Constructing A Balance Standard

[Hanuman]

What is mechanic balance? How does one make a standard of effect?

As I've been taught, balance is something to be compared vaguely in reference, but there's no real way to quantify balance that's advertised.

Share your experiences, references, articles and cites please.


I wonder if it's possible to create a standardized playtest scenario to test the balance of classes and such?

[bobthe6th]

well... there is the issue of the d20 being random.
I hear the same game test (http://dungeons.wikia.com/wiki/Dungeons_and_Dragons_Wiki:The_Same_Game_Test) is good, but I never used it.
The problem is a lot of balance depends on so many factors, the vast majority of which are random. So you have a new class? well, checking for balance depends on 1)number of sources available, 2)party composition, 3)the other players optimization level, 4)the quality of the monster...on and on.

[Grinner]

Balance is a tricky subject. It tends to be subjective by nature and heavily dependent upon the circumstances. For the purposes of this post, allow the definition of balance to be "sufficiently similar in overall power as to not overshadow other options".

The first thing to be considered is the system and optional rules in play, since some systems will naturally favor certain styles of play over others. For instance, I have it on good word that speedier characters tend to fare better in Shadowrun combat.

Also worth considering is nature of the setting. What might be appropriate for one setting would be totally inappropriate in another. Trying to mesh these things together just complicates the whole process. In this respect, the tier system is probably one of the greatest tools the D&D 3.5 community has developed.

Finally, and probably most frustratingly, the GM himself can drastically alter the efficacy of any given character option. Whether through total lenience or sheer favoritism, certain character options can be made more viable by presenting challenges that they are particularly well-suited for resolving.

Frankly, the whole idea of balance is laughably abstract for the purposes of design.

[Hanuman]

Interesting points, the SGT was really nice (thank you).

I am talking more about 3.P dnd, but I'm looking for a way to balance homebrew I make so that it doesn't outshine other classes which it's kit.

[bobthe6th]

Well for classes you could make a generic four man party for each of T4-1. Make a dungeon crawl that is a challenge for each tier group, and keep records of the groups run. Then pick a tier replace one of the four and run them through a dungeon crawl made for that tier. Then do it agian, replacing another character. Keep doing this, switching tiers when you have replaced all the characters once.
You would get some data, and could compare to the control... you would probably get a good idea of the tier of the class.

[Grinner]

Well for classes you could make a generic four man party for each of T4-1. Make a dungeon crawl that is a challenge for each tier group, and keep records of the groups run. Then pick a tier replace one of the four and run them through a dungeon crawl made for that tier. Then do it agian, replacing another character. Keep doing this, switching tiers when you have replaced all the characters once.
You would get some data, and could compare to the control... you would probably get a good idea of the tier of the class.

I'm tempted to say that a program could speed this process up a bit.

[Glimbur]

Well for classes you could make a generic four man party for each of T4-1. Make a dungeon crawl that is a challenge for each tier group, and keep records of the groups run. Then pick a tier replace one of the four and run them through a dungeon crawl made for that tier. Then do it agian, replacing another character. Keep doing this, switching tiers when you have replaced all the characters once.
You would get some data, and could compare to the control... you would probably get a good idea of the tier of the class.

This is a lot of work, but you knew that. It's also dependent on you correctly identifying who to replace: if you take a wizard paladin rogue cleric party, then test factotum by replacing the paladin, you will get a rather worse impression of him than if you replaced the rogue. Most of us know the factotum should replace the rogue, but it's less clear with, say, an Incarnate, a Binder, an Ozodrin, or other homebrew stuff.

Yours is not a bad idea, and I don't have a better one, but this is a hard problem. Which is probably why it hasn't been solved yet.

[bobthe6th]

This is a lot of work, but you knew that. It's also dependent on you correctly identifying who to replace: if you take a wizard paladin rogue cleric party, then test factotum by replacing the paladin, you will get a rather worse impression of him than if you replaced the rogue. Most of us know the factotum should replace the rogue, but it's less clear with, say, an Incarnate, a Binder, an Ozodrin, or other homebrew stuff.

Yours is not a bad idea, and I don't have a better one, but this is a hard problem. Which is probably why it hasn't been solved yet.

Yeah, it would be a pain and a half to run. I was mostly suggesting the way were you could be certain of balance. Not fesable for most classes, but an option if you have a free month(or a free week with the dungions/generic parties set up) and want to be sure it is a good option.

also, you replace all of them, and see what happens. the tier system is meant to show versatility... so if you can use the druid as every member of the T 3 group, you know it isn't tier 3. When it's role starts to matter in the T1 group, you know it is probably T1.

The power portion of the brute force test(or BFT for short) comes from comparing the run where you replaced the appropriate class to the control. So if a Home brewed Beguiler like class ends up in solid T3, and can replace the beguiler, and it does way more within its specialties, you know to dial back the power a little.

[Hanuman]

Ozodrins are grapplers and should replace frontline units. They can tank pretty decently with the defensive ab. feats and 12 eyes.
They don't have to, but should not be less injured than any other D8's or below, generally.

Good advice though, I think I'll use this data to playtest some options.