I always thought that having damage determined by class rather than weapon would solve this.
This is exactly the sort of thing I'm getting at in my posts.
The game has weapon charts on which some weapons are more effective than others. And it provides class-based access to those charts - part of playing a fighter is having access to better parts of those charts than would be the case were one playing a magic-user. This is, in effect, a damage-by-class system, but designed (
by Gygax presumably not quite; see
@lowkey13's post a little upthread) in a way that favours "realism" (swords are more effective weapons for warriors than are daggers) over sheer aesthetics (
I want my guy to be a knife fighter!).
As a result, there is no rule that stops a fighter player choosing to use a dagger and no shield (just as there is no rule that stops a MU player choosing to memorise only Affect Normal Fires, or Wizard Mark, or Detect Magic and Identify, in his/her first level slots). But I don't think the result (a moderately ineffective fighter; a moderately ineffective wizard) is what the game is designed around, although the 5e maths may be forgiving enough to cope with it.
But unless the player is actively seeking to be ineffective, if the table is prepared to sacrifice "realism" than it seems obvious that no imbalance will result from allowing the knife-wielding fighter to do class-appropriate damage.
Yeah, I believe Dungeon World does it that way, too
Correct. Although there are also "tags" on weapons that aren't class-dependent in the same way. And each class has particular weapons on its gear list which are correlated with the damage die (at least somewhat - I've just had a quick scan but didn't look at every classes damage die and gear options). Eg the paladin does d10 and has longsword and halberd on his/her gear list; the ranger does d8 and has bows and shortsword or spear.
That was my reaction when I first read 13th Age. Armour class is class dependent as well
In 4e AC is very close to class dependent, in so far as each class has an armour list with one or two optimal choices, and has a (correlating) expected use of DEX or INT as a main stat.
Because 4e feeds it through a version of D&D's traditional AC chart plus a version of the 3E idea that heavy armour limits the DEX bonus, there are some convoluted outcomes, like primal classes that for aesthetic reasons are confined to light armours but that don't have DEX or INT as a main stat and so get a class bonus to AC to compensate. (And then this patch itself breaks down in some places like the DEX-based barbarian in Primal Power, which I've seen mentioned in some threads as the basis for a broken build when obviously what it needs is a nerf on the class AC feature which is there simply to patch the issue that results from the original conflict between class aesthetics and traditional AC methodology.)