Not rebooting again, just an update this time :)
I'm writing this quick little thing as a reminder to myself: when you're in the middle of a game, keep the stuff that you are finding easy to create, the stuff that is flowing well, and drop the stuff that is hard. Even if it conflicts with the vision you have for the project.
Let the game live its own life, let the practicalities of "this just works" override your own ideas "this will happen"-- let the universe dictate to you what your game will actually be.
I think this principle can be applied to other art, but it's more essential to game design than other art forms because a game is more like a scientific experiment: you put some things together, they interact with each other and the player(s), but until they exist physically you can't really know what they will be like. In the same way that it would be unscientific to shove an experiment in one direction or another to get a desired outcome, it's unscientific to shove a game design in one direction or another to match your own intent as a designer.
I think good game designers probably know this intuitively, but advice like "build prototypes" or "iterate" actually don't go far enough-- if you are still hamstrung to your own vision, you won't easily see the part of your prototype that was contrary to the whole thing you were going for, much less will you be willing to run with it.
I think part of the reason that Paradise Never has taken such a winding path is that I have worked too efficiently on it. Going back to what I said about experimentation, I have actually don't a very good job of bending the parts of the game that don't work until they almost work, and things have always seemed to progress well— but actually this is irrelavent, because nature does not care how much you push in one direction, if you desire an impossible outcome, you'll simply never reach it.
There's a difference between "flow" and "progress". Always aim for "flow" and treat "progress" as a sign you may need to double-back on your path.
I'm building the main content of Paradise Never now, in cycles. Each cycyle contains several small parts, which can be independant to each other or relate to other existing parts.
Each cycle consists of three "background stories" and one "NPC story".
The background stories consists each of one large, medium, and small, and add some interesting context to the world. An example would be ruins, some detail for buildings, or an item.
The NPC story will consist of adding an NPC, or adding more context/meaningful interaction to an existing NPC.
The latest cycle I had, I over-reached and that's the cause of this musing. Basically, the "large" story was too large and didn't really flow. I forced it into existence, but I don't think it's really good so I'm going to drop it. The work felt hard. I don't feel like going into the gory details here but maybe I can explain at some point in the future when the game is out. It's a definite design decision to drop this particular bit of work.
I think there is a problem where a person can over-polish a game too early. But there is a corollary to that, which is that a person can leave annoyance in place too long.
Try to see the points of friction and don't leave them standing, address them early. Until you do, you can't see other opportunities for deeper interactions; you're unconciously stumbling over the same thing repeatedly.
For Paradise Never, after the last ill-fated cycle, I spent the better part of a month or two just addressing some things like this, in particular some visual stuff that was functionally not good (creating too much visual noise so that it was hard to parse what was going on) and some bugs relating to entering/exiting buildings.
Keeping in mind "what is working", I will continuing to work in cycles, each about 2 weeks. In the interim I am updating RAPALA and Venture the Void, to try an keep myself going too deeply on Paradise Never. These days I work only in my spare time, so in some ways it's slow going but it's enjoyable to me.