Solo Rpg Players: How often do you engage randomness for story needs?

edited September 2013 in Play Advice
It seems to me that for one to engage in solo roleplaying some sort of random input is needed to direct one's imagination. Without it, the exercise becomes plain creative writing, or storytelling.

Still, engaging a randomizer results in a break in the flow of imagination as you interpret the results. The break in the flow might be short, but it is there. I've found that trying to keep a cohesive feeling to the story sometimes means that the breaks are longer as I try to make sense of some random result in context of what has happened in the story prior to that.

What is your sweet spot in terms of percentages? Do you have any rules of thumb as to when to call upon Randomness?

I have created a little poll to see if I can get a feeling of how often folks engage their randomizers:



  • The 'break' is a tricky thing. I love the randomness since it offers a challenge to me to make it all fit. So, it's like a mini-game in that respect. In the style that I play, I begin a new scene with a roll of some Rory's Story Cubes and throughout the dialogue I'll reference the Mythic GME tables. But, while I randomize a lot, I minimize that break by making it very short, story cubes give you an instant result in the form of icons on the dice and I write my solo story/game log inside of an Excel spreadsheet that has the GME tables already built-in. I get results in a few clicks without having to hit the books. So, I spend the vast major of the time in descriptions and dialog even though I might use randomness a half dozen times in a scene.

    For those that are curious about that style of solo story telling...
  • @TAZ:

    If you had to come up with a very rough ratio for how many times you use random input per plot points + NPC actions + color details, what would that ratio be?
  • My preferred way to drive solo play is setting constraints on the future narrative. Random inputs come a close second, but the way I see it a random player is less fun than a player who is me, but earlier or later in time.

    As you may guess, the way you get around the Czege principle is to have the game effectively produce another player from its rules, your decisions, or random inputs. In all those cases, if that effective player is sharing the load of presenting and resolving conflicts - your play will be better for it.

    - Mendel
  • Hi @wyrmwood,

    That sounds like something I might actually take to. Can you expand a bit on what those constraints on narrative are and how you use them (as in, how does your mental process work specifically) ?

  • Alex,

    One simple example mechanic for imposing constraints is a list of resolution options (complete disaster, perfect success, complicated victory, etc.)- one of the options is to describe a new resolution option and add it to the list. The constraint comes from needing to use all the options on the list before you can re-use any of them. Combine this with a similar list of new situation prompts (preferably starting with a different number of options), and you are frequently faced with difficult decisions based on your past actions, without any need for randomness.

    - Mendel
  • If you had to come up with a very rough ratio for how many times you use random input per plot points + NPC actions + color details, what would that ratio be?
    [cross posted]

    Wow, looking over Yuki's tale, it's a really high percentage. Each roll generates a paragraph of scene, action, or dialog. It's not 1 to 1 since sometimes I'll get 'wordy' before getting back to the dice. Excel says it is 70 or so rolls and 111 paragraphs for chapter one. Given that most of those long paragraphs are extended dialog, I'll stand by 90+ as a percentage. It's not as random and chaotic as it might sound. It you read the text without the looking at the 'game track', it would appear pretty seamless.

  • Thanks @wyrmwood. That's actually kind of cool and I will add it to my toolbox. :)


    Thanks man! I find that with the Mythic style of yes/no question, things do zip by a little faster. A lot of the content is in the question itself anyway, at least in how I do it. I imagine Excel could speed me up more. :)

  • My preferred way to drive solo play is setting constraints on the future narrative. Random inputs come a close second, but the way I see it a random player is less fun than a player who is me, but earlier or later in time.
    I follow these solitaire threads assiduously as I'm constantly trying to find the right technique for my needs. I have a curse, known colloquially as "no imagination", so my approach is almost the exact opposite - I hunt for ways to synthesize the input of a player who is not me, generating seeds with a little more direction than zircher's story cubes or the GME generally provide.

    Historically I've generated a handful of randomized results at the start of each scene and tried to find a way to tie them together, but maybe I need to spam it up more and see how that works. The hunt continues...
  • @veav

    That is sort of how I've been experimenting on my latest attempt at a solo session. I front load details relating to character, setting and theme, and then continually reference those as guidelines for PC action, dialogue choices, and color so that I don't have to rely on randomness for those.

    What I've been trying to do is rely even less on the randomizers but I am still fine tuning the right balance for me in terms of what ideas I can take away from the randomizers without completely making it an exercise in straight creative writing. :)
  • Would still seem an exercise in creative writing. The 'interpretations' are just more creative writing.
  • Yes and no, most solo gamers don't talk to themselves, so there tends to be more writing, whether that is full dialog or journal entries is a matter of taste. The thing that draws in a lot of people to gaming is the interaction with the players and GM and the surprise when they do something unexpected. Randomization does that to some extent, the key is to not make everything a coin flip. For example, the Mythic GM Emulator has a graded scale from exceptional no to exceptional yes and these values are based on a sliding scale of likelihood and the chaos factor. The net result is that the logical thing usually happens, but every now and then the system throws you a curve ball.
  • A lot of early Matrix games were run as solo story games. The system used the idea of conflict between two sides. Each turn the player would make up two short stories about what happens next. One argument is the best outcome one side (the good guys) and the other what the other side would want from the story. Next the player steps back from partisanship and decides what side's argument is stronger and then assigns numbers to a roll to see which action happens. Roll to see which becomes reality.

    The story builds up one conflict at a time.

    This works well but doesn't include a random element. A player could roll randomness and then use that as the starting point of each turn. It does make all moves meaningful, which is challenging when very much randomness is used.

    Solo games and creative writing are pretty closely related.
  • Speaking of creative writing's link with solo gaming, its worth taking a look at Oulipo for further inspirations on random and non-random creative constraints on story and narrative. One of their goals was to make the creation of fiction more playful - which is nicely adjacent to making the playing of fiction.

    - Mendel
  • edited September 2013
    @Callan_S: Yeah, agreed. However, I don't want it to be an exercise in "straight", or maybe a better word is "plain" creative writing, in the sense of just sitting down and writing what comes to mind and basically leaving it at that. :)

    @wyrmwood: That link is very interesting. Food for thought, definitely.

    @zircher: I was crazy enough to try to do away with writing altogether. I was using a voice recorder, thinking that it would speed me up, but I still had the same sort of blocks, and indecision. I ended up with a lot of "Mmms" as I thought, which resulted in a lot of silence in between narration. :)

    @matrixgamer: That's the way I've been doing it lately, except that I only have myself, and Lady Luck to test the truth of my ideas (I think "argument" is a better way of framing it, now that I read your post). :)

    [Edit: I should add, though, that I don't test the whole argument, but rather one or two of its ideas (usually the last ones). ]

    I also like to have some constraints on my "arguments" such as having prior premises strongly support them. It approximates that feeling that I'm not just running willy-nilly with ideas, but have some "force" outside myself guiding the narrative. Yet, that force is not entirely or even mostly based on randomness, so I feel it flows a bit easier.
  • The recording idea has merit, I've edited a few actual play pod casts in the day and dead air/pauses are the easiest thing to remove with the right tools. Make you sound smarter too. :-)
  • Have you tried structured arguments?

    For instance:

    Action: who does what, where, when?

    Result: What happens because of this?

    Reasons: Cite three reasons to support the argument that come from past events or material from the scenario/world with only one slot left for pure logic.

    Structure and limits aid creativity.
  • edited September 2013

    Can you clarify what you mean by this fragment: " with only one slot left for pure logic"?

    Other than that, I think this is what I've been experimenting with, except I've been trying to find the reasons before I deduce some true conclusion from them:

    If Premises 1, 2,...N are true, Then Conclusion x must be true

    I've found that most of the time I can't rely on deduction alone, though, because the premises don't always guarantee a 100% true conclusion. So, in these cases I usually rely on randomness for a final judgement. This is what I'm trying to reduce, though. Either I need some sort of arbitrary cutoff where I say "if it's more likely than not just treat it as a True".

    I like your suggestion, because feels very similar to causal inference. Still, I think I will still run into the "issue" above, because the post-facto reasons may not be conclusive-- so I will still have a lot of situations where the conclusion will be "likely true" instead of absolutely true.

    Thinking about it, I have a few ideas on further constraints as far as when to use randomness:

    1. If something is more likely to be true than not, just treat it as it being true. It happened in the fiction.
    2. Only use randomness to determine truth when some conclusion is not likely to be true given the established facts/premises.

    I could combine this with an idea similar to the one you shared above where I just write a scene, or a sequence of x number of actions and then later use randomness to test the truth of each of the conclusions/actions/plot points made. If one of them is determined to be false by luck, then I need to see how that affects the conclusions after that (it may not at all). It's still going to be quite a process, but I think it might flow better for me than just asking an oracle whether something happens before I go with it.

    As an aside, I was actually thinking of creating a mini-game for the Gamechef competition based on a similar idea, except I'd be using a Bayesian probability algorithm to try and change events in the past [Edit: or, more to the point, in an established storyline.]. I thought it might be fun to try it with things like the plot of Blade Runner, with Roy Batty as the Hero, and try to change its facts to change the story. Each time one fact changed, one would have to consider the effects down the line, if any, at least in outline form. Then one would keep testing the facts down the timeline in an iterative fashion until there was some sort of end.

    However, I never considered trying to do what you mentioned some Matrix games did. I'll give that a whirl to see how it feels.
Sign In or Register to comment.