

I understand it is a big and very open ended project to think up but if the playground is up for it I'd be interested in what suggestions come to mind!

P.S. ideally it would be system independent, but realistically if things are designed in PF, 3.5, or some combination it would be simpler

[Rhaegar14]

I'd love to help you with this, especially because it'd probably help solidify my understanding of why each tier is which (I get it in theory, but in practice my highest-level game is nine).

One thing you could do for a bunch of them is evaluate them based on how many rounds it takes; even social encounters (one spell vs. ten rounds of diplomancy).

[jiriku]

It doesn't take a simulation to identify a class tier. Simple comparison to existing classes, or a few thought exercises involving the situations you suggested, should do it. If you've got something hombrewed and you're not sure, just post it in the homebrew forum and ask - a pretty reliable consensus opinion should quickly emerge.

[TroubleBrewing]

Make sure that you can't pass the tests with just WBL-mancy.

[Coidzor]

Nice idea, scenario or two for every sphere of the game, come up with a scoring system for their ability to interact with the scenario within the rules, and then based upon their score in all the spheres they get their tier ranking assigned...

Worst case it helps focus the thought process just having a couple of scenarios and determining what the spheres of the game are, even if you don't actually bother to ascribe points to various levels of dealing with them.

[erikun]

I would recommend:

An encounter with a single, strong enemy, like a giant or a titan.
An encounter with a single, highly mobile enemy, like a demon.
An encounter with a single magical enemy, like a dragon or Lammasu.
An encounter with a large number of weaker enemies, like an army of orcs.

Needing to get into or out of a heavily guarded place, with otherwise overwhelming CR defenders.
Needing to convince a heavily warded NPC of a course of action.
Needing to acquire a trinket from a locked and warded vault.
Needing to rally a large group of people to a task.
Needing to surprise and subdue a weary and flighty target.
Needing to travel a large amount of distance in a small amount of time.

Producing goods for a large number of people in the most efficient way.
Producing useful weapons and equipment with little or no resources.
Escaping from a cell or prison with your equipment in another room.


That seems to cover a wide variety of situations, some of which would be well-suited towards characters with particular skills. PvP isn't necessary, or even desired, as a lot of situations you'll see have little to do with who can beat a monster or travel a journey first.

[Psyren]

The "caster bases" are also a good way to help determine tier. In combat, a well-rounded character should have the following offensive options:

- Ranged attack roll (touch if possible)
- Melee attack roll (touch if possible)
- Targets Fortitude Save
- Targets Reflex Save
- Targets Will Save
- Allows no Save
- Not Subject to Spell Resistance or Power Resistance (SR/PR)
- Area of Effect (for fighting Swarms)
- Force Effect (for fighting Incorporeal/Ethereal creatures)
- Effective against AMF/NPF
- Effective against Globe of Invulnerability

Naturally, access to spells/powers are the easiest way to cover all of these bases, particularly prepared from a large list. Note also that some abilities can cover multiple bases at once.

Low-tier melee has a very hard time covering even half of these. Low-tier casters (like Warmage) are similarly disadvantaged. Meanwhile, high-tier casters like Sorcerer and Psion can easily cover them all with leftover spells known, while high-tier prepared casters don't even have to be able to cover them all at once on a given day.

[Anarchy_Kanya]

http://brilliantgameologists.com/boards/index.php?topic=5070.0

[Eldan]

Tier 1 test is easy.
"Can you quickly think of five things this class can not do by level 20? If yes, it is not Tier 1."

[Rhaegar14]

Make sure that you can't pass the tests with just WBL-mancy.

But WBL-mancy is what makes certain options (the most outstanding being UMD) so valuable.

[Myth]

If you design the challenges I'd be happy to play a Tier 1. I'd even help you roll it or roll it myself, avoiding theoretical optimization levels and keeping things RAI.

[dobu]

I personally would add a series of encounters without resting as a test scenario. It's easy for a Duskblade or a Psion to burn through all his spells or powerpoints and kill a Monster, but he'd be depleted and therefor worthless for the rest of the day.

[Myth]

For a no resting scenario a Crusader would mop the floor with basically anyone who didn't resort to infinite loops. That guy can go on forever so long as the challenge is doable.

edit: that, or a Warlock.

[vitkiraven]

What about some explicitly non-magic encounters, where there would be a penalty or inability to even use magic, i.e. Convincing a King to your viewpoint in an antimagic area, or dealing with an honor duel with a barbarian prince on a dead magic holmgang island?
Convincing magic fearing/hating peasants to hide you for some reason, or convince the people of Thay that they need to rise up against The Red Wizards, without resorting to magic, and thuz losing their trust?

[Psyren]

I personally would add a series of encounters without resting as a test scenario. It's easy for a Duskblade or a Psion to burn through all his spells or powerpoints and kill a Monster, but he'd be depleted and therefor worthless for the rest of the day.

By RAW, it's just as easy for a Psion to get unlimited PP, if we're assuming TO rather than a DM.


For a no resting scenario a Crusader would mop the floor with basically anyone who didn't resort to infinite loops. That guy can go on forever so long as the challenge is doable.

edit: that, or a Warlock.

Binders, Incarnates and DFA are also perpetual-motion machines. Truenamer is theoretically one as well.

[Fax Celestis]

Make sure that you can't pass the tests with just WBL-mancy.

Limiting to non-custom items should handle that, I think.

[Hiro Protagonest]

Make sure that you can't pass the tests with just WBL-mancy.

Artificers.

[Tyndmyr]

I am skeptical of any test to accurately depict tier.

However, when you finalize it, I volunteer to run through them both with a strictly RAW legal wizard and truenamer if you need calibration. Both will have notable amounts of optimization without resorting to wish loops or pun pun stuff.

[Myth]

What about some explicitly non-magic encounters, where there would be a penalty or inability to even use magic, i.e. Convincing a King to your viewpoint in an antimagic area, or dealing with an honor duel with a barbarian prince on a dead magic holmgang island?
Convincing magic fearing/hating peasants to hide you for some reason, or convince the people of Thay that they need to rise up against The Red Wizards, without resorting to magic, and thuz losing their trust?

1. Initiate of Mystra or Dweomerkeeper
2. See above
3. Silent Still Charm Person / Dominate / Mindrape etc.
4. See above

Not that kings and peasants matter in a setting with Epic level casters (unless the kings ARE Epic casters in which case the point is mot)

[A Ladder]

Do alignments factor into Tiers?

A Myth said still/silent mindrape would work, but that would be an "evil" act by most DMs I've played under. How would good characters get around social issues where they can't blatantly lie or force people to do their bidding ( i.e. Bluff/Intimidate for the non-magical variety).

[vitkiraven]

1. Initiate of Mystra or Dweomerkeeper
2. See above
3. Silent Still Charm Person / Dominate / Mindrape etc.
4. See above

Not that kings and peasants matter in a setting with Epic level casters (unless the kings ARE Epic casters in which case the point is mot)

Very true, and those can all be ways to get around the challenges. Although if there are any to witness any of the events, that might know what is going on, then that might still effect the outcome. Granted, Wizards can do anything, everyone already knows that. I know Mystra is campaign specific, so should this be limited to non-campaign specific things?

And while we are presumably looking at level 20, this has already been done in Tier tests, where the silver platter is handed to those with the games cheat codes. I'm not going to argue that wizards, clerics, etc... tier one TO are godmode by level 20. Most people know that, and many people realized that WOTC messed up, big time.
As for other classes, it might be a viable challenge to those that don't have the automatic I win buttons, and if we use FR, why can't we have a Warblade IHS the weave? (Since it does have a duration, that being however long this incarnation of Mystra lives).

[Tyndmyr]

What about some explicitly non-magic encounters, where there would be a penalty or inability to even use magic, i.e. Convincing a King to your viewpoint in an antimagic area, or dealing with an honor duel with a barbarian prince on a dead magic holmgang island?
Convincing magic fearing/hating peasants to hide you for some reason, or convince the people of Thay that they need to rise up against The Red Wizards, without resorting to magic, and thuz losing their trust?

If you use the dead magic rules, you're using faerun, no? So Initiate of Mystra right out kills that.

It's entirely reasonable to have encounters like "being identified using magic will make these magic hating people hostile". For fairness, encounters should also exist in which say, drawing weapons will have an equally hostile reaction. Possibly the same ones.

It's rather less so to design an outright inability to use it. That seems less like a fair class tier test and more of a "nerf the casters" game.

I would presume a mid-level challenge would be somewhat more representative of average? If we assume a normal level range of 1 to 20, a level 10 or 11 char is about midway in the power spectrum, and thus, is more representative of the average state than either extreme would be.

Also, the weave is not a condition. Those are things like dazed, stunned, etc. IHS is very powerful, but ridiculous things like IHSing away the sun or drown healing are not terribly realistic measurements of class capabilities, and are probably not very relevant to actual play.

[deuxhero]

Tier 1 test is easy.
"Can you quickly think of five things this class can not do by level 20? If yes, it is not Tier 1."

Swing a sword, epic magic, heal, unlimited use of ability per day without cheese and make a pie that turns people fanatical. (alternately, name 5 9th level Divine spells)

Wizards aren't tier one.

[Shadowleaf]

Swing a sword, epic magic, heal, unlimited use of ability per day without cheese and make a pie that turns people fanatical. (alternately, name 5 9th level Divine spells)

Wizards aren't tier one.
It's pretty easy for a Wizard to be able to swing a sword and heal himself/others. Just saying.

[Hiro Protagonest]

Swing a sword, epic magic, heal, unlimited use of ability per day without cheese and make a pie that turns people fanatical. (alternately, name 5 9th level Divine spells)

Wizards aren't tier one.

Epic level things don't count. Tenser's Transformation is weak, but with enough prebuffs, they're a force to be reckoned with in melee. Also, Mordenkain's Sword. As for making a pie that turns people fanatical, I think max ranks in craft (dessert) should do it.

Your alternative, however, is correct.

[Tyndmyr]

Swing a sword, epic magic, heal, unlimited use of ability per day without cheese and make a pie that turns people fanatical. (alternately, name 5 9th level Divine spells)

Wizards aren't tier one.

*picks up sword, swings it* No optimization needed.

Dusk giant, dragonwrought kobold, yadda, yadda.

Ray of Vecna.

Reserve Feats.

Ah, pie that turns people fanatical, you say? Well, we have polymorph any object(or a ludicrously spell pumped disguise check), then diplomacy. Or outright spells.

I also have multiple alternative ways to do most of those, many of which are core only. *yawn*

[Xiander]

I like this idea.

I do accept the fact that making a test which accurately shows tiers is about impossible, but if a test could be devised to approximate tier, it would have the great bonus feature of being applicable to builds rather than classes.

[OracleofWuffing]

As for making a pie that turns people fanatical, I think max ranks in craft (dessert) should do it.
Use prestidigitation, flavor all the other pies in the world to taste like crap. You're now the only guy in the world that has a good-tasting pie. :smalltongue:

One often-overlooked thing in various class power tests is that part of the versatility of plenty of classes comes from their ability to gain foreknowledge of an event before actually undergoing it.

[Dayzgone]

I am whole heartedly behind this. Mostly bc ive been trying to make something similer for a while now. I take this more as a multy build tester though. Not just a single class tester.

I do remember that versatility is what defines a tier. So maybe have certain tests be exactly the same, but u can’t use your previous method of solving the challenge

[Tyndmyr]

Note also that two people, given the exact same char, might also play it substantially differently. So, it's more of a "this char is at least this tier" system. It's always possible for someone to play suboptimally.

[MukkTB]

What level are you running the tests from? I personally feel that lvl 20 is not representative of a large number of peoples games, and that lvl 1 would be a question of rocket tag. I'm actually more interested in lvl 1 than 20 because our group plays characters that always start at lvl 1 and we don't normally make it to 20. However a lot of groups start at 3+.

I'm pretty sure that the tier system was developed for the area from around 7 to 14. I would have to go read through JaronK's article to get the exact number.

I propose that if we want a exhaustive test we should include multiple levels. Possibly 1,5,10,15,20. On the other hand that multiplies the amount of work required by 5 so it might not be worth it. Either way we should be very careful what level we select to test.



Another thought is that PCs normally operate in a group. How well something does solo is interesting, but in practice its normally surrounded by about 2-3 buddies. If you test for solo you are testing in a different environment than the environment that you actually want to have knowledge about.

We also already know the tier system well enough to place things generally where they belong. The only thing further testing could do is give us a more exact set of relationships between the classes. After all, the tier system's primary use is to explain how well certain classes will play with others.

I propose that a 'test' be a controlled adventure with 2 characters running. Over the course of the adventure keep track of how well each character does. Things like their total damage output, how much damaged they tanked by sucking it up and evasion, how many times they contributed to quest completion, how many times they acquired loot for the party ect. Just remember that when a healer buffs your damage output, the value of his buff belongs to him.

If the adventure is long enough to somewhat normalize random chance we will then have physical evidence about the difference in performance between the two characters classes.


The test adventure should include multiple kinds of problem. There should be a day with one big encounter, and an iron man day with many encounters. Diplomacy, combat, and skill should all be relevant at one time or another. The overall quest should probably be on a timer that can be failed if the party takes too long.

[Dayzgone]

Over the course of the adventure keep track of how well each character does. Things like their total damage output, how much damaged they tanked by sucking it up and evasion, how many times they contributed to quest completion, how many times they acquired loot for the party ect. Just remember that when a healer buffs your damage output, the value of his buff belongs to him..


This may solve the scoreing issue preatty easily.

And really this entire thing is going to be a substantial amount of work regardless of how well organized it may get, and why it’s never been fully seen through to begin with

[MukkTB]

There are so many 2 player combinations that its silly. And how well 1 class performs in a particular pairing doesn't translate to how well it preforms in another pairing. However if we have multiple comparisons then we can start reasoning a>b, b>c, a>c.

[maysarahs]

Wow so many posts already! I'll take a moment to answer as many questions as I can in regards to how I would do this.

I realize this isn't a great idea and that thought exercises are the basis of what most tier rating is based on, but I mainly want to do this to quell cries of Schrodinger wizard, and the like where people can quote any combination of limited resources that would solve the situation at hand. We all know Wizards can do everything, but I want to show that ONE wizard (any/you choose the build) can do it all.

To reiterate, while this would be a tool to demonstrate a characters versatility, I less wanted it to be a scale to apply to figure out tiers, and more to show players how having options is ultimately better than say, ultimate blasting ability (to show why I'd prefer a tier 3 game to a party of uberchargers and a healer). However if I ever come out with something near serviceable, it would of course be useable for anything anyone would like to run it through. You draw your own conclusions yourself.

A quick reminder for people who already posit hypothetical builds or ask about specific spell combinations designed to circumvent obstacles, this would be designed to measure the POTENTIAL of a character, if you have a build that totally obliterates an encounter, then you get an A+! If it does so with every other encounter for the remainder of testing, then yes, your build is indicative of Tier 1 material.

The scoring system I envisioned was never more complex than a simple observation of how that class played. A Tier 1 should be able to solve every encounter thrown at it, Tier 2 should solve many of them easily, a Tier 3 should never have had a moment where it was stuck or had no options, etc. Some of these ideas sound great though, I'll be sure to remember to keep track of damage dealt, and think up some "encounter solving" points system.

In regards to the proper level to run this test I agree that multiple levels would be best, but that is too much to think about now, I'd rather have a one level scenario and later add the ability to swap features in and out to scale with level. I think I will design "the gauntlet" at around level 5, 7 or 10 because that is when many builds/ classes come online (at least third level spells, one more iterative, etc) and this also happens to be the "sweet spot" of most common play. Please any advice or suggestions would be appreciated, I worry I may have gotten in over my head with this!

The idea of randomized encounters per day was my initial suggestion to combat a wizard preparing all divination spells the previous day. AT BEST (I personally would randomize every possible encounter) the most heads up you'd get is "tomorrow is combat day,", or "tomorrow, expect distance to be a factor" (if you want themes)

I think I touched upon the idea of a character who is a better power multiplier than a soloer in my original post (it was long, I understand, sorry for that) by giving any character the option of hiring people (lets say any of the NPC classes?) to help them through. With this option though, hiring a warrior for you to DFI versus hiring an an adept to teleport you where you need to be will need to be fine tuned. The simple fact is that magic has some options that not only "help solve" a problem, but simply make it never there.

This needs some organization, I was thinking 4 encounters for each of the main categories of adventuring. Combat, Movement (Travel/Stealth), Social encounters, and Non combat dangers (traps?). I am at a loss for more refined categories, but feel free to change these as you see fit. The adventuring day will have a minimum of 3 encounters (one of which was meant to use most of your resources) (so limited novas, a 15 minute day is bad form) with the option of 5 or more if a character has the stamina. I'd try to average 4 though.

Keep the suggestions coming! I appreciate the input (and any offered help of course!)

[Madara]

This makes me curious, while it is far harder to create the proposed tests, because characters are played differently by different players, I would like to propose(maybe should move it into a separate thread) a series of tests/challenges for a player and their character. So if Bob plays Smurf the Gnome wizard, you could determine his rating for the purpose of creating encounters. This would eliminate the difficulties that arise in balancing a party.
Would anyone be interested in this separate branch of thought activity?

[MukkTB]

Why don't you buy a module from a prominent source and use that to run the test? Preferably one the testers have never played before. You could run a group or players with henchmen one at a time. And then record exactly what happens for multiple types of class in moderate detail. It would give you a pretty good grip on what happens in 'practice.'

[MukkTB]

I'd volunteer to run a badly optimised Tier 3 character. We could use Teamspeak 3 and google docs or whatever method you prefer.

[W3bDragon]

I think this is an interesting idea. However, obviously, it can very quickly get bogged down with the sheer size and complexity of the required testing.

To ease things a bit, and get this off the ground, I propose the following:

* Pick a level to start. 12th level seems reasonable. Most builds would be complete by then. Full bab classes have gotten their 3rd iterative. Both wizard casting and sorcerer casting classes have gotten 6th level spells. Don't obsess too much over it. Just pick one and start. You can adjust later pretty easily.

* Design encounters with a few levels of success. For example, an encounter with a horde of goblins that are holding prisoners. Low success would be defeating the horde of goblins or just rescuing the prisoners. Moderate success would be defeating most of the goblins and rescuing most of the prisoners. Total success would be defeating all of the goblins without allowing any of them to escape, and rescuing all the prisoners. This would give a more intricate marking system that would still be clearly defined.

* Grind them to a halt. Keep piling on the encounters until a character hits total failure a specific number of times. So we could say that encounters would rotate between combat, problem solving, etc, over and over and over, with a variable time frame between encounters, until 3 failures are accrued or 10 successes. That way, you can measure the eventual breakdown point (or lack thereof) of a specific class or build.

* The whole henchmen thing I'm a bit wary of. Perhaps cap it at one henchmen of 9th level using only single-class NPC classes. Using a henchmen starts you off at a lower level of success automatically and adds a new level of possible failure in failing to keep the henchman alive. I know that some classes shine much more with others rather than solo, so you could give them the chance to compete using a henchman, but penalize them accordingly. Otherwise, I could see all melee-type builds just using an Adept henchman to buff them while invisible and stay out of the fight, which isn't very realistic. I wouldn't be opposed to throwing out the whole henchman thing completely.

* Most importantly, you'll need to very carefully pick what resources are available for character creation. For example, you could limit it to core+completes, or core+completes+ToB+any one setting specific book, etc. Just make sure your selection is reasonable and doesn't favor any class over another. You'll need to decide on 3.0 material. On BoED+BoVD and UA, etc. If you'll spend the most time on one thing other than encounter creation, I'd say its this. Changing this after the fact can nullify all of your past results. Note that the tighter this is, the better. You might remove some of the toys the T1 classes get from splat books, but they're just as broken in core-only.

* Magic items are going to be a bit tricky. Ideally you'd want to limit it to existing items in core+MiC or something. Also, you need to have a limit of consumables, or people can load up on them cheaply and inflate their power for the 10 or so encounters you'll run them through. Perhaps you can limit the amount of consumables they can purchase to 5% of WBL. Another idea is to measure how much of their WBL they used up during the encounters and mark that against them.

I'll gladly help in designing the encounters once we can settle on what the types will be. I'll be following this with interest until then.

[Myth]

I have a Whisper Gnome Abrupt Jaunt Conjurer / Incantatrix (http://www.enworld.org/forum/5209112-post2.html) with Shivering Touch etc. ready to be used. She isn't optimized beyond RAI levels, but with 2 flaws, the AJ variant and the spell selection she's already Tier 1 deluxe.

[Tyndmyr]

I'm honestly kind of looking forward to running a Truenamer through this. The tier system indicates that I should faceplant with this, but I'm rather optimistic.

Also, Id say toss out henchmen altogether, unless they are explicitly granted by class. This is intended to measure class power level, not how well a party does.

Just use the standard 10x price for consumables for scrolls and pots. Wands are fine at normal price provided you do not allow partially charged ones.

[Godskook]

Swing a sword, epic magic, heal, unlimited use of ability per day without cheese and make a pie that turns people fanatical. (alternately, name 5 9th level Divine spells)

Wizards aren't tier one.

1.Whirling Blade, Gishes, and buff spells disagree. Wizards can go CoDzilla just as hard as the divine casters can, it just takes more work since the buffs are more scattered and the final spell cuts of spellcasting.

2.If you think epic magic is off the table, you clearly don't know how strong TO can get.

3.Arcane Disciple(Healing Domain)

4.Arcane Disciple x5 will cover most of your bases. Especially since you have to do this 'quickly'(Indicating you don't have time to determine which divine spells are not in domains).

5.Cheezed out spellhoarding loredrake kobold wizard 15/exemplar 5 would like to have a word with you over some craft(basketweaving).

[Hiro Protagonest]

3.Arcane Disciple(Healing Domain)

Reserve feats also work.

[Xiander]

Also, Id say toss out henchmen altogether, unless they are explicitly granted by class. This is intended to measure class power level, not how well a party does.


Problem is, some class abilities like bardic music and many buff spells work much better with a party. Perhaps this could be taking into account by making a group discipline part of the test?

[Tyndmyr]

Problem is, some class abilities like bardic music and many buff spells work much better with a party. Perhaps this could be taking into account by making a group discipline part of the test?

Lots of abilities from lots of classes are more party friendly.

But the party isn't a class feature. Presence of a party is not guaranteed for every use of a class, nor is it's size or composition in any way guaranteed.

No, a class that requires a party, but cannot actually create a party with it's class features, is fairly limited, and should be judged accordingly.


Remember that low tier does not mean "terrible, and you should never, ever play it". It means that it is less flexible, and has less options. Requiring a party == less flexible.

[Xiander]

Lots of abilities from lots of classes are more party friendly.

But the party isn't a class feature. Presence of a party is not guaranteed for every use of a class, nor is it's size or composition in any way guaranteed.

No, a class that requires a party, but cannot actually create a party with it's class features, is fairly limited, and should be judged accordingly.


Remember that low tier does not mean "terrible, and you should never, ever play it". It means that it is less flexible, and has less options. Requiring a party == less flexible.

I am not saying that party friendliness should determine tier. What i am saying is that a build which lies on the border between two tiers, but dramatically increases in usefulness when accompanied by a party might be judged to be in the higher tier.

Also D&D is by default a team game. In most instances of play you will have comrades around at least some of the time. Which leads me to believe, that a build which functions as tier five alone but tier three with a party might in reality fall between the two rather than in the lowest of the two. YMMW.

[Tyndmyr]

I am not saying that party friendliness should determine tier. What i am saying is that a build which lies on the border between two tiers, but dramatically increases in usefulness when accompanied by a party might be judged to be in the higher tier.

Also D&D is by default a team game. In most instances of play you will have comrades around at least some of the time. Which leads me to believe, that a build which functions as tier five alone but tier three with a party might in reality fall between the two rather than in the lowest of the two. YMMW.

Synergy being what it is, two classes are ALWAYS better than the sum of their two potency. Synergy will vary wildly depending on campaign, party makeup, party size, and all sorts of other things.

But see, synergy is not part of class makeup. It is thus a confounding factor that will, instead of giving you a more precise score, increase variability and make the score less precise.

[Xiander]

Synergy being what it is, two classes are ALWAYS better than the sum of their two potency. Synergy will vary wildly depending on campaign, party makeup, party size, and all sorts of other things.

But see, synergy is not part of class makeup. It is thus a confounding factor that will, instead of giving you a more precise score, increase variability and make the score less precise.

Good points, I mostly brought it up because it struck me as a variable that would effect a characters effectiveness in a non-trivial way.

I do however get that tiers are mostly about how many different situations you you have tools for in your toolbox, and how effective those tools are at dealing with their corresponding situation.

[Prime32]

Uh... anyone familiar with the Same Game Test (http://dnd-wiki.org/wiki/Dungeons_and_Dragons_Wiki:The_Same_Game_Test)?

[Godskook]

Uh... anyone familiar with the Same Game Test (http://dnd-wiki.org/wiki/Dungeons_and_Dragons_Wiki:The_Same_Game_Test)?

My only objection with that is it was written by someone who clearly didn't understand the breadth of options Rangers have.

They can easily keep up with fighters, and with swift hunter builds, can easily keep up with rogues.

[kme]

Why don't you buy a module from a prominent source and use that to run the test? Preferably one the testers have never played before. You could run a group or players with henchmen one at a time. And then record exactly what happens for multiple types of class in moderate detail. It would give you a pretty good grip on what happens in 'practice.'
This.

Designing a test the way you imagine it will be extremely difficult and complex. It would be easier to just design an actual adventure and run some random or chosen players through. Also, it wouldn't be very useful in gauging the objective power of classes. You will mainly just determine the creativity and D&D/optimization skills of players involved.

If you want to know the potential power of a class, there is no need to go beyond(or bellow?) Schrodinger wizards and other similar concepts. After all, the tiers are assigned based on potential power, not test.

TLDR: If you want to know the practical power of classes, just run normal adventures, preferably different in style.


What would be more interesting/useful, is determining the effort/power ratio of classes. When played by a noob player how does fighter compare to wizard? We all know the answer for that one, but what if it's a casual player who is already familiar with the rules (and even reading about optimization on internet)?

[Tyndmyr]

Uh... anyone familiar with the Same Game Test (http://dnd-wiki.org/wiki/Dungeons_and_Dragons_Wiki:The_Same_Game_Test)?

Actually, it was the first thing I thought of when I saw this thread. That said, the information there is not terribly specific, and the challenges tend toward the explicitly combative.