I have thought about a bunch of things, read through the basic rules of another game, D100 Dungeon (hat tip to Andy J for the recommendation), and written a few more bits into my working document, but not made a huge amount of progress in the last month. I'll use this post to discuss some of the decisions I have made so far, and muse about a few of the other bits that are up in the air...
One of the key things I have noted from the games I have looked at in the solo dungeon crawl space is that they are almost exclusively pretty highly "statted" games with very much a D&D vibe to them with skills, attributes, equipment lists, classes, levels, and so on. This is not a bad thing - dungeon crawling as a genre feels to me like it should be a tactical challenge as much as anything, so rules for a game of this type will, of course, support this play style.
Let's not do that.
Stepping back a little, one of my favourite roleplaying systems is in Over The Edge by Jonathan Tweet (I'm basing this off the 1st edition). I admit that, while I really enjoyed reading the book, I have never played the actual game in its original setting. I have, however, used its core system in other settings (it worked well for me for a 7th Sea one-shot, for instance) and love its simplicity and flexibility. The basic concept is that you have four traits: one is a central trait (often a profession, trade, or similar), two are side traits (hobbies, interests, side-hustles, etc), and one is a flaw. The positive traits give you dice to roll to achieve things that fit within the scope of that trait (fuzzy definitions to be decided by the games master) and the flaw causes dice to be deducted from your rolls when it applies.
I'm not proposing to do this myself, but using this as partial inspiration, I think character creation will effectively be to simply define some broadly applicable strengths. I may add in some sort of flaw later if it seems appropriate. The OTE system relies on agreement between the game moderator (as OTE expands GM) and the player on the interpretation of whether a trait or flaw is applicable at any given time. That won't work in a solo game without either heaps of lookup tables or challenges in the game being defined with a set of applicable traits. I'm reckoning on focusing on the journaling aspect of my game, so the important thing is more to provide interesting writing prompts rather than a tactical game, so it'll be mostly left to the player to decide. If you want to cheat, fine, this sort of game isn't really about winning or losing, it's more about having an engaging experience and creating an interesting story.
The idea is that you play a party of two characters: a renowned hero with mighty skills, and you, their sidekick, who can help out from time to time, and who will record the tale of the adventure for posterity.
How does this all fit together to actually make a game? To be honest, I'm not entirely sure yet, but what I am working on as a core mechanism is that an encounter in the dungeon has a description, including a suggestion of what sort of challenge it is...
"You enter a cavern with a pool of water in the middle and two other exits. There are a whole buttload of goblins, counting their saucepans. There is a boss goblin who looks angry. This could turn violent."
I mean, not Shakespeare (or even Gygax), and I need a way to generate that, but this is a problem for future me, who I'm sure can handle it.
As the player, you decide how you will use the skills and traits (as well as equipment, magic items, etc.) to deal with the situation. I currently think that each of the things you can bring into play may have a number of times you can use them (boxes to tick off), and each thing you use adds +1 to a die roll to resolve the encounter. The modified roll is checked against a results table (with other factors being taken into account) and you get an outcome from that which indicates how successful your efforts have been, so then it's down to you to turn that into some narrative in your journal. If things don't go perfectly, you are likely to lose some luck or health (two scores tracked on your character sheet), and maybe be forced to retreat or something.
I'm currently in the process of writing up the character generation element, which effectively takes place at the tavern in the village while getting the call to action. Then the idea is to have a "travel to the dungeon" section, where there can be some sort of encounter that could set the tone for the rest of the game. Finding a way to link encounters together thematically will be an interesting challenge; to a large extent this can be left to the player's narrative, but there should be some help or encouragement for this from the game mechanisms, maybe something like gaining a checkbox somewhere if you make a link between certain things.
One step at a time though. My aim for the next few weeks is to get the character generation and travel sections into a first draft form, so it can be played to that point. Oh, and also work on some other projects along the way.