Well my observation is that if you care about optimization and being one of the most powerful PCs then you do certain things and you care about certain things. My point was that not everyone engages in that game and they don't need to engage to have fun. The people who have played rogues in my games in older editions definitely had weaker PCs in combat. Some people just don't care. If you care and can't get what you like then of course you are not going to have fun. My observation is that not everyone is obsessed with whether their character performed better in combat than someone elses. YMMV.
My observation is that a lot of people who think that they don't care about this stuff actually do, just not in nearly as conscious a way. That people who you think are fine with weak characters are, in at least some cases, simply not speaking up because they think it's "the way it is".
I have a specific example. One of the players I've played with for 30-ish years, for the first 10+ years, whenever we played AD&D 2E, played a Thief (single-classed, human). He seemed to enjoy it, though I think part of that was that I ensured he had magic items that allowed him to keep up with the other PCs more. He never complained. In other systems he played other things, though still usually leaning Rogue-ward. In 3.XE, he played a Rogue, and honestly, 3.XE felt so bad for literally everyone who wasn't a caster that I don't think his experience stood out (this was true even at relatively low levels - though full BAB characters also felt okay at lower levels - ones who were neither full BAB or casters though? Ouch). Skills weren't hugely rewarding-feeling in 3.XE so that didn't help. Then he played 4E. 4E rejected this notion that it was fine for some classes to "just suck". 4E Rogues were badasses. Suddenly, his demeanour changed. We always thought his slightly rueful attitude to combat and the like was merely an expression of his personality, but in 4E? That clearly wasn't the case. He was actually having fun - and he expressed that in very clear and literal terms, saying that he finally felt like he was playing the character who he'd always wanted to play in D&D, finally really having fun in combat. He even started optimizing and making really sensible decisions re: powers and so on. 5E he played a Rogue again, but didn't have as much fun as 4E, and now his go-to class is Warlock, because despite never having complained during the 2E and most of the 3E era (he did voice a concern along the lines of "I wish I had spells..." in late 3.XE), he's seen what it's like to actually have fun in combat, and doesn't want to go back to not being like that.
So I strongly suspect a lot of other people are the same way, and until they get to play a character who is highly effective in combat, they won't really express it. Some might not even then, but I'd be interested to see if they kept on playing combat-ineffective characters.
It's not really about winning the module, it's about looking good doing it.
Even if the players consider themselves as a group I think this is true. My experience over 30+ years is that players don't actually put a lot of weight on "beating" an adventure. Some, but not a lot. They put a lot more weight on having cool characters, doing cool and memorable stuff and so on, and that depending on the system, that often does interact with optimization. Players often set short-term win conditions of their own too, like getting their own back on a specific NPC for they feel as screwed them over or something, even if this wasn't part of your plan as DM.
@pemerton Still agree with most of what you're saying but I feel like Rifts is the killer example of where the "Well just picking a class or choosing a weapon isn't optimization" breaks down, at least in terms of being
meaningful. This is particularly relevant to me now because I've been comparing Rifts and the newest version of Savage Worlds Rifts. By your definition of optimization, there's absolutely no question that optimization in terms of picking specific edges, specific skills, spending stuff carefully and so on is vastly more possible within Savage Worlds Adventure Edition (SWADE) than Palladium Rifts. With Palladium Rifts, optimization is pretty limited - you pick either an Racial Character Class (RCC) or a race and an Occupational Character Class (OCC). You roll some stats and pick some skills, but the actual optimization is pretty limited. It's pretty much the same skills which are optimal for all characters, and in many cases some/all of those skills simply come with the class. But Rifts is vastly, insanely unbalanced, because the reality is, a lot of RCCs and some races and OCCs are just ridiculously better than other ones, like laughably so. It's been that case since the beginning, too, and only got worse over time. Knowing which RCCs, OCCs, and simple equipment choices were just waaaaay better than the others (combined with a limited number of choices re: skills) was huge. A lot of it required you to understand stuff that wasn't immediately obvious, too. Like, yeah, a Glitterboy is obviously powerful and doesn't require much in the way of choice-based optimization, but something like the Hunter Cat RCC buried deep in South America 1 is also ludicrously powerful, and to know that, you need to know it exists, and understand why it's so powerful, which feels awfully like optimization to me, and unbalances the game more than a lot of points-based or similar optimization. Likewise with knowing about guns and armour buried in various books, and understanding why certain ones were particularly OP because of the way they interacted with other rules
Contrast this with SWADE Rifts, where whilst you can, in theory, optimize more, because you have more moving parts to work with, more choices, and so on, and you see that actually, SWADE is far more balanced (though not without issues), and in-depth knowledge of the system and books is far less likely to benefit you to the same degree.
I don't know if this entirely undermines what you're saying - I don't think so - but I feel like you can have games where your definition of optimization is quite limited, but which are grotesquely unbalanced, and ones where there's a ton of optimization possible, but the end result is pretty balanced in most cases.
