I've seen time and again that someone mentions someone has a low or a high AC for his class. So I'm wondering: Is there something like a list for what AC the core classes should have per level (averagewise or optimalwise)?
What are tiers?
Tiers are a ranking of how "powerful and/or versatile" the various 3.5 base classes are, with low numbered tiers being considered more capable than high numbered tiers. It's important to remember that certain caveats apply to the rankings:
- Tiers assume similar levels of optimization. Someone playing an optimized "weak" class (like a fighter) and using its abilities well may be a lot more effective than a poorly built wizard played by someone who doesn't know how to make use of its options.
- Tiers attempt to describe power over levels 1-20. Classes will generally be in their listed tiers immediately, though the gaps between tiers tend to be a bit smaller at lower levels.
- Tiers are based on published material only. Homebrew and house rules can and will modify the rankings of some classes or even just negate the entire ranking system.
- Tiers are based on relatively high-magic games. In a low-magic setting the rankings will be mostly the same, but the gaps between tiers will get a lot bigger, because magic items tend to be the best way for less powerful classes to cover up their weak spots.
- Tiers look at characters' ability to solve problems of any sort, not just combat.
We frown on link-only answers, so I'll go ahead and summarize the full tier list of all published classes, originally from here. Fuller descriptions of why each class is in its tier can be found here.
Tier 1: Wizard, Cleric, Druid, Archivist, Artificer, Erudite (Spell to Power variant) — Can do anything and everything, often better than lower-tier classes that supposedly specialize in that thing.
Tier 2: Sorcerer, Favored Soul, Psion, Binder (w/ online vestiges), Erudite — As powerful as tier 1, but no one build can do everything.
Tier 3: Beguiler, Dread Necromancer, Crusader, Bard, Swordsage, Binder, Ranger (Wildshape variant), Duskblade, Factotum, Warblade, Psychic Warrior, Incarnate, Totemist — Good at one thing & useful outside that, or moderately useful at most things.
Tier 4: Rogue, Barbarian, Warlock, Warmage, Scout, Ranger, Hexblade, Adept, Spellthief, Marshal, Fighter (Zhentarium variant) — Good at one thing but useless at everything else, or mediocre at many things.
Tier 5: Fighter, Monk, Ninja (both CA & Rokugan versions), Healer, Swashbuckler, Soulknife, Expert, OA Samurai, Paladin, Knight, CW Samurai (with Imperious Command), Soulborn — Good at one rarely applicable thing, or mediocre at one thing, or simply too unfocused.
Tier 6: CW Samurai, Aristocrat, Warrior, Commoner — Objectively worse at their specialty than another (often Tier 5) class, without anything else to show for it.
Tier 7: Truenamer — Apparently received no actual playtesting, mechanics as written simply don't work. See this question for more details.
DMG 274 suggests these values for AC as part of the table for defensive CRs (HP also factors into defensive CR, but is not listed here):
- CR 0-3: 13 AC
- CR 4: 14 AC
- CR 5-7: 15 AC
- CR 8-9: 16 AC
- CR 10-12: 17 AC
- CR 13-16: 18 AC
- CR 17+: 19 AC
According to the guidelines, the ultimate CR of a creature is the average of their offensive and defensive CRs. However, the DMG encourages additional tweaking and adjustments to individual monsters beyond the listed guidelines:
Alternatively, you can determine an appropriate AC based on the type of armor the monster wears, its natural armor, or some other Armor Class booster (such as the mage armor spell). Again, don't worry if the monster's AC isn't matching up with the expected challenge rating for the monster.
There are also a number of features that a monster can modify the effective AC for CR calculation purposes. These are tabulated starting on DMG 280. For example, giving a monster magic resistance boosts its effective AC by 2, which might result in a defensive CR that's higher than the one calculated strictly from HP and AC.
Thus, there is going to be significant variation in the AC of monsters at any CR, because there are so many factors that can change the final AC relative to the CR, including the judgment of the person designing the monster.