It's desirable, when setting forth a world, to show both the predictable occurrences that give it meaning and the unpredictable ones that give it surprise. I don't have much of an interest in running a world where everything is unique and surprising, because I suspect that people would quickly bore of it. You need to get an idea of business as usual, in order to really appreciate business as unusual.
Typically, a random table will handle this by including greater- and lesser-probability outcomes. But there's another way to get this effect.
To illustrate, roll a d6 five times, then another five times. You'll get two lists of numbers like:
5, 1, 5, 4, 2
1, 6, 4, 2, 3
If we try to match the numbers as closely as possible that gives 3 exact matches - 1,4, 2 - and two that are off by 1 or 2 points - 5 to 6 and 5 to 3. More rare are sets in which the matches are way off - in which a 2 is forced to go with a 6, for example. So this seems like a good procedure for generating, not expected or unexpected lone elements, but expected or unexpected match-ups.
The point of the classic treasure types is that monsters have different kinds of treasures - orcs have coins and looted objects, sphinxes have scrolls and magic items, dragons have, well, everything. Rather than make a cut-and dried list, though, why not group encounters into units of five or so, and try to create the best matching possible within each unit? This will give a mix of expected and unexpected treasures.
So as an example, the treasures I generated with my new table were:
A well-crafted large stone statue worth 6000$
100$ of drugs and 2 gold pieces
2000 cp (what, no rats?)
Two high quality, decorated swords each worth 1000$
A Grand Hoard of 4000$ in silver, 5000$ in exotic hides and a 1000$ gem.
And the monsters of approximately appropriate level:
An NPC, level 6 (fighter)
4 gibbering mouthers
A juvenile ankheg
Some of these matches are easy. The NPC is the one with the gold and drugs (a 3 round dose of powdered haste). The ankheg is the weakest encounter, and has turfed a sack of crappy coppers into its burrow. The gibbering mouthers are the strongest challenge, so they guard the grand hoard.
Then some are less obvious. You could decide in the end that the swords belong to the NPC and the winter wolf is her pet. This leaves the statue with the owlbears, and it's just coincidence that it's in their lair. The statue is the second best treasure, but also a complete white elephant to move, and may not even be recognized as treasure.
I find that five units is about the right size to get a mix between completely appropriate monster-treasure "stories" and surprising or strange stories. It also occurs that this might be a good way to get other matches done - monsters and lairs, for example, or tricks and their effects.
Terrible Campaign Ideas: Too Many Gods
2 hours ago