Monday, February 22, 2010

Chart Granularity

I've mentioned before that I love using and perusing some of the great random charts created by the OSR. I've also pondered on the blog about the OSR paradox of do-it-yourself versus using all these cool things people in the community make and share. My solution was to try and split the difference, to make more abstracted DM "Spurs" that could help each of us do-it-ourself, but also function as an aid that, like old school play, allows complexity to emerge from randomness.

Also, keep in mind that I have settled on the idea of using Grim's Roll All the Dice method, because it is elegant and a fun game-like way to involve players in a lot of these randomization discoveries. I don't want to stop during play and consult a book of charts, or at least if we must the book should have the fewest possible charts to work elegantly and still be fun.

So, the problem seems to be how granular do you make these spurs. If we would like a magic item generator do we separate out the "Item" part from the "Magical Effect" part?

I think so. An Item Spur can have lots of important info including Size, Function and Material in addition to item type that is hard to squeeze into a spur that is also concerned with Duration, Range, Charges etc. of magical effects.

But even there, how granular do we make the Item Type? Think of jewellery. We could have a spur that would determine item type solely by the location on the body it is worn:
  • Head = Tiara
  • Neck = Necklace
  • Arm = Armlet
  • Wrist = Bracelet
  • Hand = Ring
That works well, and we could probably fit clothing and armor to the same system. So, you might have one die determining Location (arm, neck etc.) and the other Item Type (jewellery, armor etc). Fine, simple, compact, functional. But . . . it loses all the cool flavor, the details, that I would want to have if I were the player receiving treasure. This system will give you a Ring result, but what about pillbox rings, or gimmal rings, or signet rings? If you try to include each of these interesting real world examples, our elegant spur will turn into a multi-page chart very quickly.

I think one possible solution is to have a Function category, so, if a random roll indicates that an item Depicts, Commemorates, Hides, Stores something, the our spur might be able to produce something similar to real world examples.

Take a look at this cool object that I'd never heard of before yesterday:

It's a Golden Hat which, aside from the fact that it is shaped like no other hat I've seen, apparently has solar and lunar Calendar data recorded on it.

This is a kind of touchstone. I want any Spur I create to be of the right granularity to be able to produce a result such that we might say "Aha, it's like a Golden Hat."

That would be cool.


  1. Let's see how the community figures out a way to go about doing so. :D

  2. Ah, granularity.

    I have no idea what that is, but it sounds like something I could wear my teeth down on.

    I like lots of random generation tables though.

  3. :) Granularity is just a way to talk about specificity. Maps are a good example: a map of a whole country is useful for visualizing voting patterns and such, but lousy for helping you find the airport in your city.

    With charts, I'm basically reacting to the chart of single items. You know, lists of cool treasure items or books with specific titles. They are single use. Once you roll that result you will want to cross that option out or replace it with something else.

    That's a lot of work.

    Essentially you as DM have to generate every item and the chart just randomizes its selection. On the other hand, is having a chart that also helps generate the result. These are cool because they save you work, but because they are necessarily more abstract, if you aren't careful in how you craft your chart your results can be homogeneous, boring.

    That's the problem I'm trying to solve: how do I make a chart that will result in as near infinite results as I can make it while still being interestingly specific; creating unique seeming items.

  4. Telecanter: I began (and haven't abandoned, but have put off completing) a d100 chart where each of the 100 entries then had 10 sub-headings which could be read as a whole, or each sub-entry could instead be randomised from among the other 99 entries' sub headings.
    --Someone calculated the permutations and it was ... a large number. Granted, not near infinite, but likely more than would be exhausted by any one GM in a gaming lifetime (although no table can exclude rolling the same results).

    Likewise, in my Life Lab creature creator table, it is a roll everything at once, but it cross-indexes dice (d8 and d4, d6 and 10, d4 d8 and d12, etc.) all from the one Roll Everything At Once roll.

    Perhaps the combination of those two sorts of tables can be integrated into your maximum extensibility, minimal repetition table. :)

    Either way, I look forward to what you come up with. :D

  5. Timeshadows, those sound interesting and may be doing something different than I've ever seen with a chart. Do you have a blog or forum post with more discussion of them you could link to, or perhaps drafts that can be downloaded somewhere? I'd love to look at what you've done.