It's really varied quite a bit:
- AI War was something that nobody saw publicly, as it was before Arcen officially existed. But it absolutely went through this amount of experimentation, if not even more. It was difficult, but ultimately a really positive process because all the movement was overall forward and resulted in something cohesive.
- Tidalis did not go through this process at all, after the prototyping phase. It was a much simpler game; though the early prototypes were very different.
- Valley 1 went through the same sort of process as AI War, but x10, and really publicly. Unfortunately I wound up really making a lot of mistakes by listening to players a bit too much, which was mainly a problem because players were saying all sorts of contradictory things. Listening to players is super important, but I was really giving up the reigns and lost any sort of central vision. Bad bad bad. It was a very toxic process that I would describe as "flailing about" in the main.
- Valley 2 went through the process to a smaller degree, and it was a more positive process. It didn't result in nearly as fun an experience in the opinion of most people, but I don't think that was an issue with the process.
- Shattered Haven went through a lot of iterations over several years (mostly in 2008, prior to AI War even being developed at all), and its split identity is perhaps one thing people were turned off by somewhat.
- Skyward Collapse had a lot of iteration and changes because it was such an odd duck, and it had a few delays because of that. A lot of player feedback really helped finish shaping that one in a really positive way. The result was something I'm very pleased with.
- Bionic Dues was a very quick project, although we had some drastic prototyping experiments. But after we had something that was fun, it was just something we flew through. Players helped us polish that to a nice sheen by launch, and I'm definitely pleased with the result.
- TLF went through what felt like a million really drastic changes in the combat model. Some of them players never saw, and others are ones that players saw and helpfully shot to pieces (seriously it was very helpful). It was a very stressful process and something that had both its positive and negative points. But overall it was a really positive collaboration between testers and staff.
- If I had to pick a game that SBR is most similar to in terms of the process, it would be AI War, but with the feedback coming from a wider selection of people than just my dad, uncle, and my uncle's friend. It's one cohesive game, unlike TLF (which is basically two games in one), and it's getting closer and closer to being something that is truly fun. AI War spent about 9 months in development, and it was interesting during that whole time, but it wasn't
truly fun until about the last month or month and a half. I feel like SBR is on the cusp of reaching that point itself.
