I'm not calling 4e monsters bland. I'm calling monsters without weaknesses that show up in game bland. I put equal weight/blame on encounter design for 3e and 4e in that regard.
You asked if we were talking about encounter design, and, in a sense, no, I'm not talking about
encounter design. I'm talking about
monster design.
I
am talking about monster design primarily as it relates to the encounter, however. And you're correct to not read too much into my examples...they're just there for facilitating understanding (well, that was the intent, anyway).
I'm, in a nutshell, primarily saying that a monster isn't there to fulfil a role. It may "fall into" a role, sure. Naming that role helps to categorize monsters (especially for dm selection). But my point is that monsters are there to be evocative monsters...and that any given monster might fulfil multiple roles. It might also be poor at some roles, and rather than searching out a different monster, or creating a variant, it can be good adventure design and good writing to use the monster in a sub par way to highlight its weakness.
Regarding the ogre archers and the distinction between encounter design versus monster design...I'm designing a monster who has pros (strength and hp) and flaws (clumsy). The rock throwing/archery point was meant that the ogres would be bad at it. I'm making it an easy encounter to highlight a flaw of the monster. The Ogres do not want a ranged encounter (but it makes sense in context that one would occur). The player (characters) are happy to have a ranged encounter against these up-close monsters. If there are "ogre rock thrower" variants who are bad up close, but great ranged attackers, then we don't have this phenomenon occurring.
I'm not really discussing 4e or 3e in any way...I'm mostly theorizing about what I believe to be good design. You (Pemerton) state:
But if you're saying the game will be better if monsters have more uniform statblocks, and all the work is done by them exploiting numbers plus the terrain around them, I think I want to see some evidence for that. Because it is somewhat contrary to my own experience.
I think that's what I most need to clarify.
No, I don't mean exploiting numbers or terrain (at least not necessarily). I mean, simply put "let goblins be goblins".
What would the goblins do to keep invaders out of their lair? Some would do x, some y, some z. I think there's an advantage to having "regular goblins with bows, regular goblins with short swords, and regular goblins holding a defensive line". The point there is that if they are good enough at more than one thing (rather than breaking them down into super-specialized sub groupings), then more than one meaningful encounter can take place, and it will be different, but the same. It's the sameness I'm arguing for as a good thing...that players develop a base understanding of "goblin" and know that the, say, archery might be bad for themselves, but that they'll wipe the goblins when the critters are forced to fight hand to hand.
To put it another way, I think there is something lost when there are goblins who are good archers for the archery combat, and goblins who are good tunnelers for the tunnel ambush, and goblins who are brutish for the elite combat. While this varies the combats, it makes goblins more meaningless as a race (because in 2 levels it's orcs who are good at all three, and then bugbears, ogres, giants, etc). But those are comparisons amongst very similar species.
What about elves versus goblins, versus dwarves, versus mephits? Each has their own schtick, generally, but that schtick is lost between the tough elf, the flying goblin, the nimble dwarf and the tricky mephit variants (if they exist).
I've not thought much about edition so far. I think everything I mention has been a problem in both 3e and 4e. Designing a monster for a role (named or not) is the issue I take ubrage with.
However, let's bring up 4e. One great element is that goblins, orcs, elves, mephits, etc. are all, on a base level, quite different from one another. I'm not arguing to erase that difference...I applaud it. What I'm arguing against is goblin skirmisher versus goblin lurker versus goblin elite (here we're editionless again, as this also occurred in 3e). If its a goblin only campaign, then that variety is welcomed...the players will deal with a lot of e.g. goblin lurkers and get to know them. But if it's a one shot adventure to hit a goblin lair...then let the players learn the enemy, and give the enemy more than one thing to do. Sure, maybe their king is tougher (an elite or a brute or a solo). But hit players with the same goblins several times so the players learn what they are good at, moderate at, and poor at...and let the players adapt to this. Allow players to force goblins to get close because they're weaker that way. Do not make each goblin encounter essentially a "new monster" by introducing too many variants of "goblin". Yes, every encounter will be new and challenging, and different, but the continuity and the learning curve suffer.
Well...I tried to be more clear...but it's late here, and I'm not sure I was. Hit me with questions for more clarity if ya like and I'll post a response tomorrow.