Letters as Dice: Gaming Without Moving Parts

This scheme allows one to use a printed (or written, or memorized) section of text as a set of dice.  You take the next letter in the sequence, and look up where it goes in the table. If you see a dash instead of a number, move onto the next number.

This isn't so much an alternate dice system for fun as it is a way to play in cramped quarters with limited resources (no dice, no computer, no cards). Though if you want to use it for fun, go ahead ...

The Table

History

I have a long history of doing 'table-top' gaming without a table. It started on long walks home from school with my brother. We just used the rules by rote. We needed to get home relatively quickly (rather, 2 miles is a long walk and if you dally you never get home), so we couldn't slow down to roll dice. And our dice were precious and we would never risk them out there. So instead we build a spinner out of a piece of foam board, a protractor, and some nice, new bailing wire. This was D&D, so one roll sufficed for most purposes. The first spinner worked well enough. Then it got smashed or something, and we never could make a fair one again.


So we tried random number picking. When it was time to roll, the GM would pick a number, and the player would pick a number ( both in the range of the die results). Then we'd add the numbers, remove any full dice we got (so a 2 and a 5 gives a 7, which is greater than 6 by 1, so the roll was 1). I don't think my players ever figured out the exact mechanism I used, but they trusted me. By that time I was playing in all sorts of weird places -- hiking trips (weight is a premium) and especially family road trips. We'd have the back two rows of the van filled with players. It was pretty cramped, not only with people but with luggage. Using dice was a good way to get them dropped in cracks, under seats, in luggage, and generally lost (there was no space to go digging around).

That was when we brought out the dice bottle. We'd stick all of the dice in a clear plastic container (the shell of an old 1 quart spice bottle) with a flat surface inside, and shake that -- kind of like a game of boggle. It worked well for travel, but it was exceptionally bulky and you had to remember to bring it. And my siblings had limited six sided dice.

When I started playing gurps, I'd been away from pretty much everything for about two years (lds mission) and the dice bottle was long gone. Many of the dice were gone or horded by individuals. And picking three numbers for each roll was just long and tedious. Then in school I ran across hoffman encoding in school -- a large part of which relied on the frequency of letters in the English language. I knew I had my random number generator. It took some fiddling, but I got the conversion down well.

warnings about usage

One of the most important things to select is your text -- the written words you will be using as your random strings. Its important to use every letter, not just the first letter of each word, and I wouldn't trust myself to pick truly random letters off a page -- humans trying to be random often fail. Start at a place and move from there. Most texts will work. A few won't. Poetry is often a bad place to look. Some common phrases are a bad idea: "The quick red fox..." has every letter in the english language on purpose. I personally suggest a section from the rulebook. If you're playing from memory (which , get a good section of text. perhaps from conversation or a memorized quote.

The results will be fair overall. They will resemble dice overall. But they won't have quite the independent effect of dice: language has patterns in it. In particular I'd advise against using the 1d table 3 times instead of the 3d table: I can't vouch for the accuracy of patterns of three letters. In fact my gut says it will certainly be off by a bit, I just don't know how much and which numbers are favored. This isn't to say its unfair. Its just to say the odds may not be quite what you expected.

The assumed statistics about the english language are from here. They are close enough for my purposes: your selection of the english language won't match these numbers exactly, and even if they did, the odds for specific numbers are merely close, not exact. I'm distributing 26 pieces over the potential die results -- I can't get that fine grained. The 1d table is within 5% of the actual value, the other two are within 10%.  (if a number has a 1/8th chance to be rolled, at most its 1/80th off the chance dice would give you).