Warren Spector speaks: part two of our exclusive interview

Whether considering the moral ambiguities involved in slaughtering Greasels, or mocking spacestation romps where you have to save humanity (what could he be talking about?) Warren Spector's enthusiasm for all things Deus Ex hasn't waned - just don't compare him to Miyamoto... You can read part one of our exclusive interview by clicking here.

Have you balanced the difficulty this time, so that different approaches don't offer hugely varying degrees of difficulty?

Spector: I think we did a much better job of that this time round. In the first game, because we built on top of the Unreal engine - you know, Unreal was fundamentally a shooter, a shooter with a very powerful engine, but still a shooter. And the resulting game, if you tried to play it not as a shooter, was way hard [laughs]. Way hard.

So this time what we did, and thank god we had that foundation, was take all the elements that weren't exactly what we needed for our game, and we ripped them out and replaced them with the elements we knew we wanted.

And so the AI is ours, the lighting model is ours; the stealth experience is better, more robust than it was in the first game. Sadly, pulling a virtual trigger is always going to be easier than adopting a stealthier way of playing, it's just a fact.

So if you want the easiest experience, you're still going to stick with the shooting approach. But it's certainly better balanced this time, maybe as good as we could hope to make it.

So the game constantly gives you different ways to approach it?

Spector: Totally; it's funny because I was playing through the game once, and I had an absolutely magic moment, which I won't go into detail about it, but this unbelievable, unpredictable thing happened, as a result of a decision I made in the game.

And so I ran out, I was running around the halls screaming, through the studios, and so I had everybody start sending me their own moments in the game when something magical happened to them.

I have got a folder full of moments from the guys on the team, saying "here's how I played through this bit, this is what I did here," and that's what I'm trying to achieve, you know, a game with options.

So people are playing it in all sorts of ways - it's all about finding a way to play that's personally meaningful to you. A perfect example of that - my wife was playing the game. Frankly I'd be divorced right now if I hadn't left my debug kit at home; I was not gonna take that thing back to the office, she made that abundantly clear.

She was playing through and we had these transgenic creatures called Greasels from the first game, and there's a place called the Greasel Pit that's basically awash with Greasels, and she found herself in the basement where the Greasels were and she couldn't quite bring herself to kill them even though she was supposed to.

So she ended up taking a side quest, interacting with the Greasel leader, and when she entered the leader's apartment, she found that Greasels eat cats. And she loves cats! So from that moment on she became a full on Greasel hunter. Every Greasel had to die! Her identity had to be that person who took out all the Greasels.

So that's an illustration of the way unexpected things happen as a result of your interaction with the game, and it's stuff the designers aren't even expecting. You know, you're the last guy on the space-station who can save the world from alien invasion - who cares!

There needs to be some personal reason for what you're doing, and it's just not there in so many games. Again, we're all about people expressing themselves, and people can find all sorts of ways to express themselves in this game.

How have you made the AI so impressive?

Spector: Okay; about the best answer we have is that it's gruellingly hard work. You do a lot of playing. But the bottom line is that this whole real life stuff is really catching on, as in Knights of the Old Republic for one example.

It seems a lot of people are jumping on the bandwagon, and we've had the advantage where we've been doing this stuff since the late eighties. And so we design games that are specifically designed to allow things to emerge naturally from gameplay running its course.

So we know kind of what to think about, we know what to include; once we've created game systems that allow the characters to try different things, we know enough to predict where problems might come in, and so we plan our AI to cater for the stuff that we know could go wrong.

But fundamentally we just make the game playable - it's no fun at all. You test and you tweak, you test and you tweak, and eventually you lose your mind!

But eventually you get the game to a point where you're supplying almost every player's choices every step of the way. What we're just starting to do now is sort of work on some of our testing processes; we can now track every step that a player takes on every map, we can track every shot, we know exactly how much ammo you have at every moment in the game.

And so when we're testing we can flick a switch that just turns on the metrics and tracking stuff on, which will let us know stuff like "no human being has ever set foot there, on that part of the map. Let's do something that will encourage players to explore that area." Or we could use it to determine that every single player runs out of ammo at this point, so let's add more ammo here.

Are you still excited about the whole Deus Ex universe? Do you think there's still much more to come?

Spector: Let's put it this way, we've already started thinking about the third game in the series... That should tell you about the level of enthusiasm we have for the whole Deus Ex universe. We're not even close to burn out.

In certain circles you're described as the Miyamoto of PC games design - what do you think of that?

Spector: No, that's crazy! Miyamoto, I've never had the privilege of meeting him and I don't even want to because I'd probably start drooling; I respect Miyamoto to the degree where I would gladly have his children; don't even get me started!

I think the difference is those guys, they're artists. The way I see myself, I will pat myself on the back and say I have a very clear vision of what my games should be, and I can't even tell you how passionate I can get about realising those ideas.

So I try to surround myself with similar people who get that vision, who are convinced that that's the future of gaming, who feel like I do. I'm not a Miyamoto, that's crazy; I'm just good at collecting good people.