The Dream Machine is a fantastically-original indie point-’n’-click adventure game made with real life objects, featuring a dark and trippy story that involves exploring dreams. The final piece of the episodic series, Chapter 6, is currently in development by Anders Gustafsson and Erik Zaring, who together constitute Cockroach Inc. I recently got a chance to talk with Anders about the game’s development. Check out the (more or less) spoiler-free interview below!

A conversation in The Dream Machine.

Jeremy: First of all, I just wanted to say thank you for speaking with me. Chapter 6 of The Dream Machine is one of my most highly anticipated games and I am very excited to learn more about your development process.

Anders: No problem. Thanks for contacting us!

J: My wife and I both love The Dream Machine even though I’m a very traditional gamer and she is a much more casual/non-traditional gamer. Was this broad appeal a planned part of the game’s development?

A: Not as far as the themes or the story in the game, but we certainly wanted a low barrier of entry to the game.

We basically set out to create a game we’d want to play ourselves. We wanted a story that would appeal to more people than pre-pubescent boys. Not a gun-based power-fantasy. Instead something an adult might be able to feel engaged in. Something familiar with hints of twisted darkness in it.


Generally speaking, we want gaming to be a democratic pleasure, accessible to as many as possible. Not just for people with twitchy reflexes and big pocket books. There’s this ongoing race to get people to reinvest in gaming hardware. I understand why that’s done, but we don’t want to be part of it.

Back when we started, plastic peripherals were all the rage. Guitar Hero and such. Not only did you have to own an expensive console, you had to buy expensive extra equipment for it as well before you could even play the game. We wanted to go the opposite direction. If you own a computer that can reliably browse the Internet, you can probably play The Dream Machine on it.


We also wanted to remove time limits entirely. The player can take their own sweet time, without harassment from the game. And we wanted to map all player input to one mouse button, so the player never had to think about which button does what.

And accessibility has always been important to me, for some reason. Games are getting better and better at catering to for example color blind people, but there’s still a lot to do. We have some accessibility features, but we could do better there for sure.


J: It seems like each chapter has been larger than the last and expanded the scope of the game. Can we expect this trend to continue with Chapter 6?

A: That’s true. Our ambitions for each chapter have grown the longer we’ve worked on the game. I think that’s true for any long term project.


It’s certainly true for Chapter 6. We originally didn’t want Chapter 6 to be very big, because it’s the climax of the game. And it’s hard to sustain a climix for very long without loosing its emotional impact.

So we imagined Chapter 6 as a pretty lean sequence of trippy vignettes, but the more we worked on it we realised that trippy is very hard to achieve if you’re trying to cap off an already trippy experience.


I think Chapter 6 might be the largest chapter if you look at the environment tally, but it doesn’t have as many puzzles in it as Chapter 5. The game is a huge six stranded braid, where each chapter is a strand that eventually ties into Chapter 6.

It’s one big riff on everything that’s come before.

An environment from Chapter 5 of The Dream Machine.

J: The characters in The Dream Machine are full of personality and it really feels like each one has their own unique life and history. Did you base any of the characters on real life individuals?

A: In some cases, yes. Off the top of my head:

The design for the old mother in Edie’s dream comes from a portrait of a relative that hung on my grandparents living room wall: A very stern judgmental old woman that was always looking down at me disapprovingly whenever I was in there.


Erik based the design of Selma’s grandfather off Jean-Luc Picard from Star Trek: The Next Generation.

There’s more examples, but those are the ones that come to mind right now.

J: What different aspects of development do you and Erik handle? Do you each have defined roles or do you both work on all aspects of the game?


A: The game is a love child from my and Erik’s womb. We share one womb between us, that we each get to have every other week. The game has been gestating in this womb now for almost eight years and it’s soon time for our beautiful mutual baby to dislodge and break free.

As far as roles goes, we have very different roles. Erik found the money to kick this project off. He was also the instigator, caretaker and visionary of what is the look for the game. He builds, paints, lights and photographs each set or character in the game.


And I do the rest. That’s a very reductive way of explaining it, but it’s also quite true.

J: At what point during development was it decided that real world objects would be used in place of traditional computer generated graphics? Were there any unexpected challenges that came with this choice?


A: We decided to handcraft the graphics very early on. I had been kicking around a prototype for the game, when Erik suggested building everything by hand. Naturally I resisted, thinking it would take way too long time. But when he sent me some primitive test environments it was love at first sight. He’d been up all night putting five sets together and they were just the greatest. He’s made them so fast the paint hadn’t dried yet, so there were fingerprints all over the, but that only made them look better.

There was no turning back after that.

An environment from The Dream Machine featuring some lovely popsicle-stick chairs.

J: How do you go about designing the puzzles for the game? Does it start with the actual logic of the puzzle or does the puzzle come out of the story?

A: In the best case scenario, the puzzle comes from the set and the story. In what situation is the character at the moment and how does he or she get out of it? If you start that way, the puzzle and the story will push the player along in the same direction. And when the player encounters the puzzle, it will hopefully feel like a natural obstacle, rather than something put in his or her way to artificially pad playing time.


The few time I’ve tried to ”pre-generate” puzzles and then retrofit them into the game, the results have been a bit hit and miss.

J: The Dream Machine has been released using an episodic/early access model. What have you learned from this and what advice might you give other developers looking to release a game using such a model?


A: For the release of this game we’ve utilised a combination of both the episodic model and the model today referred to as the early access model. That name hadn’t yet been established when we released Chapter 1 back in 2010, but the similarities are striking.

We chose to go episodic because we didn’t have any other choice, as I think is the case with most episodic games. It’s certainly true for us. We wanted to make a very ambitious game and we didn’t have the money to do it. Given that we didn’t want to scale down our ambitions, it left us with only one option — to cut the game up into chapters and start selling them early rather than at the end. It’s not the ideal way to release a game with a linear continuous story, but we simply didn’t have the money to develop it as one big chunk up front.


It worked out well for us, but I’m not sure I’d recommend it. Gambling your future reputation on whether or not you manage to complete the game, is pretty daunting.

It’s worked for us. We’re just about to make due on the promise we made 6 years ago: to release the game we started all those years ago. Without scaling back the scope. Without compromising the quality.


It’s been a long crazy ride.

J: The story of The Dream Machine reminds me a lot of H.P. Lovecraft’s Dream Cycle stories. Was this an inspiration for the story? What were some other inspirations for the game?


A: I’m aware of Lovecraft’s Dreamland concept, but I found that particular notion through John C. Lilly.

While at animation school, I read a lot about John C. Lily and his experiments with LSD and Ketamin. He had this wonderful notion that he was visiting an alternate reality during his drug-induced hallucinations; a place he thought had a coherent geography. Once he regained consciousness, he would draw maps of what he had experienced – noting down things like landmarks, geographical features and coastlines. He thought that if he had enough map pieces and then spliced them together, he would be able to slowly chart this new reality.


To me, that idea just sounded so naïve and beautiful. In our game we approach dreams in a similar way.

J: Are there any plans for what you might do with all the models you’ve built once development is completed?


A: We’ve already started exhibiting them at art galleries. So far the game has toured Los Angeles, Las Vegas, New York, New Jersey and Trollhättan. We have a few more exhibitions coming up this year and the next.

After that, we might start auctioning them off if there’s interest. They take up quite a lot of room and Erik has a tight basement. No pun intended.

Clay character models from The Dream Machine.

J: I know it may be too early to ask this, but are there any plans for Cockroach Inc.’s next project after the conclusion of The Dream Machine?


A: For sure! We’re developing a game about an old lady riding the subway. It’ll be a very different game from The Dream Machine, but after 8 years in development I think we deserve to do something different.

You can keep up with the development of The Dream Machine on the official website, on Twitter, or on its Steam store page.

Share This Story

Get our newsletter