What Went Wrong

1. Kickstarter: Negatives

With $200,000 in funding and nearly 10,000 new fans, public expectations are bound to change. Many people felt that with the extra financing, the game should be bigger and better than previously planned, but we had set a release date that was only five months from the end of the Kickstarter. Nearly every way to expand the project (hiring more help, licensing better technology, and so on) also required more time to get it done, so we had a hard time balancing the scope and time expectations. We feel that we were reasonably successful in that regard; the game was greatly expanded and released only two weeks later than expected, but it was far from a smooth experience.

The extra fans and publicity also meant a lot more public relations work, and since we didn't have a PR firm or marketing manager, that inbox full of email fell entirely on our shoulders. We set up a forum to start building a community space for our new (awesome) fans, which involved additional technical challenges we've never experienced. In retrospect, we should have had someone manage public relations (and our website) for us so that we could more efficiently devote our time to development.

In addition to contracting out help and building a community, we were setting up a company, finding a lawyer, discussing contracts with distributors, and much more. Every day we had to learn how to do things we've never had to do before. We were wholly unprepared for all of it, and relied heavily on the advice and help of our extremely generous friends and family. We set out to make a game and didn't realize we had to learn how to make a business, too.

2. Quantity of Events/Limited Development Time

Since FTL started out as a small experimental project, the original vision was quite limited. We thought we would have a dozen or so basic event types with a dozen different flavor texts for each type. Essentially, we wanted to create something similar to a deck of event cards in a typical board game. Our (perhaps naive) plan was to simply add more text if we needed more variety. How hard could that be?

After the Kickstarter, we decided to expand the game universe by adding a number of alien races, which meant more locations and events. At that point, we had something in the realm of 10,000 words worth of events. By the end of the project, FTL had nearly 20,000 words. That's a lot of text, but we discovered that even this amount would not be enough. When you divide the events between the sectors, even 20,000 words start to spread pretty thin.

One of the most common issues reviewers and players have with FTL is repetition of events. Even with a writer working for six months prior to release, we couldn't create the variety we wanted. While text is perhaps easier to create and integrate into a game than unique animations and art, it's hard to pump out the sheer volume needed to keep it fresh for hundreds of replays. Perhaps our time would have been better spent finding ways to make common events more compelling rather than adding tons of unique events that lose their impact after the first encounter.

3. Multi-OS Launch

Cross-platform development is generally a great plan; we wanted to support alternative operating systems and expand our user base to reach as many interested players as possible. But attempting a simultaneous, three-platform release for your very first game project is not a great plan.

We thought we were prepared -- FTL was planned from day one to be a cross-platform game, so all of the libraries and code would (mostly) easily transfer to other systems, and there was even a development version of FTL for Linux as early as five months into the 18 months of development. We believed, foolishly, that it would be an easy task to finish off the OSX and Linux builds once the Windows build was cleaned up and ready to go. But we underestimated the amount of work "finishing" FTL would entail.

The crunch of release involved the typical 12- to 15-hour workdays to just finish the Windows version, much less the last-minute cross-platform work. Even once we thought we'd succeeded, four days before the launch we discovered the Linux build wouldn't run on a substantial number of the different Linux flavors. Fixing the problems involved intense all-nighters and frantic phone calls to Linux-porting experts. We ended up launching smoothly, but it was hard-earned.

Supporting extra OSes causes problems during the immediate post-release support. We released two patches within a month of release to solve system-specific issues, and each one required making and testing four builds (Windows, OSX, and both 32-bit and 64-bit Linux) before we could release it. As new developers, our build pipeline wasn't ideal, and testing four builds with unique quirks with just two people doesn't speed up the process!

Simply delaying the Linux/OSX releases until two to four weeks after the Windows release would have made things more manageable. With a game under our belts and a build pipeline cleaned up, it wouldn't be as much of a hurdle in the future, but stumbling into it clueless from day one was not ideal.