I mapped a L1 Dungeon of 63 rooms, using the random tables and then laying it out as made best sense. I got 40,063 XP in monsters and 88,422 GP in treasure. I intend to go through and stuff additional treasure in, of course.
But I was curious: how did it end up so far off?
One possibility: Do the treasure tables count the magic items toward their total value? I spot checked two, and they added up without the magic items. So I’m inclined to think “no.” That also makes sense, because magic items usually aren’t a source of treasure-based XP.
Another possibility: Were the champions, subchiefs, chiefs, and fighting non-combatants ignored? That seemed more possible, especially since the fighting non-combatants often represent almost half the fighting force of a village or lair, and two huge sections of my dungeon are an orc village and a lizardmen village.
A third possibility: The “special” goods are undervalued. But again, I spot-checked the Gold table, and it actually comes out to an average of 1,035.2 gold per 1,000 gold. So it’s probably not that.
Lastly: It could be that the treasure tables are swingier than they look. Really, really swingier. Enough so that even 63 rooms isn’t enough for results to start to approach the average. Unfortunately, short of coding an entire dungeon generator, I don’t have a way to test that one.
More lastly than that: Something that hasn’t occurred to me. Maybe I did something wrong? Can anyone think of something I might have done to arrive at almost exactly half the treasure I would have expected to have?
My guess is (Another Possibility) and (Lastly) combined. If you replace your treasure piles for the “average” result listed on the column, does it add up correctly? If not, try not counting the conditional monsters. Does it work now?
Anyway, my /single/ experience with this came up with a very similar result to yours.
Plus, there’s a lot of room for things to get swingy in. Even if you were to use the “average” treasure rates religiously, it’s still possible to end up with a dungeon packed full of pit traps and fire beetles and thus, no treasure.
The formula we used to develop the treasure types are esoteric and fill up many many pages worth of spreadsheets. But, more-or-less, the answer to your question is that "it's very swingy." In most cases, you will want to do a second pass to correct the treasure allocation in the dungeon. That's why we provide specific guidelines on how to do so. We could have made the treasure types less swingy, but then you've boxed the GM into a corner for those moments when he wants rich orcs, or poor dragons.
I know some GMs who do play with "the dice are always right" mentality, but I personally prefer a less gonzo game, so I tend to adjust treasure allocations by hand to match up to my desired ratios.
I let the dice (or the treasure generator) decide how much coin is actually there. Like Alex I then adjust the allocation to shoot for the 4:1 ratio, but I do this in play rather than by placing more treasure. For example, if the treasure haul from a few sessions is 1,000 gp short, I might decide that one of the special items is an antique worth 1K more to the right buyer, or that one of the monsters killed had a 1K bounty on its head. I like this approach because it creates opportunities for roleplaying and side adventures in town (might that antique be worth more elsewhere? does the bounty-giver have other monsters on the list?) and lets me preserve the self-directed illusion that the dice describe a real world.
When treasure is more than expected, I'll look for opportunities to generate some more XP from combat to bring the ratio back into line - ruffians hearing about the party's big score and looking to jack some of it are a logical way to do this, for example.