Synthesizing Portal 2

Valve's first-person puzzle platformer Portal became an instant classic upon its 2007 release as part of The Orange Box, in large part thanks to the game's uniqueness and charm. That makes developing a successful sequel a tall order -- if your game was a hit because it felt so new, how do you top it when there's a "2" in the title?

The Portal 2 team -- a group several times larger than the one that created the first game -- is addressing that challenge both with concepts that they didn't have the resources to include previously, and with brand new ideas that hopefully broaden the gameplay range without impeding on Portal's crucial elegance and simplicity.

In this interview, Portal and Portal 2 writer Erik Wolpaw discusses how Valve approached the notion of a sequel, how it is building on the game's established mechanics, and how former DigiPen students have once again become an integral part of the team.

How quickly did you start working on Portal 2?

Erik Wolpaw: In some capacity, we started working on it basically right after Portal, but there have been breaks along the way. The team would jump off of Portal 2 to work on Left 4 Dead or Left 4 Dead 2, which are games we shipped subsequent to The Orange Box.

Portal was a very unique game. Coming off of that, did you have particular design goals for Portal 2?

EW: Well one of the things was -- and this was early on -- right after Portal shipped, we started getting these stories from employees of Valve, who played it after The Orange Box shipped. All these stories were like, you know, “I played it with my wife,” or, “I played it with my girlfriend,” or, “I played it with my kids,” and they were sort of playing a "co-op mode" where they would solve the puzzles together. So, we knew pretty early on that we wanted to formalize that and actually make a real co-op mode.

I used to play point-and-click adventure games that way, back in the '90s.

EW: Yeah, yeah. It was a perfect game for that. We figured, “Well, there’s got to be some way to formalize that co-op play.”

That was the big initial idea. We want to introduce new puzzle elements, we want to make it bigger along that axis. Those things are sort of a given; you want to do all these things. But the co-op was the part that was like, “Wow, we’re really going to have to figure out what this means -- what co-op means.”

Did you have to go through a number of iterations of the design even to determine, on a basic level, what co-op in Portal meant?

EW: Yeah. Partly it's just wrapping your head around. It doesn’t hit you until you actually have to design puzzles for it. It’s a very different puzzle design with the two extra puzzles, especially because the co-op puzzles need to require two people to solve.

In other words, if you have a co-op puzzle that one person can do on his own, it doesn’t pass muster as a co-op. It took the designers a little while to really wrap their heads around it. Some things cropped up immediately -- they seem obvious in retrospect -- that hadn't occurred to us before we sat down to play it.

For example, there's the environment tagging. You sit down to try out your first co-op level and tell your partner, “Put a portal here,” and then you spend five minutes trying to explain to the other guy where "here" is. So some of that stuff, little stuff that again seems obvious in retrospect sort of became apparent as we started testing the co-op stuff.

[Ed.: Portal 2 features a mechanic that allows players to temporarily mark the world with glyphs indicating basic actions, allowing simple non-verbal communication between partners.]

The paint surfaces with various properties seem really similar to the IGF game Tag: The Power Of Paint. I assume they're involved somehow?

EW: Yeah, those guys work at Valve. They're on the Portal team. They got hired at Valve and they were working on some Tag stuff. Then, the Portal team was looking at it, and after a while, we thought, “Wow, maybe there‘s some way we can mix these two things together.”

We were already thinking about surfaces. There's an interesting thing in Portal where a lot of it is about interacting with surfaces. But it was kind of binary, in the sense that either you could put a portal on a surface or you couldn’t put a portal on it. And we didn’t want to make Portal more complicated, in the sense that you had more buttons that you had to hit. We like the elegance of just, “I can put portals down.”

At some point, it clicked. We realize, “Oh! If we change surface properties, that lets us do a lot of new things.” So, yeah, it was just kind of a natural fit. Again, it's obvious in retrospect, but it took us a while to come to that insight.

So, unlike the Narbacular Drop team, the Tag guys weren’t hired specifically for that purpose? It just emerged?

EW: Well, they weren’t hired specifically to incorporate Tag into Portal 2, but they were hired to develop Tag in an interesting way. And it turned out that that interesting way was to merge it with Portal 2.

So Portal is almost becoming your college graduate training program franchise.

EW: Yeah, I mean I don’t know if, going in the future, that will hold. It held true for this, at least. Those guys came in and it just made sense that their particular, experimental gameplay actually fit into the Portal gameplay mechanic pretty well.

I'm sure the success of Portal was part of what encouraged Valve to pursue that path again, hiring a student team to develop a concept.

EW: Oh, sure. Yes.

Did it impact the game at all when [Portal designer] Kim Swift left for Airtight?

EW: No. Everybody at Valve gets to work on whatever they want to. A few of the Portal people -- and if you think about it, the Portal kids were pretty young -- had spent a third of their life working on Portal, so a couple of them wanted a break after Portal shipped to try something else. At the time, Kim was helping out on Left 4 Dead. She left, and we miss her, we wish her the best, but it didn’t impact the development of Portal 2 at all.

Are you the primary writer on this game again?

EW: Yeah. I’m writing a lot with Jay Pinkerton, another one of the writers at Valve. We’ve been tag-teaming a bunch of stuff lately on the Team Fortress 2 updates.

I think some people, including myself, feared that a potential Portal sequel might really over-exploit, or browbeat the meme explosion that surrounded the first game. Is that ever a temptation?

EW: If you thought you were sick of the memes, I was sick of it way ahead of you. For instance, cake. I had enough cake jokes, I’m not going to...

The cubes are in there because they’re a gameplay element, and obviously, GLaDOS is back, but there's a bunch of new gameplay and we want to tell an interesting new story. We didn't jettison everything, but I absolutely do not want to try and resurrect a three-year-old meme. That seems like it would be kind of sad. It's not a good idea.

The broad arc of Portal was that you delve deep into the weird underbelly of this facility and it collapses by the end. It looks like the premise of Portal 2 is just the opposite -- you’re starting in this destroyed environment and GLaDOS is rebuilding it.

EW: Right. Basically, you’ve destroyed the environment in Portal and GLaDOS comes back to life through a sequence of events, and starts rebuilding the facility. So that’s the first part of the game, and then it goes off in some other directions from there.

How did you end up with that? Did it just seem sort of an elegant thing to do, mirroring it in that way?

EW: Yeah. It seemed like it would be an interesting way to show Aperture Science. We wanted you to see the effects of what you’d done in the first Portal.

And part of what we always envisioned in Portal, but really didn’t have the manpower to do, was to imagine the facility as this living, breathing place. We always had this idea, in our heads, that the test chambers were modularly assembled and can be reassembled, but we didn’t have the manpower to actually create any of that, and now we do this time.

Having her rebuild the facility before your eyes gave us an opportunity to show off some of that stuff. And then once she rebuilds the facility this time, there’s a lot more. Besides being the voice of the facility, you can really feel that she’s in control, because she can manipulate the environment in real time as you’re doing things.

How much has the team size grown?

EW: Well, Portal was seven or eight people. This one’s probably, right now, 30 people working on it. It’s weird at Valve, though, because you get to poach somebody from another team for two days to give you a particle effect. We have these resources that. They’re not full time team members, but they’re just there, ready to do something.

And the team will expand as we get ready to ship. People jump in at the last minute. As you know from looking at games in development, they don’t really look like the final release until a month before it’s time to go. It's a lot of work, and a lot of polishing gets done at the last minute.

That new technical ability you have in the game seems to interlock with the fact that GLaDOS maybe sees the player as more of a threat than she did in the first game. Is her overall relationship with the player more aggressive, as a result?

EW: Certainly, the relationship has changed since Portal. She does know that you’re dangerous. There’s certain stuff it would be spoilerish to talk about, but yeah, your relationship has changed.

That is a focus that we want to keep. You know, when you talk about the cake or whatever, the one thing that we took away -- or at least I took away -- from Portal, part of the reason it was successful, was because despite being a game it told this intimate story about your relationship with this one other character. Even though we want to introduce a few new characters, we want to keep the core piece of the story as your relationship with GLaDOS, and how that’s changed by what you did in Portal, and then have some place to go as the events of Portal 2 unfold.

Part of what made Portal so good and memorable was simply that it was so unique. That's a trait that's particularly hard to recapture in a sequel. How do you think about that?

EW: Like you said, it’s always a challenge. The best way to deal with it is to power through and just do something that is going to be good. If you set yourself up to think, “It‘s gotta be revolutionary!” you’ll probably just make yourself crazy and never finish it.

We had a bunch of ideas left over from Portal, and we had a bigger team that we could put to work on it, we definitely have a lot of new puzzle elements, and the paint stuff turned out well. Everything’s looking good. You just try and make the best game you can.

I remember Tom Leonard from the Left 4 Dead 2 team saying that he felt so much more confident in designing scenarios and mechanics the second time around, because he'd had so much practice in that world. Do you find that to be the case as well?

EW: Right. For a lack of a better word, there’s a vocabulary that was established to talk about designing puzzle chambers, and we knew a lot of things that didn’t work, so we didn’t have to go down those paths. Having said that, some of the new elements that are pretty transformative, like the co-op and the paint, brought it into brand new territory. So there was a new learning curve, again, that the designers had to go through.