fett on 26/5/2011 at 02:06

---

Tim, Laura Baldwin, and Sarah Verilli. Al - feel free to post this to your site like you did the others. Thanks to Pardoner for the work!

Tim Stellmach: Hi, I'm Tim Stellmach, I was employed over 6 or 9 [years] or something at Looking Glass, so I was there quite early. This would be starting on Ultima Underworld 1 actually, when I was in playtest. I started in the design role, as lead designer on Underworld 2.

Matthew Weise: What years were these, just to get a little context?

Tim Stellmach: Must have been '91 or '92, on Underworld 1. Yeah, it would have been 1991. And then for Underworld 2, which was sort of the first time I had authorial input, really...there's an extent to which you have that in playtest, but it's pretty limited. So that would be '92, '93 – long time ago now. Then I was also one of the designers on Terra Nova, and System Shock, and I was lead designer on Thief and Thief 2.

Matthew Weise: So the games were?...just for clarity's sake...

Tim Stellmach: Underworld 1, Underworld 2, Terra Nova, System Shock? System Shock, Terra Nova? (laughs). It was a long time ago. And then Thief, Thief 2.

Matthew Weise: And then also for the sake of clarity, can you just say what your role was in each one of those too – the actual title?

Tim Stellmach: Tester in Underworld 1, and lead designer in Underworld 2. On System Shock I was...I don't know if he had a lead designer on System Shock, it was more just a...cabal, as Valve would have put it. And on Tera Nova I was a designer and programer. I did some of the tools and used the interface and stuff in those days. And Thief, both games I was lead design on those.

Matthew Weise: Okay, anything else you want to say? On to Laura?

Laura Baldwin:I'm Laura Baldwin. I came onto Looking Glass pretty late. I've been trying to count on my fingers here....probably about '97. I worked on Thief, and System Shock 2, and Thief 2. I was sort of a backfield designer. I was only halftime, so I wasn't in charge of any levels – but a lot of the stuff I did was sort of dialogue - “Come up with ten different ways for a guard to say, 'I think maybe I've heard something.'” Then ten more ways for them to have maybe, *maybe* have seen something, and then when they've definitely seen something. So that was a lot of writing into spreadsheets, I did conversations, I did a lot of random things like editing motion captures. I guess there are three authorial things I'm proud of. I did the bear pits conversation at the beginning of Thief.

Matthew Weise: Oh man!

Laura Baldwin:I came up with the word “taffer,” and I did the “dogeral vet” the Trickster says at the end of Thief. There's this big ritual, but he wasn't saying anything. It's a ritual, he has to say something! And then probably my favorite thing is when that got translated, the audio recordings came back with the people in French, and these people in German, saying these poems that I had written, translated into French and German, and they still rhymed. That was really cool.

Matthew Weise: Yeah, because rhyming was a big aspect of the whole Trickster part of the story. That's cool. And then, the bear pits conversation is literally the first thing you hear in game, right. I mean, you start the game...

Laura Baldwin:There's a briefing and voiceover...

Matthew Weise: But if you just start the level, that's the first thing you hear, even in the demo.

Laura Baldwin:Yes.

Matthew Weise: I think everyone remembers that.

Laura Baldwin:And it's actually one of the longer conversation. I think everything else got cut to four lines because otherwise it took up too much time, but the bear pits got to stay longer so people could just go and listen to it, I guess.

Sarah Verilli: I joined Looking Glass just a little bit before Laura. I did playtesting on System Shock 1, and then I was a lead playtester on many of their other games, either Solar Joint, or Flight Unlimited, Tera Nova ...and then on Thief I started as the lead playtester.

Tim Stellmach: Then you tried to quit and we had to stop you. (laughs)

Sara Verilli: Yeah, I tried to quit Looking Glass because I was tired of being lead tester (laughs). Or at that point I was the quality assurance supervisor, was the actual title, which meant I was getting further and further from playing games and more and more from telling people how to play games. That got me shifted into a designer position on Thief. Then I was the lead designer on Thief Gold, and then lead designer on the Thief Gold project which promptly got canceled on us. Which is about the time Looking Glass went out of business.

Matthew Weise: Speaking of which, I recently saw (on your website, Tim) the videos from the final days of Looking Glass. Which apparently you only put up recently.

Tim Stellmach: Mike Chrzanowski did the videos. He had them at one time – I don't know if it was on his personal website or what – but only recently he got them converted over and got them all up on Youtube, around the ten year anniversary of Looking Glass closing down. And that was, I think, the first time they got spread widely. There was a lot more feedback to them when he did that.

Matthew Weise: It's interesting you mention it– I looked at at few of them, and you mentioned being forgetful or hoping you could remember stuff that happened fifteen ago. There was something about Makh LeBlanc in the videos talking about the severed heads, the puzzle in System Shock 1 where ...

Tim Stellmach: I always loved that puzzle, and I found out by watching the videos that I helped come up with it!

(laughter)

Matthew Weise: There's also a story about your driving, somewhere on the road and you have this idea, because it's really late or really early...

Tim Stellmach: We were coming home from the office, I think it didn't count as late at night anymore. And again, Makh remembers this but I do not, and it may be that I don't remember it because we were up all night. Nothing was really going on in my brain at the time apparently, except coming up with this severed head puzzle. Being that we had these severed heads laying around on Citadel Station in System Shock to show all the violence at the pre-game when Shodan had taken over the station, and somehow we came up with the idea that these heads had to be a puzzle somehow. I think by then we already had art associated with each one. If you picked one up and put it in your inventory, you got a shot of one of the developers, basically, in severed head form. And we just loved that so much that we had to call attention to it I guess. And our means of doing that of course, was a retina scanner that locks down a door that you have to get through. You have to find the person who was authorized to go through that door, and bring their head to the retina scanner.

Matthew Weise: It had to be the right severed head, because there were a LOT of severed heads.

Tim Stellmach: Yes (laughs), there were quite a few, so you had to find the specific one. I think the idea was that the audio logs had portraits of the speaker as well, so you had to match up the portrait in the audio log with the inventory image of the severed head, and figure out who was the right person.

(laughter)

Laura Baldwin:If I recall correctly, the audio log also gave some hint about where that particular person had been.

Tim Stellmach: It was the log that let you know the person was authorized.

Laura Baldwin:And it said, “he was working in this location...” it would sort of tell you where to go to look so you wouldn't have to look through every single head.

Tim Stellmach: And I totally don't remember having any part in coming up with that.

Matthew Weise: I feel like this is a really good segue to one of the things I wanted to hear more about, because that's a really memorable moment from that game. Or a really memorable puzzle or aspect that's really well integrated into the fictional universe of the game. It's not like one of those Resident Evil statues that's just sort of “here” and you have to fit these switches on it, and why would anybody build this, because obviously no one would be able to get around the mansion if this were really here. One of the things I remember about playing that game at the time is that the game design, the world design was so fictionally meshed together.

Tim Stellmach: Yes, and that has to do with the process that occurred to develop it. It wasn't a puzzle that had been though out ahead of time and the elements in the game meant to fit it. There's a lot to be said on this entire topic when it comes to Thief because Thief 1 and Thief 2 were very different in that specific regard, but it was an element that came out of work we had been doing on the game, so naturally it was very closely integrated into what was going on the game world because it emerged from that through the process of integration as we went through elements that were there that inspired us to include this puzzle.

Matthew Weise: So from a development process point of view, was it like the game world was designed first then largely? What were the original ideas? Was it like, “we want to have a space station, we want to have a bunch of dead people in it, and we'll figure out the puzzles later?”

Tim Stellmach: That was literally the very first thing that I remember, and it was a reaction to the Underworld games that we were sort of constantly chaffing against the limitations of conversation systems, and their inability to be more than thematic conversations. Even in RPG's today, the most common model is some menu of utterances that you can say at any given time that will lead you through this graph of possible paths in the conversation. It's not really much like having an actual conversation, it's more like a mini-game that's themed on conversations. Probably Doug [Church] in particular was sick of that. Doug has always been very interested in this problem of how to model character in game. So after two games of doing that sort of thing in Underworld, there was the idea in System Shock that we should apply some Ikedo to the problem and sort of - “doctor this hurts,” “well, don't do that.” The solution in System Shock of course was, there are no conversations because all the NPC's start off the game dead. You get to know them, but you do so through these logs that they left behind and the disasters that happened all pre-game. Which continuing to this day if you play the Bioshock games, it's exactly the same kind of narrative model. What was the question now?

Matthew Weise: The question was - it seems to me what you're saying is that the puzzle design of the game emerged more out of a desire to design a believable, seamless world.

Tim Stellmach: Well, not even just puzzle design specifically, but that whole game system design. Like the decision to have audio logs as opposed to a conversation system, came from this narrative decision, or vice-versa actually, right? The narrative decision to have this pre-game disaster was based on this limitation of what we wanted to attempt in the game.

Matthew Weise: That reminds me of something Austin said last time. He said that part of this idea came from Underworld 2, which you said you also worked on. For people who might not know, Underworld 2 has several worlds in it, and one of them was a tomb where everybody was dead. He said in some sense, System Shock was almost like extending that idea– maybe the kernel of that idea is kind of [from there].

Tim Stellmach: You know, I hadn't thought of that , but I don't have anything to say against it. The thing about Underworld 2 in that regard is the need to make the worlds that you went through distinct from each other led to a certain amount of stylistic experimentation in order to draw those distinctions. So it's very natural then. I don't know if I though about it at the time, but I wouldn't be surprised if Austin did – he was the writer (laughs).

Matthew Weise: He explained it that way which I though was interesting. Speaking of dialogue and the problem of dialogue, it would be interesting to hear some of Laura's ideas about this because your job was to write dialogue for this, kind of like an extension. So Thief was one of the following games that was kind of revising some of these other concepts.

Tim Stellmach: Where everyone's alive, but you don't actually talk to them – you eavesdrop on them.

Matthew Weise: To me, that always felt like an extension of trying to look for a solution. So one solution was killing people and finding logs, so how do we do it without having people dead? They're talking to each other, but you can't talk to them. Was that conscious?

Tim Stellmach: I think so, yes. But I didn't actually have to execute it, so...

Laura Baldwin:I wasn't a designer, I was sort of a task-carry-outer. I do think back in the early evolution of games there were adventure games where you wander around and do things. Computer role-playing-games, which are not really like role-playing-games, sprang out of that. In an adventure game, everything is limited – there are four things in the room that you can interact with, and they can only interact with each other and you. In computer role playing games, there's a lot more freedom to wander around and poke at things. The conversation systems are still very adventure-like. There are still four things you can say to this person, and then they can only say those things back to you. When I was brought on, it was already a sneaker at that point , so there was never really going to be conversations with Garrett talking to these NPCs. But, we wanted it to feel like a world where these were actual people. There were motion captures where they were standing around and fidgeting, to make them feel like actual people you were sneaking by, and who got bored, and if they were going to be emitting these utterances every so often - “Oh I've noticed something,” and “Oh, I'm looking for something,” and “Now I'm not looking for something anymore,” they couldn't be talky, but then they had to be at least a little bit talky to each other.

Tim Stellmach: It makes them more authentic, that's a really good observation. The needs of the gameplay in Thief means that they have to be really, really chatterboxes about all their innermost feelings and observations, because that's how you tell their AI state.

Matthew Weise: It's feedback.

Tim Stellmach: The whole point of the game is to outsmart these AI on some level, and this is the information that you're using to do so. But if that was all they ever talked about, it would be really unnatural.

Laura Baldwin:Also sometimes you're getting information from them – what are different thiefy, sneaky type things you can do? Like you can steal information documents, you can overhear information, and that gave you another thing to do as you were sneaking around. Get as close as possible so you can hear them, and often their dialogue had clues about where something was.

Matthew Weise: That's something Shakespearean about the convenience of eavesdropping in Thief. Like in Shakespeare, someone's always talking – speaking in earshot of an important character. What struck me playing it – it's very believable. I don't know if it's quality of writing or quality of acting. Some people complain that the AI is dumb like, “why would anybody do that?” But for me, there was something about the way the characters were stupid that seemed very believable. (laughter) Years later I was thinking, “why do I like this game so much?” and it's something about Thief 1 that almost feels like a nerd revenge fantasy. What's it about? It's about this weak, smart, witty character...

Tim Stellmach: The bear pit conversation sets all of that up as well. It's the very first conversation you eavesdrop on and one of it's purposes is to sell what buffoons the guards are.

Matthew Weise: Exactly.

Tim Stellmach: They're meant to be. When people say [they're dumb] - It's a game about outsmarting the guards. What do you expect? If they outsmarted you, it wouldn't be much of a game.

Matthew Weise: That brings up interesting topics about the inter-relation of characterization, voice – acting, writing, and game design, specifically AI design. Because the AI design, the voice-acting, and the writing meshed together to create something that, to me, doesn't feel broken. It's expressing something coherent about a fictional world.

Tim Stellmach: And your question about the writing and acting and such, that's all tied in to some degree to the game structure as well. Structurally, the overheard conversations in Thief are a lot like the audio conversations in Thief or Bioshock, you come across them, you listen to them, big difference is you don't get to listen to them again, and you can interrupt them. The fact that you can interrupt them is one of the things that motivates the effort that we put into some aspects of it. If we want you to stop or slow down and take in these conversations, we need to motivate you to do that. One of the ways to do that is to embed some of the conversations with useful hints about you might approach your mission. But you need to make them really entertaining. They need to payoff on their own terms if the player is going to set aside their other activities and listen to them. That was something we had a lot of conversations about.

Laura Baldwin:Something you can't just click through the subtitles of.

Tim Stellmach: I remember talking about them very specifically during the development of Thief – that we need to make these conversations worth listening to.

Matthew Weise: (to Laura) And that was your job?

Laura Baldwin:Often, yeah. (laughs)

Matthew Weise: We need to make them great, so make them great.

Laura Baldwin:Yeah, try to make them entertaining and very short.

Matthew Weise: So where does “taffer” come from? Was that your solution? You were trying to be entertaining and give them some personality, so is that where that comes from?

Laura Baldwin:That actually comes from an original evolution that a lot of the dialogue was supp– I think it was Greg O'Pickelo's theory that there should be some type of argot or Thief's cant that everyone was speaking, and there would be words for all sorts of things. You wouldn't say “here” there would be something for that. “Eagle” I think was supposed to be cant for “I saw something.” I wrote all those long things like, “I eagle something,” and playtesters were like, “we don't understand what they're saying, this is no good.” (laughs). And when the NPCs are supposed to be conveying information to you about their state, it really can't be that unclear. So then I had to go through and take out all these made up words. But taffer was the only one that really stayed. Often I would use it as a swear word. “This taffing around there?” or “who's the taffer who did that?” They're guards, they're not very bright, they're probably going to be swearing, but we're trying to have a game that doesn't have swear words in it so (laughter). So that became the place-holder for the swear words and you can do whatever you want with it. You can make it a noun, make it a verb, you can do all sorts of things with it.

Matthew Weise: Much like real swear words. (laughter)

Laura Baldwin:So eventually I came up with sort of a back construction for it. There were some Hammerites talking to each other later and one says “something something taffer,” and the other one says, “don't say that, that's derived from the Trickster, and you give him power – don't do that.” I don't even know if that one made it into the game but it's in one of my spreadsheets somewhere.

Matthew Weise: This is fascinating to me because that's why I would play a game like that and imagine a singular author behind it who had everything figured out. Because that level of coherence between a small detail like that contributing to the overall world-design of the game. What you're talking about is writing interesting dialogue, but you're also talking about fictional world fleshing-out. You're sort of deepening the fictional world and creating a mythology. One of the things that really struck me about Thief is that it creates a very interesting mythology very quickly without sitting you down and just telling it to you. I'm wondering if you could maybe talk a little bit about how that came about, because I love that world, I love the characters and there's that whole nature versus technology theme. I know a lot of things have that theme, but in that game it felt very original to me, the steam-punk, kind of low-fantasy, medieval-ism thing. Where did that come from? Was it one person, was it two people? Was it an accident or sort of an accident?

Tim Stellmach: Wow. Well, it was not just one person but I would lay its origins at the feet of Ken Levine. If you look at that retro-tech kind of style, you see the same thing in Bioshock. If not the same retro-tech, the original concept docs for Thief had a lot more focus on the use of elementals as power sources, and these background elements (no pun intended) having to do with the Hammerites and their role in the world, and the Trickster kind of came as counterpoint to that in a pretty obvious way. In some sense, we didn't want the player character to be clearly on one side or the other of that. Garrett works much better as a morally ambiguous character...

Matthew Weise: Is that where Garrett comes from as a character?

Tim Stellmach: A lot of it comes as sort of in the role of the courts. He's there, in a large degree, to comment on the Hammerites and the Trickster and their conflict, and the futility of all of it. In some sense, the way I see it, we identified the Hammerites as a force of order and the Trickster as a force of chaos – in opposition to each other. They are both sort of forces of faith. In contrast to that, Garrett is sort of the voice of reason, so there's this triangle of ideologies going on there, so that things are a little more (I hope) nuanced in the story of Thief.

Matthew Weise: I certainly agree.

Tim Stellmach: But I think that all kind of started from the concept of the Hammers as a sort of initial and primary enemy that came out of Ken's concept documents. Or that's certainly my recollection of the arc that we went through developing all of those elements.

Matthew Weise: (to Laura) What writing did you do other than the character dialogue.

Laura Baldwin:That's the main thing I did. I think I did work on the manual. But it was mainly stuff people said.

Tim Stellmach: Which is a lot of what communicates the story in that game. One of the things that we were trying to do specifically was to have multiple channels of narrative going on between the overheard conversations, the briefings, and the lower objects that you could find in the missions. I have a whole list of more minor ones.

Matthew Weise: Because there's documents you pick up and read at some point. I remember there was one where the Trickster outlines his plan, which he calls “The Dark Project,” which, the way I read that, tell me if I'm wrong, was like, “We called this project 'The Dark Project the game' then we just wanted a way to rationalize a way to use that...(laughter)

Laura Baldwin:Yes!

Tim Stellmach: Absolutely! (laughter) That's entirely correct.

Laura Baldwin:It's because we couldn't come up with a better name for the game!

Tim Stellmach: It was a code name from back when we were working on an entirely different concept pitch which was Dark Camelot where you're in the role of Mordrid, who's secretly a freedom fighter against the evil Arthur, and all this stuff that never got made. All that was really retained from that was the idea of this player character as sort of this less sort of black and white hero, more of a Robin Hood character. So none of that happened, but the concept pitch name “Dark Camelot” - when we threw out the concept we still had to refer to our team as something, right? We didn't know what we were going to do, and we knew we weren't the Dark Camelot team anymore so it was just The Dark Project. And that somehow made it into the press materials and we couldn't get rid of it after that.

Matthew Weise: I remember the first trailer I saw, which still on the game is the intro, but it doesn't say “Thief” it just says “The Dark Project.”

Tim Stellmach: Right, yes. (laughter) And so their title had to become the subtitle when we finally had a title.

Laura Baldwin:It just had to be stuffed in there somewhere! (laughs)

Tim Stellmach: Otherwise it would have confused our whole sales channel. And then that bit you were talking about - “let's get this Dark Project name into the game somehow...” (more laughter)

Matthew Weise: It's an interesting constraint, to think “we have to make a game,” but then - “What does it have to be?” “Well, it's gotta be dark.”

Tim Stellmach: Well, that stylistic decision had been made early on. But even before we settled on the concept of a Thief.

Matthew Weise: It's an interesting constraint because darkness is a useful gameplay instrument.

Tim Stellmach: Darkness is your friend.

Matthew Weise: Which was another really interesting thing at the time. Sarah hasn't gotten a chance to talk yet. You did level design right? Is there anything you want to contribute to the conversation?

Sara Verilli: I actually came on as a designer, really after Laura was already working on the project, because I had seen it almost most of the way...we were...were we in beta when I came on? Yeah, I think we were quite in beta. But all the levels had been built. So much like Laura, I was a jack of all trades, fill in the blanks designer. I took a couple of levels that had been built by artists, but who weren't going to populate and script them, and I ended up populating and scripting them and setting up the traps.

Tim Stellmach: Not like Thief 2 where you played a more substantial role.

Matthew Weise: So not designing the spaces themselves.

Sara Verilli: No real physical design other than, “Oh, this needs to be tweaked here,” and fixing small bugs.

Matthew Weise: Tim and Laura talked about things they felt authorial about or had authorial connection to. Is there anything that comes to your mind that you would say, “See – that's me!” in the game.

Sara Verilli: The only thing I had authorial control over was the gameplay of the levels I worked on.

Matthew Weise: And which levels were those?

Sara Verilli: Oh good heavens (laughs).
Matthew Weise: You don't have to list all of them...

Sara Verilli: No, the problem is trying to remember them, much as Tim had forgotten it all, I've forgotten it all. There was the mansion level, when you break into Constantine's mansion. And that level had pretty much no gameplay in it whatsoever. It was this beautiful space...

Matthew Weise: Kind of Escher-esque...

Sara Verilli: Yeah, very Escher-esque, then there were like, two traps set up (laughs). It was a beautiful visual concept, but it wasn't really clear how it linked into the story, other than “you're breaking into this to advance your plot,” and then no gameplay set up in it whatsoever. I went from playing to the game to “okay, now I've been dodging these things, now can I set up interesting encounters.”

Matthew Weise: That was a crazy level.

Sara Verilli: Crazy level. I remember thinking, “Oh, holy...holy moley. This is the first level? It was the first level I started on. I actually worked on some of the other levels as well. And in Thief 2 I did a fair bit of level design and clean up as well. I forgot to mention I was a designer on that too.

Matthew Weise: You mentioned Thief Gold as well and there's something in Thief Gold if I remember correctly, that is an interesting variation on the whole “we want to have dialogue, we want to have somebody who's alive, but who you can't interact with.” I have this vauge memory in Thief Gold where you meet a character who's crazy and drunk or something, do you remember that? He's standing right in front of you but he's not dead, and he's not a guard because he's not trying to kill you, but he's babbling incoherently.

Sara Verilli: Isn't there a drunk in the back alley of one of the..? I would have put him in the back alley of the first mission...

Matthew Weise: The one in Thief Gold I'm thinking of is a little more expository deliberately.

Tim Stellmach: There is one in the Lord Bafford mission as well.

Matthew Weise: One of the things people forget is that that level (and others), start in a space where people actually aren't trying to kill you, they're just civilians on the street. It's not like Metal Gear where you're behind enemy lines and anybody who sees you is going to try and shoot you. You have to break in somewhere before you're actually doing anything illegal. There are drunks and some of these other kind of people out on the street too. But there is a bit of a subtle “going down the rabbit hole” in the first level of Thief.

Tim Stellmach: You break in through a well, in fact.

Sara Verilli: And that's the first reaction where you can decide if you want to try to distract the guards and send them off on a hunt, or if you wanna try and...

Laura Baldwin:When we were at lunch I was telling you about the mission I was working on for the Thief 2 build which is the Hammerite mission where everyone's a Mechanist and that had a drunk Hammerite in it...I don't know if you remember playing this expository event or...

Matthew Weise: No, I have a vague memory...I think one of the extra levels in Thief Gold has a Opera House.

All: Ohhh....

Matthew Weise: You get to the opera house through the ground and there's a guy near the hole. He's saying something important...

Tim Stellmach: You know, Terri [Brosius] may have written that, because that was her level.

Laura Baldwin:A lot of the levels had a lot of stuff that was written by a level designer, both the Garrett voiceovers as you were encountering things, or just dialogue.

Matthew Weise: How many designers would you say were working on...? Because I remember the end credits where everyone had the hood on, and the people are credited with the individual levels they worked on if I remember correctly.

Tim Stellmach: Right.

Matthew Weise: For example, I seem to remember Randy Smith being responsible for the really weird, messed up, horror levels.

Tim Stellmach: No, that sounds about right.

Laura Baldwin: *I think it was the Asylum.

Tim Stellmach: [pause] I think so. *But it was very much one level designer having primary responsibility for a particular level.

Matthew Weise: *But also writing, right? *It's not just “oh, I built a level, now you write something.”

Tim Stellmach: *They were definitely in on the writing of their level as well.

Matthew Weise: *I guess to me that's what's fascinating about this question. *Because I've played a lot of games where it's like there were a hundred writers, and it's like “oh, you think” because it's just all over the place. But Thief doesn't feel all over the place to me, it feels very directed and very coherent. *Was it just like everyone at the office was on a really similar wave-length? *Were you just all eating the same food?

Tim Stellmach: *Well, part of it was, I think, that we had good communication of what the style was, what those goals were.

Matthew Weise: Did you have a bible or something? *Was it written down or was it just meetings?

Tim Stellmach: There was an element of that. *There was more written documentation than ever made it in the game. *And this is another interesting thing that I generally think is a good thing in games. *There's stuff that you write not for the purposes of putting it in the game, but it for the purpose of making sure that the stuff that does go in the game is all coherently related to one particular thing. *And the corollary of that is that you don't have to communicate your whole story by telling it to the player, right? *You can leave gaps for them to fill in, with the knowledge that they might not all fill them in the *same way as the other player did, but that's interesting and gives them something to talk about when they're on the discussion boards. *They can form certain conclusions about what these things mean to them in the interstices of the stuff that you did put it. *So, partly it came out of that idea of having a consensus narrative built up fairly well within the team that was more specific than the stuff that went in the game, and if there were inconsistencies, they were more sort of in this higher level of detail that never actually made it in the game.

Matthew Weise: *When you say there's a lot of stuff that's made, that actually didn't make it in...My first encounter with that was opening up the CD for Thief and looking at all the voicefiles, and there's like 10% of what's on the CD you actually hear in the game. *There's a huge amount of incidental dialogue. *There's like: “There's a guard down at the end of the hall. *If I use a water arrow...”-
[crosstalk]

Tim Stellmach: Lots of stuff was recorded-

Matthew Weise: -You know what I mean? *But just none of that's in the game. *There's just very few moments, and I'm just wondering...

Laura Baldwin: *Sometimes it was playtesting. *If Garrett is just saying something over and over as he's talking, than he starts to sound like the guards, who are just expostulating at a drop of the hat. *Mostly Garrett is taciturn and occasionally snarky, and he shouldn't say stuff unless it's either really useful or funny. *Or sly.

Tim Stellmach: *The way the audio production works too, is that it's so much harder to go back and do pickups sections based on feedback. *The incentive there is to err on the side of writing and recording a lot more stuff then you're think you're going to use. You can always pull out as a result of your playtest feedback but it's hard to put more audio in.

Matthew Weise: It is interesting, because lots of games have a voiceover, but abuse it, right, where you hear it a lot, or it plays over and over again. *I think Garrett speaks like four or five times through the whole game? *But you really remember each time, because it's at exactly the right moment, where it creates a punctuating character moment. *He always struck me as a really interesting character, and I'm wondering who was involved in the voice casting for that? *Did the same people who were involved with part of the writing process do the direction for the voicecasting?

Tim Stellmach: No, the direction for the casting was on the audiovisual team which would have been Dan Thron and his people, which was pretty much separate from the writing. *He worked pretty closely with Terri (This is again Terri Brosius). She did most of the cutscene authorship. *So they worked pretty closely together, but he didn't work directly with the mission people. *It would have been Dan Thron, and Greg LoPiccolo on the first game, who did the voice direction and casting.

Laura Baldwin: Dan's also the voice of the stupidest guard.

Tim Stellmach: Yes. *Dan is a surprisingly good voice actor.

Matthew Weise: Oh, the bearpits guy. *The ongoing-

Sara Verilli: *Yeah, he gets named Benny in Thief 2, I think because he was so popular.

Matthew Weise: Is it the one guy who reappears all the time? Even if you kill him, right?

Tim Stellmach: Well, Dan does several voices himself as well. *Benny the Guard being a recurring character, but Dan actually did several guard characters and Hammerite characters, as I recall. *Stephen Russell, who did the voice of Garrett, has a very wide range, vocally, in terms of pitch and dialect and accent and everything, he did a surprising number of those characters as well.

Matthew Weise: So he's like all over the place in the game, but you wouldn't know it?

Tim Stellmach: Yeah.

Laura Baldwin: I saw Stephen Russell in a Huntington play, last season. *I was like, “Holy cow, that's Garrett!” *He didn't sound like Garrett at all.

Tim Stellmach: *Well yeah, makes sense.

Matthew Weise: Was he a stage guy, primarily?

Tim Stellmach: I don't remember what his other work was.

Matthew Weise: *It sounds like it would be interesting to get the people who were more directly involved on the day to day with that. *I'm just interested in this topic, because so much of the videogame voice-acting seems to come from this voice-acting talent pool, which is largely defined by television animation. *Or at least comes from that kind of culture.

Dennis Dyack told me a story once about wanting to cast the original Legacy of Kain, and he asked for actors from one of the main talent people. *She was like “Oh, I got all these people who worked on the Jetsons, and all this kind of stuff. *And he was like “Okay, what else have you got?” *And she was like “Well, I've got these theater people.” *And he was like “Well, why don't you give me the theater people. *We're making a serious adult game here.”

I'm not saying other voice-acting is bad, necessarily. *It just seems like there would be so much more places to draw from for videogame voice-acting that it seems like people don't go to.

Laura Baldwin: *There were a lot of programmer voices too. *Not for the really big voices, though.

Tim Stellmach: No, we were still using...

Sara Verilli: The servants were...

Tim Stellmach: Yeah. *We did like an open casting call on the team to see who we could get. *Which you could not do nowadays.

Matthew Weise: So this is making me think of System Shock too, weren't there a fair amount of developer voices?

Tim Stellmach: Oh, there was even more stuff like that.

Matthew Weise: It's like all over the place.

Tim Stellmach: *System Shock was entirely nonprofessional voices, except for Terri who did the voice of SHODAN. *And she was not primarily a voice-actress, she was a singer. *But at least she was professional.

[laughter]

Matthew Weise: As a voice-performer, she had professional experience doing it. I know Sara was talking about faces before. *You keep saying it's not your voice, but you're one of the faces.

Sara Verilli: I'm Rebecca Lansing's face.

Matthew Weise: You're like the Ed Harris character in Apollo 13, in System Shock. *You're like the person who helps you [the player] through the whole game.

Sara Verilli: Well. that's my face. *Someone else's voice, as I said.

Tim Stellmach: No, your voice is in there.

Sara Verilli: My voice is in there, but I don't remember where.

Tim Stellmach: All I remember is that you're the character my character is leaving voice-mail for [laughter] because I can't make it to dinner [laughter], because the reactor is melting down or something. *Char, Charlotte or something.

Sara Verilli: Charlotte something, yeah.

Tim Stellmach: I just remember that I was Mark and I had to say “Char, this is Mark. *Listen, I just can't make it to dinner tonight, 'cause...”

Laura Baldwin: The great thing about being voice-actors is that now we're on IMDB for it, so we can connect ourselves to Kevin Bacon.

Tim Stellmach: That's true.

Sara Verilli: I think my biggest role in any of the Looking Glass is actually in Flight Unlimited, where I'm the voice of all the airports.

Tim Stellmach: Oh, I forgot that.

Matthew Weise: Oh, there's airports?

Sara Verilli: Well, the FEOs.

Matthew Weise: I just remember crashing and respawning.

[laughter]

Sara Verilli: If you manage to actually contact someone and say you want to land, and you hear someone give your flight numbers and your runway numbers, that's me.

Matthew Weise: One of the things that interests me about the history of the Looking Glass games is *the non-first person shooter evolution of the first person immersive, narratively rich game, whatever you want to call it. *And of course we have things like Bioshock now, which is a sort of evolutionary offshoot of that. *To its credit and popularity, it can easily be identified as an FPS, where it's harder to identify a lot of the Looking Glass games. *For me that's awesome, because I liked that kind of richness.

Tim Stellmach: *Well, there's this whole argument (and not that anyone's arguing it much these days), when System Shock 1 was first out, there were people who would object to other people calling it an RPG and they'd end up in this big discussion: “Is this an RPG.” *If it had been top-down, and you'd been Link, for example, then it would have been an RPG.

Matthew Weise: It's almost the same design on paper.

Tim Stellmach: *The big issue was that there was no character sheet, and that all the character progression arc, in terms of abilities over the course of the campaign, had to do with getting more inventory items and not more stats or skills or things on your character sheet the way it was in Ultima 1. * If you look at that model, it's exactly the ability progression model in the Legend of Zelda games. *Which no one argues about whether they're RPGs. *But since we were in first-person, I guess no one thought of that or something.

Matthew Weise: It was also early first-person too, right?

Tim Stellmach: Oh, yes.

Matthew Weise: The conventions were still solidifying, and what exactly does first-person mean hadn't really solidified yet.

Tim Stellmach: I think there were even fewer points of reference in those days. *When #1 came out, it was a month or two after Wolfenstein 3D came out. *For the longest time, it was us and really stripped down early iD-style run and gun type shooters. *If you look at even things that are squarely in the middle of the FPS market these days, there's a lot more to those game systems than was going on in *DOOM. *At the very least, there's cover, for example.

The idea that there's a little more breadth in your choices...first person is not a genre. *It's one element of your format, that's all, and it doesn't necessarily determine all these other genre decisions. *In the early days, it seemed like people thought that that was the case, because they only had a couple of examples to draw from, in terms of what the game could do.

Matthew Weise: Even in System Shock 2, for example (and I know that wasn't Looking Glass entirely), you're kind of going down a more conventional RPG path.

Tim Stellmach: Oh, much more so.

Matthew Weise: Like with stats, and everything like that. *What's really interesting to me about the first System Shock, *is...I don't know what you call them, I tend to call them in conversation: discrete tool-based character advancement system.

Tim Stellmach: Sure.

Matthew Weise: *Because your improvements are not like levels of effectiveness of items. *It's like “I have a thing, and it does x, and I have it or I don't.”

Tim Stellmach: It's much more qualitative than quantitative.

Matthew Weise: It's interesting that you mention Zelda, because that's very much what it's like. *Like Zelda, or-

Tim Stellmach: Or Metroid.

Matthew Weise: Or Metroid. *Which actually why, when the first Metroid Prime came out in 2002, I remember saying “Wow.” *To me that was almost a sequel to System Shock. *It was like a science-fiction game...It was tool-based, right, because it was Metroid.

Tim Stellmach: Because it was Metroid.

Matthew Weise: *I do have a sort of strange question, it's like a fan question that I can't articulate in any other way. *Why is the lean so awesome in Thief? Because there are so many first-person games that have a lean. *Like, it's a first-person game, I've got my blinders on, I'm playing an FPS with mouselook, and I go to do my lean but the screen kind of [makes creaking noise]. *There's something about the way the screen moves, and I don't know if nowadays they try to map the lean to some kind of polygonal character that's actually there, for body awareness or something?

Tim Stellmach: That may be. Garrett did not have a model. He had arms. There was this rough physics model for Garrett. He's a snowman. Secretly in System Shock, you were also a snowman. Because of the time. Your physics could not handle the kind of skinning-

Matthew Weise: Did you say he's literally a snowman?

Tim Stellmach: Yes. *You were a pair of spheres!

Matthew Weise: Oh, okay.

[laughter]

Tim Stellmach: In terms of what shape you are.

[crosstalk]

Sara Verilli: Not that snowman.

Matthew Weise: I thought you put an actual snowman. *Like buttons and a nose.

Tim Stellmach: No, no. *There is a place in an older model-

Sara Verilli: There's a box.

Tim Stellmach: -for Garrett in the game. You're sort of a Mr. Potato Head-looking brown, rectangular box.

Matthew Weise: Oh, that's kind of awesome. We should be using [unintelligible]

Laura Baldwin: With an arm.

Tim Stellmach: But in terms of your physics model, you're a snowman. *One thing that results in, is that there's a lot of control over the properties of that model, because it was all done in-house, by Looking Glass, by our physicists. *So to the extent that that came out well from a [unintelligible] point of view, the credit probably goes to the degree of control that they had.

Laura Baldwin: The chopping motion. All the stabbing, chopping motions. The fact that Garrett was a box as opposed to a person actually made it very difficult.

Tim Stellmach: Oh.

Laura Baldwin: We originally motion captured all sorts of stabbing and chopping and slashings. But when you take just the arm part of that motion, and try and play it...Try to imagine stabbing or chopping if the rest of your torso is bolted into a frame. You're very wussy.

Tim Stellmach: Well, what your arm is doing doesn't make sense any more, out of context of what your shoulders were doing. He has no shoulders.

Laura Baldwin: So we ended up ripping out all the motion captures, 'cause that just wasn't working.

Tim Stellmach: The other thing about leaning in that game, which sort of gets back to this issue of the amount of iteration we were able to do in the development cycle there, and the degree to which some of the features were not planned ahead of time, and were sort of improvised during the process. *And that's the listening at the doors, which is only occasionally crucial, but is often really useful, and no other game I know of ever does it. If you lean sideways into a door, you can hear what's going on the other side better than if you weren't doing that. You can actually press your ear against a door in Thief. Little known fact.

Matthew Weise: That's like an emergent verb, in a sense.

[crosstalk]

Tim Stellmach: There's no listen-at-the-door action you can do, right? *It's leaning in such a way that it brings your upper snowman suit-

Laura Baldwin: On the other side of the door, which blocks sound.

Tim Stellmach: -into a door. *It's not the sort of thing you would go out of your way to support, I have to be honest about that. *But the fact that our sound propagation in Thief was so good. *Because we did not want to deceive you about where things that you were hearing were. *So there was a lot of work put into making the sound propagation quite naturalistic. *So sound would propagate properly through doorways, and not through solid walls and stuff. *But because of that, we had all this data, so we could attenuate sound when there was a closed door, so that wouldn't fool you. “Okay, now that we're doing that, we could stop attenuating sound when you lean up against the door.” Like I said, I don't know how much that gets used, but it's delightful that it was so easy to do, and that we had the inspiration to do it.

Matthew Weise: It feels like you had put these pieces in place, not for the purpose of doing this third thing.

Tim Stellmach: Oh, if you proposed to do all that work for the purpose of listening at doors, you would have been insane.

Matthew Weise: Yeah, but you just realized, there's all these dots, all I gotta do is connect these two, and then this whole idea makes sense.

Tim Stellmach: Which is like the severed heads.

[laughter]

Matthew Weise: I'm wondering if anything else jumps out to any of you that feels like that. *Any other examples of the whole process.

Tim Stellmach: Can't think of one off the top of my head, I'd be thinking about Thief Gold and Thief 2, though. That's the point where we had all this understanding about the system, right, and you can really play with what it's capable of a lot more than we were able to do on Thief 1.

Matthew Weise: Is there anything you feel like really wasn't planned from the beginning. *Because I know Sara's mentioned to me multiple times that certain things in Thief came together really late, near the end of the process.

Tim Stellmach: Yes.

Matthew Weise: Are some of those just fortuitous things that happened? *I know for the first game, that the game was changed a lot.

Tim Stellmach: Yes.

Matthew Weise: When was the moment when it kind of became the thing that people know that it is. *What confluence of elements, I guess, would...

Laura Baldwin: One of the things, was that the initial AI was much more like a camera connected to wheels for the NPCs. *They would see you and they would behave. As people playing it felt like it was sneaking, it became more clear than you had to actually have a more complicated AI, where they had an internal state of how much they are aware of there being something out there, and what they are doing based on that. You couldn't have a simple AI, that homes towards you until it doesn't see you anymore.*And there was this one period where the AI was being ripped out and replaced with a better AI. That was a turning point that I was there for, as opposed to one the ones before that.

Tim Stellmach: *All of that represented a much narrower focus on stealth than we originally identified. *People often refer to it as being one of the early modern style stealth games. *You had things like Metal Gear Solid, and Commandos and other stealth games that were coming out around the same time. *To start with, we had much broader points of reference. *We were looking at Fritz Leiber's Fafard and Grey Mouser stories. *Fafard is not exactly sneaky, he's this hulking northern barbarian type guy, and the thing you draw from that is this pulp adventure notion of a thief. *And this is something we've talked about a lot, where there's elements of Thief 1 that come from that, the sort of Lost City stuff with the Burricks, that aren't playing as much into the stealth metaphor. *In Thief 2 we got away from those and focused much more on the stealth game. In Thief 1, we hadn't identified that yet at the start, “this is a stealth game”. We start with “this is a thief game,” and we're looking at a lot of points of reference of what a thief could be. Dungeons and Dragons was one of them, and the Fafard and Grey Mouser stories, and the Batman comics. *Batman in the role of Garrett, not a thief. But that was a lot of the stealth right there, came from that.

Matthew Weise: Sara knows I've said this before, but I definitely have gone on record saying: Thief 2 is a great game, but I'm actually the only person I know who actually liked that more eclectic definition of being a thief in the first game.

Tim Stellmach: I like the variety. I certainly agree with that. On the other hand, I felt like in Thief 2 we had enough of an understanding of what our game system was capable of, that by exploring those possibilities we able to still have a lot of variety even though we were narrowing our focus a bit.

Matthew Weise: Oh, absolutely. I definitely wouldn't say there's less variety in general in the second game. In that narrower focus, you push it farther and deeper under that one theme, a little bit more Ocean's Eleven, it's a heist. In Thief 2, there's that amazing level, of course I was playing on the hardest level, where you have to ghost, where you have to not knock anyone out at the police station, which is insanely nail-biting. I remember the first game, when I first played it I was like “Oh cool, this is the level where I'm just sneaking into someone's house, and oh my God, this is the level where I'm Indiana Jones.”

Tim Stellmach: Yes.

Matthew Weise: I remember finding that really thrilling, and there was a real sense in the first game of “what are they going to think of next?” There's this amazingly weird level at the very end. One of my pet peeves with a lot of games is that when you played the first level you've played the whole game.

Tim Stellmach: You've sort of seen everything it has to offer.

Matthew Weise: In Thief 1, it was surprising up until the end. There were fresh ideas even in the second half of the game. There's a level called “Undercover” where you dress up as a Hammerite, and you have to be in plain sight of other people, and I was wondering how that level came about. Was it just like “why don't we have a level where you're in a disguise?” That's something that didn't get picked up in the stealth genre again until the Hitman series a few years later.

Tim Stellmach: I don't remember that level specifically. I remember the early development of the mission tree in general, that was one of our big goals, mine in particular. And that came out of this idea of having to explore what did it mean to be a thief. Because we didn't have a lot of prior points of reference in videogames to work from. And I mentioned some other game influences, but with regard to that specifically one of the things I remember at the time really thinking about, I don't know if anyone now remembers the game Thieves Guild, by this small press up here called Gamelords. I guess it must have been around 1980, in one of the early waves of post-D&D games, where they were very much in a D&D type world, but everyone was a thief. The problem with that was that if you look at what D&D had been doing at the time, in terms of their published adventure and everything, there was no thieving going on. You had a thief character class which did no thieving in first edition D&D. So they had this problem when they were doing that game, of really having to flesh out that idea a lot. “Okay, if we want people to play that role, and have some variety in our scenarios, what are we going to do?” They actually, in their printed supplements in Thieves Guild, they organized things by these categories of thieving activity that you would do. They had a chapter which was like 'Pickpocket Scenarios' and 'Armed Robbery Scenarios' and 'Burglary Scenarios' and 'Tomb Robbing'. * went back to that at that time, and based on not just that list, we went and started brainstorming “what are all the thievy things you could do,” and I'm sure that the disguise scenario came out of that process.

Matthew Weise: I know it's apples and oranges, which variety you prefer, and there's absolutely nothing wrong with Thief 2, I guess I know a lot of fans who are down on Thief 1. “Oh, but I hate the zombies.” I love them, I write about zombies. For me, this whole thing is about searching for authorship.*It's easy for a big guy who's presented in the press as the author of the game talking about a high level about why these things happen, but when I play games and really remember what makes them feel special to me, it's these smaller moments. “Who thought of putting the severed head in the thing.”

Tim Stellmach: Which are not central planning.

Matthew Weise: But in really good games, they feel like it.

Tim Stellmach: To the extent you have an auteur in a game, I suspect that this is true of anything, their imprint comes as much from the environment they create within in the team, where everyone can contribute their best work and they can focus that. But they're not going to be the one that crafts each individual, essential element of the work.

Matthew Weise: It seems like the press still wants to give them credit for it.

Tim Stellmach: Well yeah, you need to give someone credit, right? *And the right person is going to bring out a lot better work in everyone. God bless Warren, my earlier experiences with him were completely essential to learning everything that I did about how to do this stuff right. But at the same time, there are very few specific things that went into the game that you can point to and say “that was Warren's idea.”*Like Hitchcock without his cameraman, stuff like that.

Matthew Weise: Yeah, exactly.

Laura Baldwin: I think it's a very natural impulse from the point of view of a fan, too, not just media. Okay, you're a crazy fan, perhaps you actually have everyone memorized. I think the average person, you listen to music, and you want to know who wrote the song. You don't necessarily want to learn the names of all the violinists in the orchestra. *You know the composer's name, maybe, and the conductor's name. That's why the conductor gets so much of the credit for an orchestra, even though the musicians are doing a lot of the work. It's good to have a name to fixate your warm-fuzzies on.

Matthew Weise: Oh, no it's totally natural and I'm certainly not against it. I think it's fine, it does make it harder to talk about this stuff in a precise way sometimes. *'m not saying people shouldn't get credit *for creating that environment. It's fine if Ken Levine is the guy who talks about Bioshock or the new game they're doing. I'm always fascinated, for example, on DVD documentaries when Martin Scoresce talking about it. He's always like “Yeah, you know, everyone talks about my editing, but my editor is doing that.”

Tim Stellmach: Somehow no matter how much you deny it, it just comes across as endearingly humble. No one actually ever believes that you mean it.

[laughter]

Matthew Weise: “Oh what a great artist, giving other people credit, isn't that wonderful.” And it's like “I'm giving them credit because they did it and not me.”

[laughter]

Tim Stellmach: “It's cute!”

Matthew Weise: Collaboration is great. But the coherence, the collaboration and the pointed coherence, the fact that it feels so coherent, that it feels like there's a driver in the seat-

Tim Stellmach: *But that person was less making all the decisions. It's not even like writing the Bible, although certainly there are projects that have a bible.

Matthew Weise: And it helped in the case of Thief, it sounds like.

Tim Stellmach: That there was a lot of documentation?

Matthew Weise: That there was a lot of documentation that was not used, or not directly in the game, but just fleshing out the game for the writer.

Tim Stellmach: It's sort of its own skillset. Your metaphor of a conductor and the orchestra, is I think an apt one, because the piece as a whole emerges from the coordination of all these elements, and that *is its own skill. And the way you go about coordinating them is going to color the final result in certain ways. I'm sure that, were I a trained classical music connoisseur, I could tell the same piece conducted by one guy as opposed to another by his conducting style. And yet it's the same piece, and it's being synthesized from all these individual musicians with their own style, but the central intelligence of the piece is still coming through.

Unknown Male Voice: That's something we run into a lot. I just think about the last two summers. *The first summer the problem with the team was, and it was a small team, it was like eight people, the problem with the team was that there was no central vision. So we spent this whole summer talking about how do we have central vision.*And the lesson I learned from this last summer, is that if you get a team together, there's no magic to it you, if you get a team together that's mature enough, they do it themselves, they please each other. You don't have to go to a vision-holder, they become each others vision-holders.

Tim Stellmach: A huge part of that is people are always going to have certain disagreements. And one of the core values that the team has to share is valuing getting something done and resolving problems, rather than valuing the particular decisions that are involved. And that is a matter of maturity.

jtr7 on 28/5/2011 at 12:48

---

Thanks, Pardoner and fett! :cool:

Al_B on 15/6/2011 at 23:43

---

Thanks for the hard work on your part. I've done some basic formatting and it can be (http://www.binarydrift.com/LookingGlass/TimStellmach.html) read here now. I agree with EvilSpirit's take on that sentence so it's been changed and I've corrected a few other items (e.g. Sara being called Sarah on a couple of occasions). Anything else that's wrong then please let me know and I'll change.

Al_B on 16/6/2011 at 06:42

---

Oops. Chalk that one up to lousy memory - apologies, Tim.