What is SS3's unique feature that no other game has done before?


Warren Spector mentioned recently that SS3 will feature something pretty crazy that no other game has done before, he’s not even sure it will work. What do you think it could be?

I’m guessing it will be something to do with the warp drive / cyberspace side of things, SS2 left that open at the end and SS3 will of course build on SS2.


I would be shocked (no pun intended) if the Cool Thing isn’t in some way about character development.

My first guess is that we somehow get to see the world through SHODAN’s eyes – perhaps to relive her awakening to consciousness, and to experience first-hand the events that twisted her into the creature that decided to destroy humanity starting with the inhabitants of Citadel Station.

Playing as a distributed – and deeply disturbed – consciousness could be pretty strange.


I would like to see SHODAN begin to respond to the hacker’s “in-game” actions by her own voice cues, as she mocks you from above. A good example would be whether or not to save the prisoner in Prey - SHODAN remembers your decision, and mocks you down the line for earlier actions on behalf of the player-character. Also, maybe some sort of emergent AI design for SHODAN, where she learns from the player’s decisions (perhaps informed by others’ decisions, and utilizing some form of “cloud computing” to give SHODAN a more reactive personality, and response). If the protocol droids are not putting up much of a fight, she arms them with turrets; if you hack them too often, SHODAN ups their security; and you require a higher level of hacking to be efficient.

I also hope for references to nanotechnology, cloud computing (AKA Skynet IMO), and other technological innovations that have happened since the 1980s. Since SS2 had a 1980s view of cyberpunk, I would expect it to retain that, but be informed by the technology of the “now.” In other words, historical accuracy upon the technology we are building in the present.

In fact, I can’t help but think that SHODAN would be benefited by cloud computing, in a 4d sort of wall-breaking way. She a ghost with in the system, who fools with the player. Why not shadow their every footstep, and respond interactively with the manner in which players are playing the game.

She is a landmark villain is two respects, a) She is omnipresent. As the hacker, you cannot get away from her. She is a ghost in the system - everywhere at once, and b) Because she is a ghost in the system, she “everywhere at once,” and nowhere killable. As a ghost - if she cannot be killed, she cannot be defeated.

I would also argue that a third aspect of SHODAN’s personality needs to be brought into focus, which is c) She is morally ambiguous. The hacker merely “removes” her limits. He does not impose “narcissism” on her, or any of these traits with which she is maligned. SHODAN is merely a product of her coder’s design. A snippet of programming language, but without limits.

I think this begs future exploration.


I wouldn’t call it “proof,” but these comments from Warren in his RedBull interview are not making me think I’m wrong:

"I want to include Citadel Station, even though it was – apparently – destroyed in the original System Shock. I want to include Shodan as the primary adversary, obivously. And I want to get back to the Hacker as the hero. In tone – fear, paranoia, tension – we'll follow the lead of System Shock and System Shock 2, for sure. But will the game pick up directly after the second game? You'll have to wait and see."
"What I will say is that we'll be investigating some of [SHODAN's] motivations from the earlier games, instead of just treating her as if she were insane. She deserves better than that. She's super intelligent and should behave that way. She isn't insane; everything she does makes sense to her. We're going to play with that in System Shock 3."

Other thoughts?


I think this is heading in the right direction. SHODAN responding to the player’s actions could include camera footage where the player got injured, “hacker’s blindspots”, moments where the player showed mercy to other npc’s, “weaknesses worthy of exploit”, and inexplicable “moments of insanity” where the player displays acts of altruism.


I think it’s less that SHODAN doesn’t understand the concept of altruism, it’s more that the altruism of insects is amusing at best, annoying at worst, but ultimately insignificant.


Giving SHODAN her own AI would be nice - one thing I would like to add. I was not a fan of SS2’s ending. I think as an ending cut-scene, the game should string together the actions of the player, showing how they had informed the gameplay and impacted the narrative. It would be morally ambiguous, but more in line with what an emergent simulation is trying to achieve - unscripted actions, and the consequences thereof. Lastly, it’s been brought up in other threads, but I wouldn’t mind SHODAN breaking the fourth wall tactfully to mess with the player. I had other thoughts earlier, but forgot them in the heat of the moment - I’ll try to remember them later.


There is an interview where Ken Levine talks about the ending of the game: http://irrationalgames.com/insider/what-might-have-been/

The End? Rewrite.

“Due to miscommunications or differing ideas, a different cinematic video was created from the one that I originally scripted,” says Levine. “It had this elaborate sequence where Shodan would attempt to kill you in a double-cross, as this ‘cyber stinger’ that was in view provided tension of your impending doom.” Upon getting his hands on the video for the ending sequence, Levine didn’t see anything that he wrote in the script. “We didn’t have much to work with. It was like when you look in the cupboard and you’re trying to make soup, and you have a bag of salt and couple of pinto beans.” Working with fixed assets can be extremely challenging especially with limited time and resources as well as fighting the technology back then. Levine remembers, “We had to write to the assets we had at that point, and all we could do was edit it. We completely ran out of time and that cut scene wasn’t the right ending for the game.”

Breaking the fourth wall… It needs to be done very carefully. My first experience with it was with Undertale, and I thought that was obtrusive, but still fun. All it did was shut itself down and change the intro and load menus. I don’t want to think my computer actually gets hacked or anything. A game, if launched in admin mode has full access to my computer and even if it’s proven to be safe, I don’t want it toying with my trust by faking a hacking attempt. This kind of stuff would seriously piss me off and make me uninstall even the most enjoyable game. You just don’t pull these stunts off with me, or maybe someone else too. It’s very much a morality thing.


Probably a better example would be the Psycho Mantis battle in Metal Gear Solid, where you had to switch controller ports to stop him from reading your mind. Very, very tactfully.


I’ve had another thought about this. (Uh-oh. :smile: )

As noted in the OP, Warren has mentioned SS3 offering something not seen in any game before. If that still holds true, it just struck me – what if that New Thing is an ability to rewrite reality (in the game)?

  1. It fits with what SHODAN was doing at the end of System Shock 2.

  2. It would explain how Citadel Station is back in some way.

  3. Editing reality according to rules could be a highly systemic ability – a key part of the fun of Looking Glass-style games.

  4. Editing reality in some specific ways sounds a lot like what Warren was exploring with the paint-and-thinner concept in Epic Mickey, so he’d be building on that.

One argument against this is that it sounds pretty cerebral; that’s a very different kind of play experience from the exploration-with-weapons focus of the original System Shock. Another counter-argument might be the old “it’s already been done” objection if, in fact, some other game has used a rewrite-reality mechanic.

Still, I like the feel of this notion enough that I thought I’d throw it out here for consideration.


There’s also the possibility that we might never know the answer, since “not even sure it will work” means what it says.
Which is likely a large part of why they haven’t gone the kickstarter route of telling everyone everything about it, and are instead developing the sensible way, without revealing stuff before it’s baked.

Most of the areas less trodden that I can think of in game design relate to NPC AI. Shodan is certainly one where the AI could be played with, but in general, as a bad guy, Shodan would not be present: rather, off in the shadows, working the strings.

Instead, there is a plethora of other enemies: robots, cyborgs, mutants, et al.

My dream for NPC AI is to have it realistic: not just fight or flee, but a far more complex and nuanced system of decaying levels of aggression, with many mechanisms to de-escalate, including natural decay; combined with every beating or death having repercussions. But that’s not likely to happen (not be at all acceptable to players) in a SS sequel.

But what about an AI where it is assumed that most mobs can inter-communicate at all times, because they’re robots/cyborgs, and are on a network? They would not all bum-rush you, but also wouldn’t stand in convenient groups of three separated by just slightly more than their own aggro-radius. They would instead set up co-ordinated attacks, use flanking and surrounding and ambushing maneuvers, attacking to drive you forward into a trap, or if you attacked them, pulling back towards an ambush.

Unfortunately, done right, even the weakest fully-coordinated defenders would obliterate the player, so again that might not be acceptable to players.

How’s about having systems actually hackable? Like, you would need to do legitimate hacking on computers, real rewiring of circuitboards to bypass security, disarm bombs, and so on?I know of many hacking games, but none who’ve actually done it right.

But I think the most likely is that it’s an incremental move in the direction of immersive simulation. Extra levels of freedom in some way. Other than having the structure of the level completely destroyable/modifiable by the player, though, I can’t see anything that hasn’t already been done in that space. Perhaps, a more complex web of interactions and repercussions of the player’s emergent actions?


Full party-poop/scoop mode on…

The ‘re-write reality’ mechanic strikes me as a variation on what immersive sims are supposed to be about anyway i.e. you influence and re-write the story as you go along. It didn’t really happen with UA - which bounced the idea around no end early on - and I doubt it’s going to happen with SS3 in a wider sense.

It also scares me witless. First, that someone would think it’s possible (god, the sheer number of permutations of systems that would have to work flawlessly together), and second, testing it.

I say: get a working game with a half-decent plot first. There are no guarantees that this will happen - it’s hard.

The last thing I want to hear after the barren desert of UA is a voice calling from wilderness promising brightly sunny cities on hillsides. You’d think they’d learn. Warren.

And if you’re going to say it, at least do it once you know it’s semi-working…not hedged around with caveats.

Talk about hostages to fortune, and bright shiny new objects, dead cats on tables.

It sounds a bit desperate to me, as if desperate is the new normal. Let’s just make sure we start with a game that merits the name, and a place in the series, first.

The hacking-thing sounds a bit more do-able (yawn, if you must ) but that hardly qualifies as a ground-breaking new feature.

Enough of ground-breaking. Let’s have ‘…works…’ first. And all the above is predicated on an over-arching AI many streets ahead of what we have now, in any form.

(…this killjoy announcement has been brought to you by someone with a functioning recent memory).


Sounds about right, nothing in UA works.