Some thoughts about process and SeptNoWriMo

So, I have this idea that maybe I'm subconsciously discarding processes that work. Not sure why: some kind of weird self-sabotage, most likely. To try to fight back against this, I'm trying to figure out what works and what doesn't work, and I'm starting by looking at my work over the past few years. I've tried a few different processes with varying results.

* I've written five 1000-word stories (Death-stream, Drunk/Goat (untitled), The Goat Man, Alone Together, and Haunting Perfections of the Lightless Void).
* I've written one 9k short story (Swanflight/The Last Dance Before Midnight).
* I've written 40k of a novel or series (The Nightmare Box).

Of those, the most successful would probably be either SSSFFSS or the flash fiction. Or, most worked on would be TNB. Which is the deciding factor: most complete, or most effort? If I go with most effort, that makes me feel more like my spreadsheets are some kind of use, so I'll probably try that for now. So...

2017: TNB is most worked on, TLDBM (SSSFFSS) is most complete.
2016: LOFB is most worked on, TWEOTR (SSSFFSS) is most complete.
2015: Placeholder is most complete. Looks like most worked on is Fire and Ice.
2014: Most complete will be Children of Light (SSSFFSS). Most worked on is either The Artist or Necroverse in general.
2013: Most worked on and most complete is Seeker: second drafts of Doom, Despair, and Desolation.
2012: Most worked on would probably be Doom. Most complete is probably the second draft of Darkness.
2011: Most worked on will be Dispersion. There's no clear contender for most complete as all 12 drafts that year were completed first drafts.

Next question: of the projects listed, which are most indicative of the process I was following that year? If I go by that sort of criteria, it's going to be like this: Dispersion, Doom, probably Doom again, The Artist (although with reference to the interference of the wider Necroverse), Placeholder, LOFB, TNB. Although, if I break out of the thing with years slightly...

Alternative: can I identify a generic set of processes as suggested by these stories?

1. Deadline and wordcount goal, linear pantsed. This is what I was doing in 2010, 2011, and 2012, with 15 first drafts completed in that time.
2. Deadline and wordcount goal, non-linear pantsed. I switched to this in 2013 and although I can reach the wordcount goal within the time, I don't seem to be able to make a coherent story out of it.
3. Wordcount goal but no deadline, linear pantsed. I've tried this off and on from 2013 to the present and it doesn't work, because when there's no reason to keep poking at it I tend to quit at the first major hurdle.
4. No wordcount goal, no deadline, non-linear pantsed. This is the 'take out everything that works and wonder why I'm not getting anywhere' one, which I tend to interweave with #2 in the hope of fixing it.

That's four variations. Looking at that list, I reliably complete drafts when I have both deadline and goal and I write straight through from start to finish. I've finished a couple of shorter things by dropping the linear part, but it doesn't seem to work for longer stuff. And, there really isn't much point starting a project with no constraints. So, based on my own historical data, this is probably what my optimum process looks like :

1. I need a clear idea of the format/wordcount required. E.g. 'novel', 'short story', '50k'.
2. I need a work structure somewhat like NaNoWriMo where I have a daily wordcount goal and a fixed start/finish date. It doesn't have to be the length of NaNo, but needs a certain level of "you can't back out" built into it.
3. I can write random scenes occasionally, but I should mostly focus on producing a linear narrative.

So, I'm thinking of getting back to something like this, which possibly means I need to declare next month a somewhat official SeptNoWriMo.