That’s my feeling as well, although I’m not so down on System Shock 2.
Have you read Jon Chey’s post-mortem of SS2? It’s got some good info on why certain aspects of SS2 are the way they are. The IncGamers interview with Jon Chey and Dorian Hart is also illuminating.
I disagree with you a bit on the linearity question. If you think about it, traversing the individual levels of Citadel Station was pretty linear. In addition to backtracking, some levels required you to complete specific plot points before the higher levels would unlock. That’s not very different from how System Shock 2, and Ultima Underworld 2, and many other games are structured.
On stealth, someone from the original team can correct me if I’m wrong, but my impression is that the Looking Glass and Irrational folks who’d worked on Thief felt they’d already done stealth in Thief, and so didn’t want to emphasize that in their next game. That said, I lean toward agreeing with the feeling that, despite the very specific training paths at the start of System Shock 2, the Navy path never felt as fully fleshed-out as the Marine and OSI paths. You could specialize in combat, or psi, but you pretty much had to pick up the Hack skill no matter what, leaving the roguish playstyle not so much a choice as a necessity. (This is an observation others have made over the years.)
Really, for me the single most meaningful difference between SS1 and SS2 was the addition of the RPG stuff. Adding learnable skills in SS2 expanded the mechanical play options enormously over SS1, which emphasized power progression through the acquisition of new gear (mostly weapons). It also added to the narrative by giving SHODAN the opportunity to express her character by awarding (or retrieving!) cybermodules for upgrading skills.
I’m honestly conflicted about this. I actually think the RPG elements were handled very well in SS2 – they were fun to optimize, they created interesting choices, and they mattered to the play experience. I enjoyed them!
And yet… I find the original System Shock more memorable, not just because it was first but because that experience was, as you say, more focused without the RPG choice-making. As well-integrated into the game world as upgrading Stats, Weapon, Psi, and Tech skills was, and as enjoyable as using Psi and Tech skills was (Stats and Weapon skills were essentially passive), it’s fair to say they were so integral to success that focusing on them pulled me out of the world somewhat. They felt a bit game-y.
That wasn’t the case in the original System Shock. No RPG elements meant less “stuff” one could do in the game, and less opportunity to build a character that felt functionally unique or personally relatable. But it also meant that what remained constantly front-and-center was the world of the game itself. There certainly were mechanical moments (“Nice jump, human”), but they were moments – by far most of the game was a continuous emphasis on “Oh, man, what’s around this sharp corner, and am I wielding the best weapon to meet that threat?”
In a funny way, having fewer tools in System Shock ratcheted up the challenge, required more ingenuity, and emphasized the aesthetics of the world more constantly.
So why do we complain that the BioShocks, which were also simpler than System Shock 2, “dumbed down” the experience?
This is where I come back to the level design of System Shock, which I think was geometrically and dynamically much more complex than any of its descendants. If you’re going to eschew RPG mechanics, making the gameworld more continuously prominent to the player, then you’d better build that world to be complex and diverse and reactive – and I think that correctly describes System Shock, but less and less so the “Shock” games that followed it.
Basically, you can simplify the mechanics, or you can simplify the world, but you can’t do both if you want a game that’s appealing to thoughtful, creative gamers.
So: my personal preference for System Shock 3 would be a game that minimizes RPG skills and maximizes the active dynamics of the world. But my guess is that the OtherSide Austin team will wind up making a System Shock 3 sequel with RPG elements, because they think fans of SS2 will expect that, and whose world is only moderately dynamic, because they’ll believe they should be making “a game, not a simulation.”
That prediction could be wrong if the SS3 team can reuse a good chunk of the “Improvisation Engine” code from Underworld Ascendant. I think there’s a fair chance of this, given Warren’s stated interest in player-centric game design… but maybe economic reality dictates a more traditional structure for a System Shock 2 sequel that needs to earn back its development and advertising budget.
Sure would be nice to hear some official news on the System Shock 3 front.