Ernest Adams

If we’re not going to use inventory-based puzzles, what kind of alternatives are there? Because inventory puzzles are currently closely linked to adventure games. What could you say that sort of sparks the imagination of the readers?

Well, there’s still of course exploration puzzles, which we’ll still have. There will be puzzles involving talking to people — trying to persuade to tell us something or trying to persuade them to do something. We have some of that now but it’s crudely implemented. Most of them are conversation trees. You steer the conversation in the correct direction and either you get there or else you’ll just try every possibility exhaustively.

I think our biggest opportunity lies in what I would call conceptual challenges, which is to understand something new. We’ve already done this in adventure games. In detective stories, for example. You don’t carry a thing around. What you do is you talk to people and understand a sequence of events by putting together the clues.

I remember a puzzle from Escape From Monkey Island which I found really refreshing. You had two parrots, one was lying and one wasn’t. You had to figure out which one was which.

Right! (laughs)

That’s something that happens more in your mind.

Yeah, that’s a conceptual sort of puzzle.

How significant will AI be for making those dialog-based puzzles more advanced?

We’re a long way from that. I think AI will be significant from the standpoint of trying to simulate NPCs who have emotions and react to you, and the situation, in a meaningful way. Even more than the static NPCs who are always in one place, [I see potential in] having a sidekick character who goes with you. As we’ve seen in games going all the way back to Planetfall or other kinds of games today where you have a person who goes with you and offers advice.

Obviously generating natural language is a bigger problem. Understanding natural language is even worse. But I think there’s a lot of opportunity for having characters whose emotions, reactions and behavior you can come to understand — and I’m sorry to say manipulate (laughs) — in order to achieve the results you want to achieve.

Could AI be used to act out something instead of telling the player what it is? Instead of exposing the story through text, you might actually show some of that through the behaviors.

Yeah. The behavior of people can be used to illustrate particular things. There are some people who simply don’t want to help you and you have to find out what is necessary to get their attention. Nowadays that’s still a lock-and-key thing. You know, “tell me about this” and then all of a sudden they open up and start telling everything. It’s not subtle. It’s actually binary — just zero or one. You have the right key or you don’t. I think more subtle interactions with artificial characters holds a lot of potential.

In the lecture before this interview you made a comment about how in Sonic it appears that the player is Sonic, but after some inactivity the avatar on the screen will actually wave at the player. I thought that was really interesting because it really applies to adventure games. There is a character on the screen who sometimes has his own will, you as a player are there, and some games also have a narrator voice. Sometimes you can even hear the designer speak through the character, when it says “I can’t go there yet”.

I observed this peculiar relationship between the character and the avatar a long time ago because there was a transition at the very beginning. Games like the original adventure Colossal Cave and Deadline, and some of those early Infocom games, assumed that it was really you and you had no personality. Also in Myst, you know, if you looked in the mirror there was nothing to see because it didn’t know anything about you. The difficulty with that is that the world does not know how to react to you because it doesn’t know who you are. Once you start getting an avatar to represent you then that person has a character, an appearance, an age and sex, and all kinds of other things. And so the world reacts to them in a way we recognize. That introduces a sort of peculiar ambiguity of are you “them” or are you “with them”.

It is a very conflicting thing. It goes from one perspective to another.

But it doesn’t disturb the player, I think. Players have learned to deal with this convention.

Like suspension of disbelief in a movie — they lock it out?

Yeah. In the same way that they understand that if you are shot in an action game, you don’t slow down. Injuries hurt you but they don’t actually slow you down. That’s just a convention that we accept because it’s needed to make the game balanced. I think this odd relationship between the player and the avatar is again something that we have come to accept. In a way we’re moving even farther towards the idea that the avatar is simply a separate person who generally takes your advice. (laughs) The girl in The Longest Journey is not me. She really is her, and I’m a separate individual who’s kind of interested in what she’s doing and what she has to say. When I say: “go over here and pick this up” she’s willing to do it unless it’s impossible or it will kill her or something like that. Then she says “no, I won’t do that” and I’m happy with that.

There was an article on Gamasutra published last week by Randy Littlejohn about drama and games. Do you think that’s a direction adventure games should pursue?

I think someone should investigate it. I don’t feel as if at this point we should close off any avenue and say “no, don’t do that.” I think what we should do is explore all of our new options. Some of them will fail, but that’s okay. It’s important to find out what doesn’t work by trying it, as opposed to saying we’re not going to try that.

I guess my last question would be: do you see adventure games as a core genre that is still significant in the industry, or do you think action-adventures have taken up that place? I see a lot of examples from adventure games at this conference. I also just leafed through your new book and you dedicate a whole chapter on adventure games (like many other game design books have). So it’s still significant here, but will it be in the industry?

It’s not very financially significant right now. I think they will come back but they may not be in the form that we’re used to. I think that whatever they do come back as, it may not be Monkey Island or Indiana Jones. Indiana Jones changed from being a story adventure to an action-adventure really, once it went over to 3D. But yeah, I think they will be back in some form.

I personally happen to think that adventure games are the only truly new computer-generated form of entertainment. Before there were computer flying simulators there were mechanical flight simulators, and mechanical racing games. I remember mechanical soccer and sports games. Before RPGs there was Dungeons & Dragons on paper and pencil. Adventure games are one of the very few truly new forms of interactive entertainment. I think they’re very, very important.