[In this postmortem, first published in Game Developer's 2011 Career Guide issue, the team of DePaul University students and graduates reflect on the development of the bizarre, yet undeniably creative Octodad.]

In the summer of 2010, a group of DePaul faculty began a project known as DGE 2 (The DePaul Game Experience). The purpose: to create for students the experience of working at a game studio. The goal: to develop a game that could become an IGF Student Showcase winner. Approximately fifty students applied and interviewed for the team. The faculty advisers, game designer Patrick Curry, animator Scott Roberts, and sound man Robert Steel selected twenty students they considered the most qualified to take part in DGE 2.

In June, those of us who were selected gathered in the meeting rooms and the game labs of DePaul's College of Computing and Digital Media. We began a month of pitching, refining, and prototyping new game ideas. The previous year's team, the DGE 1, had taken a spot in the previous year's IGF Student Showcase with its game Devil's Tuning Fork. We were under pressure to find an idea that could match Devil's Tuning Fork's success, while also being completely and utterly unlike it. Finding the right game was essentially a quest for our identity as a team. We ran the gamut of platformers with twists and physics puzzles, and many of them were great ideas, but lacked that certain spark. Then came one silly suggestion about an octopus.

Octodad is a slapstick comedy adventure game about a father who is trying to keep his family from realizing that he is an octopus. It was practically a joke at first, but we grew to love the idea. By the time the first prototype was finished, we could see that certain spark in Octodad, and we knew that it would be the game that would not only define us, but would get us to the IGF Student Showcase. Production on Octodad began in July; it was submitted to the IGF at the end of October.

It was a whirlwind of development, filled with glorious triumphs and tragic missteps. The lessons we learned on the path to making Octodad were invaluable, and we'll keep them close to heart as we head into our future endeavors.

What Went Right

1. Uncanny Team Chemistry

The team composition and dynamic is what made Octodad work. We all began this project as strangers, unsure of what we'd be working on or how we would work together. However, as we began to present our ideas, the atmosphere became more and more lighthearted and bizarre. We came to realize that we were all kindred spirits, united by our collective insanity.

We fed off of each other's strangest, silliest notions, reaching a creative height that none of us could have achieved alone. We kept the bond strong throughout the project by going out to lunch every day, playing our favorite music in the background, holding after-hours Halo matches, and sharing joke after joke after joke. We came to find that near the end of the project, our team had a shared language that outsiders could barely understand, and we briefly wondered whether we'd ever be able to talk normally again. Still, our unchained atmosphere of odd friendship fed directly into Octodad as we were making it, until it became as strange and funny as we were.

2. Generating Concepts

Despite our wacky personalities, our approach for coming up with new game ideas was grounded in a systematic exercise set up by our advisers. We would come up with a series of one-page game ideas, present them, come up with more one-pagers, present those, then team up with somebody to make a more detailed eight-page idea, and finally pick three of those concepts to prototype. We found that first ideas tend not to be the most creative or interesting ones, but that each rejected idea may spark something more. Octodad itself came out of the third round of pitches. We took that approach and ran with it throughout development. Everything from the way Octodad walks, to the kitchen challenges, to Octodad's manly tentacle-moustache was the product of an exhaustive list of possibilities, and seldom did we go with our first idea. Our dedication to exploring every avenue served us well.

3. Playtesting Like Mad

From our earliest prototypes to our final levels, we were always playtesting. We advertised weekly tests through Facebook and word-of-mouth. We had an open door policy, and allowed anybody to sit down and play the game at any time. We were voracious in finding fresh players and new demographics; if there were high-school kids, visiting Japanese students, professional game developers, or notable animators in the building, we would seek them out and invite them to play. Octodad is about riding the line between frustrating and fun, and we found out very early on that our definition of "easy" did not match up with most players. We constantly needed to balance the control scheme, seek out the most fun challenges, and find out whether our jokes were funny. With our rapid-fire testing, we were able to gather and respond to this information on a weekly basis. That's not to say we always did what a particular tester wanted, but regardless, they helped us shape Octodad into a game that others would actually enjoy.

4. Setting our Sights and Sounds

Ensuring that the sound and visual design of our game fit the silliness of the mechanics and narrative improved our game immensely. We had a dedicated sound designer from the beginning, so we were able to avoid the common pitfall of shoving all the sounds in at the end of development. Our sound designer, Seth Parker, went wild making ambient music, collision sounds, hilarious placeholder voices, and octopus noises (often times with his own body).

Furthermore, our team's early attention to sound gave us the time to seek out and work with professional voice actors, and to obtain permission to appropriate Dalmation Rex and the Eigentones' "Octopus I Love You" as our theme song. The sounds, voices, and songs did more than make Octodad extra funny and charming; they also brought a sense of reality to the team whenever we listened to them, bolstering our confidence about the whole project.

From the beginning, we consciously thought of visual design as a way of reinforcing the hilarious experience that we wanted to create. We made a decision early to follow mid-century visual aesthetic and design. During pre-production, we pulled a lot of inspiration from Cartoon Modern, a book all about 1950s cartoons, and the Incredibles art book. For the environments to reach the same level of inspired design we looked to famous modern architects and industrial designers from that era: Mies van der Rohe, Ray and Charles Eames, Dieter Rams, and others. While we developed our own feel, this style was at the core of our design decisions along the way. We chose this because the team felt it would match the lighthearted comical feel of the game. The clean lines and bold colors serve to create an idealized world for Octodad to explore and destroy. Going with this visual style also inspired the art team, because they were required to explore graphic design, typography, animation, color theories, furniture design, illustration, and architecture.

5. Octo-ber Crunch Time

No matter how well Octodad began, what really mattered was how well we finished it. We had a great number of troubles in the middle of development, especially when school came back into the equation in September. Schedules were shattered, priorities were mixed, everything was grinding to a halt, and the November 1st deadline only loomed closer. We made a stand, and declared the final month of development "Octo-Ber." The whole team would meet on Saturdays and work from 8 AM till 6 PM. With about four such meetings possible, we analyzed all of the things we had left to finish, and broke them into weekly chunks. Amazingly, we got more done each Saturday than we had done in weeks previous. The ominous pressure combined with our love of the project forced us to focus, cut, and refine the game in bursts of activity.

We continued working up until the day of the deadline, but we got it all done. When we say that this was a positive aspect of development, we don't mean to glorify crunch. We weren't actually working many more hours than we had been previously. However, we were able to recognize that school had impacted our ability to communicate and work effectively during the week. Making a short-term sacrifice of a weekend day in order to gather at the same time and place allowed us to make more efficient use of our precious time, which had a positive impact on our ability to finish the game.


What Went Wrong

1.The Vertical Slice

Our first attempt at a vertical slice of Octodad was an unmitigated disaster. We were feeling great after some successful pitches and prototypes, and our vision for what Octodad could be was outrageously out of scope. We decided that Octodad's entire house would be the first "level," and set out to make it in a scant few weeks. When the time came to show it to our advisers, they were livid. It was forty minutes long, it was terribly buggy, the art was completely unfinished, there was no work done in the UI, player direction, or camera control, and worst of all, it wasn't the slightest bit funny or fun. In no way did it represent what we wanted our game to be. We had left behind everything that made our prototypes good in an effort to make the game larger and longer, and in doing so, failed to accomplish anything.

After our defeat, we took a good, hard look at the state of things and realized that it would take the entirety of our development to finish Octodad's house alone. And so we did that. Our lesson thoroughly learned, we focused on making a "real" vertical slice out of Octodad's kitchen, and it was the total opposite of our previous effort.

2. Physics

Building a character entirely out of physics objects and having him slam into other physics objects is the crux of our game. And it was a headache to deal with. With so many different kinds of bodies and meshes flailing around, there was an endless stream of physics bugs to sort out: objects not colliding, objects colliding too much, objects jettisoning from the room, and in one strange accident, a room jettisoning from itself. Each bug corrected and each feature added sent ripples across all development, invalidating designs that had worked earlier, or required new designs to be put in place. Any game that tangles with physics is bound to have these troubles, but in our case, we had no game without it, so there was nothing we could do but slog through it all.

3. Failure to Communicate

Whenever our team got together to talk in person, we were excellent communicators, and our meetings were fruitful. Once we were in different places working at different times, our ability to communicate broke down. We collectively had a bad habit of not recording our conversations on or offline, so it was easy to forget what we had talked about without being around to remind each other. Our system for storing information online was also a confusing mess. Each discipline had its own task list pages hidden among many nearly identical superfluous pages in a wiki. Nobody would read other group's pages, as they were difficult to find or possibly an inaccurate duplicate.

Because of these communication issues, we lost a lot of time to team members waiting for other members to complete tasks they were dependent on, not realizing that either the task was never communicated, or that the task had been completed already. This problem was the most ruinous in September, when school resumed and the team spent more days apart than together. Our final crunch snapped us out of it, for the most part, and demonstrated to us just how bad our communication had gotten.

4. Late, Late Bug Tracking

Part of what made our Octo-ber crunch successful was the inclusion of bug tracking software. We suddenly had a dependable way to see who needed what done, and a steady meter for how much work remained on the project. We were foolish to wait so late in the project to set up such a clearly beneficial system. Doubly so because we realized the need for bug tracking at the beginning, but pushed it aside because it took more effort to use compared to talking in-person. The need for one consolidated source of updated information was crucial, and personal communication is less reliable than we thought. We won't be doing that again.

5. Minecraft

No, seriously, it was a problem for the Octodad team. We're fun loving people, and an accessible game with infinite replay value is the death of productivity. For two weeks after the Minecraft Beta was released, it steadily took over more and more of the team's time; sneaking in a minute of Minecrafting could swiftly become an hour lost. Incidentally, this was also around the time of our vertical slice. It saw a lot less play after the slice. While Minecraft itself was only a short-term problem, it was a symptom of something more.

In general, we had issues with getting distracted. During our online meetings, people would have pointless side conversations, or they'd post links to silly videos. In the middle of work days, we'd often play games or stand around watching even more silly videos. The downside of having such a fun, friendly atmosphere is that it's difficult to take things seriously.


A Game With Legs

Octodad was officially submitted to the IGF on November 1st, about four hours prior to the submission deadline. We simultaneously made the game available to the public and asked a few members of the press to check it out. We felt the pride of accomplishment in seeing our development all the way through, and a nervous anticipation as our bizarre brainchild was released into the world. We also held quite a party in celebration.

Prior to the announcement of the IGF finalists, we received a lot of positive attention. In the two weeks after releasing the game, Octodad was mentioned on Kotaku, Joystiq, IndieGames.com, RockPaperShotgun, and many other news outlets. The trailer that we prepared for the launch of the game, which has been viewed 220,000 times as of this writing, made it easy for the media, as well as players, to share our game with others. The game has been downloaded 160,000 times from our web site, with mirrors and torrent files popping up all over the world. We were also able to get feedback from players by keeping an eye on Twitter. Releasing the game for free has allowed us to gain a reputation as a group of talented developers with insane ideas before we even finished school.

Ultimately, we achieved our goal and won a spot in the IGF Student Showcase. We've gone to GDC. Soon enough, we'll have to close the chapter on this first incursion into the cephalopod fatherhood simulator genre. The friends we made, the lessons we learned, the trials we overcame, and the experiences we had will stick with us even as school ends and the team goes their separate ways. In the future we'll look back, shake our heads, and laugh as we remember that strange and silly time when we became the fathers of Octodad.

Author Credits

Brian O'Donnell was the lead programmer on Octodad. He is enrolled in DePaul University's game development masters program. Jake Anderson was the lead designer on Octodad. He recently graduated from DePaul University's game development program. Nick Esparza was the lead artist on Octodad. He is a recent graduate from DePaul University's animation program. John Murphy was the executive producer on Octodad. He is a recent graduate from DePaul University's game development masters program. Kevin Zuhn was the writer and project lead on Octodad. He is a recent graduate from DePaul University's game development program. Kevin, John and some of the other energetic, playful and rebellious Octodad developers have started an independent studio, Young Horses, where they will develop a commercial followup to Octodad as well as other unusually innovative games.

[Originally posted on sister site Game Career Guide.]