I'm the first to admit that the Valley 1 process was development hell. It was overly ambitious for our budget, and there were a variety of other problems as well. It wound up losing focus because we didn't have sufficient time for prototyping.
Valley 2 went way over budget, but turned out more or less how I intended, based on the long experience with Valley 1 making Valley 2 easier to envision.
Shattered Haven's art was a disaster for a variety of reasons, and we started running out of time for the extra alternate endings, and it's a niche sort of appeal in general. But beyond that it's what I wanted it to be.
Skyward Collapse was completed very quickly, but under extreme time pressure due to our ongoing budget problems with our growing expenses of larger staff and larger art spend, etc. That made for a less-polished launch than I wanted, which in turn made the 2.0 version of that game (which launched alongside the expansion) a lot more expensive to produce than it should have been. There was a lot of polish and cleanup that was happening there, and the 2.0 version was really the definitive version of what I had hoped for with 1.0. That was actually true of AI War, too, really.
Bionic Dues went pretty much perfectly, from an internal standpoint, as far as projects go. That made the financial performance of it all the more frustrating, because despite the problems with the process in Skyward, it sold so much better. Go figure.
The Last Federation had some scope creep problems and some prototyping problems, but the biggest problem was the freaking combat. That took us back by a couple of months, and really caused incredible amounts of crunch and rush, and the loss of staff, and so on. Despite all that, I feel like we managed to pull it off. There were inevitably still things left to deal with at launch, but it all feels like it is in the same category as the ongoing tweaking that AI War has been getting for 5 years, not the unpolished nature of Skyward.
Have we had a history of trouble with having to rush the end of projects due to budget issues? Yes, absolutely. Pretty much every project except AI War, Tidalis, and Bionic.
Have we had a number of projects that got lost in the weeds due to trying to split across multiple genres? Yes, again, absolutely. That was the biggest problem with Valley 1's development, perhaps with Valley 2's reception, and with TLF's development (although that was not a catastrophic thing in terms of the game itself, it was just very very costly and unfortunate for the company).
Have we had some projects that simply got lost in the weeds design-wise? Yes, definitely, because on top of being in a time crunch, as the lead designer I'd often get sucked into programming and balance minutia rather than focusing on the big picture and doing proper amounts of personal playtesting, etc. I have something like 4000 hours of playtime into AI War, and that's an important thing, I think.
Have we noticed these things? Yes. After each project Keith and I talk about what we think went well and went poorly, and what we think could go better. I didn't want to lose the staff that we did, but in terms of how that affects our monthly expenses and thus the amount of time pressure we get put under, it was a healthy pruning from a project standpoint. We just always -- since the start of the company -- had been in a budget fight for survival, and finally that's not the case. It's our game to lose, essentially, which is a new thing that I'm wary of.
We also recognize the importance of having the lead designer actually playtest and have an appropriate amount of focus on the design, as opposed to getting bogged down in the programming. Much as I love programming, there was a good 3-4 month period during the development of TLF where I did literally none. That was definitely to the benefit of that game. But when budget issues hit, then unfortunately we had to kind of abandon that and that was not to the benefit of the game. With Spectral Empire, we have our ducks more in a row in terms of being able to handle this properly.
The other big thing is making sure that we aren't scheduling our project to basically be feature-complete at the time of release, or only even a month in advance, if we can help it. By having several months of lead time as well, that allows both for some slippage, as well as for an extended period of just kind of quiet stable testing and small fixing that is a very good thing.
This isn't one aspect of our company that we talk about too publicly much, but it is something that Keith and I talk about internally a lot. The others and I talk about production-related stuff based around their areas as well (art, music), but those haven't had the sort of schedule issues and whatnot that core design/programming has.
Anyway, life is a learning process, and that's something we're attuned to. We know we don't have every aspect of our process down pat yet, although we have figured out a great many things over the last 5 years. But each project teaches us new things, both in terms of new things we do well and lessons from things that we did poorly.
The tragedy is a person or a company that never learns and just plows ahead, and I don't think that's something we've ever done. There are some things, like staff cuts, that you could argue that I should have done a long time ago. And it's true that would have been better for the budget pressures the games were under. But I did feel it was necessarily to go one thing at a time before just throwing people under the bus for project reasons. And even in the end, I didn't do it until my back was absolutely to the wall and I had no other choice. So from a business sense that's probably been my biggest failing, although from a personal standpoint it's not something I feel bad about -- looking out for the staff in my charge as much as I can. But with each project, there have been new things that we tried in order to address the various issues. Some experiments were successful and were kept, and others were failures for reasons internally or externally.
What I can tell you is that at this point in time I think we have the institutional knowledge and the staff and budget to do the game that I want to do, and the timeframe needed to have hopefully several extra months at the end beyond what we think we need. We're also not trying to split genres,w which was such a pain point with Valley 1 and TLF.
So, again, it seems to be our game to lose at this point. Will we make some brand new mistakes? Probably. Will we make the same sorts of ones we've made in various projects in the past? Hopefully not, or at least hopefully not in the same ways. So we'll see what happens!
