My series, The Chained Adept, began as an exploration in overall writing technique, that is, the dreaded divide between writers who outline (plotters) and writers who fly by the seat of their pants (pantsers).
I'm an old software engineer and company builder, so (as you might imagine) I'm a natural outliner. Say what you will about software — in the end, if you didn't plan (plot) it right, the program won't run.
Of course, for my first series, I found that as an outlined plot progressed, less and less of the original outline was relevant. In the end, all I was really left with were echoes of the original goals and plot inflection points (the inciting moments, the setbacks, the crises, the resolutions, and so forth). So I thought I might as well start with just that much, or at least the end goal, and try the alternate approach.
The great virtue of writing as a pantser is that, if you don't know how it's going to work out as you go along, then neither do your readers, so you're likely to keep surprising them as you surprise yourself.
You have to trust to your subconscious which has read a lot of books in your genre. It's very good at putting together the clues you've already written (inadvertent or not) and speculating about what might come next. Writing becomes more like reading — you write to see how it's going to come out.
Each time you pause and add a bit to what's been written already, your subconscious adds that to the mix and continues to churn. Every now and then, though, I find I have to give my subconscious a good thump — I've put the coins in the machine, but nothing's coming out.
Right now, I'm headed for a big setback in act 3 of Broken Devices. It's not the final crisis, but it's significant. I've got the villains and at least three other sets of characters all headed for the same general area, with good reasons to be there and serious purposes, and I know what the result will be, but the actual paths that will tie them all together are being a bit… elusive. Like the center of that maze above, you can see the goal but you can't get there down any of the existing paths.
My subconscious is doing one of those whirr-thunk, whirr-thunk moments you get when you turn the key and the car doesn't start. I'm going to be stuck here until something shifts. I need some nice mindless tasks so it can churn away and spit out useful choices that don't depend on implausible coincidences.
I've been here before and I know it'll sort itself out, but they don't call this approach “writing into the dark” for nothing.