Last week I committed to talking about progress on level 4, regardless of how much progress there was to talk about. The good news is that there's plenty! Level 4, the Factory level, is very nearly done. All told this level was actually the quickest to build so far, which was very surprising, as in the initial draft I'd labelled it as potentially one of the most ambitious. I started less than three weeks ago, and now with a third of the month to go it's almost ready to move to the polish and playtesting stage.
For each major gameplay sequence I've been trying to experiment with different methods of construction. I'm interested in the idea of "method acting" for game design, trying to help bring about moods in gameplay by experiencing those moods as a designer. Level 1 was intended to be gloomy and pessimistic, so while designing it I listened to Adorno's "Minima Moralia", an extremely depressing book. In level 2, which is themed around drugs, I did most of the initial level drafting while very stoned, then went back to clean it up after sobering up. Then for level 3, which is a hyperactive, energetic experience, I locked myself in my room with a bunch of candy and energy drinks and tried to blast through the whole thing across a few manic creative spurts.
Well, none of these methods were especially conducive to constructive work, which is maybe a sign that game developers should not try "method acting". But since this level is themed around mechanical labor, it was in keeping with my policy to construct it in a methodical, architectural way. I drew out a draft of the map on graph paper before starting to construct it in-engine, and took notes beforehand on the precise function of the gimmicks and whatever else I'd need to keep in mind for the upcoming build. It wound up looking like this:
It shouldn't surprise you to hear that this was a much more efficient method of design than "get stoned and vomit geometric patterns onto a tilemap". Indeed, I'm a little worried the process has been too smooth. Previous levels have taken much more drafting and iteration to even reach the point of functionality, and I think this gives them a certain lived-in quality, an accumulation of small details added over the course of weeks. The Factory currently has fewer of these interesting details, simply because I haven't spent enough tedious afternoons in the debugger to daydream them up. I also suspect it'll make the level easier to complete, since the overall design is simpler, but this isn't a problem; it was already intended to be on the easier side.
Another thing that made this design process relatively smooth was that, to some extent, this level had already been drafted! My previous game, Remote Perception, has the same theme and intended tone as the first half of the Factory. So, for the introductory gimmicks -- conveyors, player-platforms, looping spike tunnels -- I could simply copy from what that game had already done. Plus, the switches and teleporters are featured in the nearby overworld (see devlog #2), so those didn't need much additional work to function here as well. That's not to say I could rest on my laurels: everything in the second half of the level, when the power cuts out and the map fills up with robots, had to be constructed from scratch.
It's this second half that had initially led me to suspect the Factory could be over-scoped. To make the map function, every room would need to serve a dual purpose: first as a rhythm challenge in the first half, and then as a stealthy low-lighting puzzle sequence afterward. I'd need to plan around the possibility that the power might cut out in any room, at any point in level progression. Lastly, I'd need to worry about lighting for each room, or so I'd thought. In fact, I realized while testing that, without a timer to worry about, the rooms were still perfectly navigable even in near-total darkness. As long as there was a small light around the player, plus lights on the trawling killer robots, there was little need for me to apply any more environment lights.
What's more, since the extreme darkness created its own interesting emergent puzzles in each new room, it took surprisingly little revision to keep things fresh. The robots themselves are intended as general-purpose hazards, so they could be added arbitrarily without breaking the flow of existing rooms or needing to rejigger the layout. They're a little janky, but not in a bad way: like in a stealth game, it's sometimes possible to 'get away with' getting caught by them, but it's highly risky and you're much more likely to succeed by staying out of their line-of-sight.
There are still a few issues. I haven't finished building the Guidance Spectre's interventions yet, and there are a couple potential softlocks if the power dies when you're in the middle of a switch puzzle. Lastly, I still need to hook up the level to the overall game flow: updating overworlds and saved global stats on completion, so it fits into the full experience without issue. Once these things are done, I can start bringing in testers to try it out. I'd also like to give the level a unique piece of music, but this is lower-priority; I don't mind saving it until after the testing rounds are complete.
It's interesting for me to contrast this against the previous level, which is much more aggressive and intense. In that one, my focus was entirely on building a linear set of challenges to match a certain difficulty threshold, and then blasting the screen with visual effects to up the difficulty even more. In this case, the visuals are simpler, the level is much less linear, and - even though there are more than twice as many rooms to get through - each room is under much less pressure to be uniquely difficult. I suspect many players will prefer this level, for these reasons; though I'm personally happy with both, the layout of the game only necessitates completing one of them, and I suspect that this will be the easier pick for most of them. Until next time.