Hidden in Shadows

How do you stock your dungeon?  The current and long-standing understanding of dungeon stocking in Dungeons & Dragons is that the design intent of the game is for the Dungeon Master in their wisdom to create most if not all treasures and monsters out of their own imagination.  

This essay argues that's bunk.  Instead I will show  that the initial design of Dungeons & Dragons by Arneson and Gygax was built around a deliberate stocking method using controlled randomized rolls giving a particular range of results tied deliberately and directly to the intended progress of character growth.  

1) Reasoning from History

Both before D&D was written, and long afterward, Dave Arneson stocked his dungeons randomly and he devised different methods and  applied different ways to do this.  Monsters, for example, he stocked through a random "Protection Point: system.  Gold he rolled dice for and items he created random tables for.  The best early example is his The Loch Gloomen stocking list from 1972, reprinted in his First Fantasy Campaign booklet. 

When typeing up the first draft of D&D, Gygax created the Treasure Types tables, apparently by expanding on Arneson's dragon treasure tables (FFC81:66).  In any case, both Treasure Types tables and the corresponding Monster Reference table (as found in M&Tp3) are present in the original 1973 draft of D&D.  

As a somewhat side issue, it is important to note that at this point in the game's development, there was no hint that the Treasure Types tables should only be for wilderness "lairs" or that a lair was in anyway a wilderness related term.  This idea didn't take hold till years later.  There is no separate method in the original draft for stocking dungeon and wilderness treasure.  The Treasure Type tables given are meant for general application.  Nor is there any suggestion anywhere in the rules that you should ever just stock by making things up.  

Think about that for a moment because it is very significant.

Now, if we were to stop right there in time; if no more had ever been written about D&D, would anyone ever have argued that the Treasure Type tables were designed exclusively for wilderness play?  Would anyone argue that the rules expect you to stock treasures by just making them up on a whim?

No one could argue that.  There is no hint of any such thing.  In the draft, the Treasure Type tables used in conjunction with the Monster Reference table appear to be for generating ALL treasure ALL the time.

We also have to ask if it even makes any sense to think "Wilderness Treasure" in the first place.  The treasure types are fairly complex, so why would Gygax - in his very first (complete) game draft - have thought it necessary to create the Treasure Type Tables unique to the relatively rare wilderness play, but not have any tables of any kind for the much more common dungeon delving?  

That's nonesensical.  There is really no question at all that the Treasure Type tables were intended for general application, following Arneson's practice of randomly generating his treasures.

Deductively, the history settles the argument as to the intended purpose of the Treasure Types.  Case closed on that.  But knowing the purpose of the treasure tables doesn't settle the question of how often to use them, which is really the crux of the matter, so let's go on.

2) Reasoning From Game Design

As noted above, every instance we have of Dungeon stocking in Blackmoor, both before and after D&D shows that Arneson stocked rooms with monsters and treasures randomly.  In some cases he decided what kind of monster he wanted, but he still generated numbers of monsters and the content of their treasures randomly.  Do we have any reason to think Gygax rejected this approach in their co-written rules, or did he embrace it in the rules?

Consider first, that there is no need to make any kind of treasure tables at all.  D&D plays perfectly well without them, a fact amply demonstrated by the overwhelming creation of treasure through DM fiat in countless published modules.

So clearly, Gygax was following Arneson's lead when he included random treasure generation in the game and he must have felt it was important.  Why did he do so?

Consider this passage from the Dungeon Masters Guide:
"…the MAGIC ITEMS table is weighted towards results which balance the game. Potions, scrolls, arms and armor are plentiful....this is done in order to keep magic-users from totally dominating play...what they gain from the table will typically be used up and discarded [while items for fighters are permanent]....This random determination table needs no adjustment, because of its weighting, and weighting of the MAGIC ITEMS table....this is carefully planned so as to prevent imbalance in the game." (1979: 120, 121)

Now, the Magic Items table in question is a subtable in the treasure generation process.  If this portion of treasure generation is carefully weighted, it follows that the rest of treasure generation process was also designed with character growth and balance in mind, and that's precisely what Gygax tells us in a paragraph he added to the Holmes rulebook:

"The tables are designed to maintain some sort of balance between the value of the dungeon's treasures and the risks involved in obtaining it. It is highly recommended, for this reason, that neophyte Dungeon Masters use the tables."  (p33)

When Gygax is speaking of risk/reward balance, he is harkening back to David Megarry's design principles for the Dungeon Boardgame, in which the amount of treasure and type of monster was carefully gauged on each level.  That is, monsters are weakestand treasures of least value on level one and strongest/most valueble on level six.   Megarry's design pre-dates D&D and had a significant influence on the rules Gygax designed.  

In D&D, treasure and monster values determine Experience Point values.  Since Experience Level in D&D is tied closely with treasure acquired, it follows deductively that Gygax is telling us the treasure tables determine not just the quantity and type of magic and gold in the game, but also how many Experience Points a group of characters are meant to acquire in a level by level, risk/reward fashion.

I'll say it again for clarity.  Contrary to common belief, D&D was designed from the initial 1973 draft to have a central mechanism regulating how quickly characters advance via the amounts of treasure they can acquire in relation to the risk they take.  There is not much point to having an Experience Point system based on point totals if there is not a correspondingly controlled method of how points are earned.  The treasure tables were designed from the start to be the regulating mechanism for how experienced points are earned.  

Replacing the treasure tables with DM fiat effectively kicks a leg out from under the system.

With the publication of the Holmes rulebook we see the end of the OD&D era and Gygax is already rethinking his approach to the game.  By the time AD&D is published, he has clearly abandoned the "follow the tables" approach for stocking treasures in favor of DM fiat.  Likely he did so for a number of questionable reasons such as wanting to slow down character growth in the game and an irrational fear that " the hoard of a dragon could destroy a campaign if the treasure of Smaug, in THE HOBBIT, were to be used as an example of what such a trove should contain." (DMG p92).   I'd say the last is irrational because it reflects a difference from Arneson's approach, who was perfectly happy to let a character of any level (such as Fredigar Cripps who builds the famous Comeback Inn) achieve instant wealth through a lucky score.  For Gygax, the game consists of a uniform controlled ascent to wealth and power, and somewhere along the line he decided the treasures he had previously devised were too generous.  Much as he also decided the original rule of 100 XP per monster HD was also too generous ("ridiculous" in his words.) and reduced it to only 10 XP per HD.   So we see an extensive Essay in the DMG on pages 91-92 on the placement of treasure - an essay that is next to useless in practical terms in my opinion as it boils down to "use your judgement and be stingy".  Nevertheless it is interesting to see him reiterate some of the same points echoed in the Holmes statement:

" First, we must consider the logic of the game. By adventuring, slaying monsters or outwitting opponents, and by gaining treasure the characters operating within the milieu advance in ability and gain levels of experience. While AD&D is not quite so simplistic as other such games are regarding such advancement, it nonetheless relies upon the principle of adventuring and success thereat to bestow such rewards upon player characters and henchmen alike. It is therefore incumbent upon the creator of the milieu and the arbiter of the campaign, the Dungeon Master, to follow certain guidelines and charges placed upon him or her by these rules and to apply them with intelligence in the spirit of the whole as befits the campaign milieu to which they are being applied.

A brief perusal of the character experience point totals necessary to advance in levels makes it abundantly clear that an underlying precept of the game is that the amount of treasure obtainable by characters is graduated from small to large as experience level increases ..." (p91)

So while he has abandoned the table approach here, he nevertheless acknowledges that there was an intended "logic to the game", which the tables must surely have been meant to reflect when they were created.

3) Reasoning From Example

Despite the fact that D&D has two authors, Arneson gaming is frequently not given due consideration, or is simply not of interest to some.  That's fair.  People can't be made to take an interest in something they aren't.  So while I, or anybody inclined to take the time, can demonstrate that Arneson used the D&D tables to randomly generated the treasures in both the first 6 levels of Blackmoor dungeon and the 2 dungeon levels of the Temple of the Frog, seemingly many gamers couldn't care less.  They only want to know what Gygax did.  Unfortunately we have less published OD&D era dungeon material from Gygax than we do Arneson, oddly enough. 

I'm stressing published here because it is material prepared for public consumption that is most likely to conform to the rules as intended.  The earliest dungeon published by Gygax was just prior to the AD&D advent: the tournament version of the Caverns of Tsojconth (1976).  For this adventure he used many monsters which would be appearing in the Monster Manual and so I've checked the data here against that volume.

Rooms	Monster	Treasure Type	Treasure	Conforms?
A	Blink Dog	C	N	Y
B	Stirges	D	N	Y
C	Displacer Beast	D	360 PP 1,100 GP	N
D	Flesh Golem	Nil		Y
E	Lurker Above	C Y	100 sp 50 pp 1300 gp 4 gems	N
F	Green Slime	Nil	1200 Jewelry, 3 gems	N
G	Cockatrice	D	Scroll	N
H	Giant Turtle	Nil
I
J	Fire Lizards	BQST	100 cp, 2000SP, 3000GP, 700 1GP Gems, 6 5000GP gems, 2 potions	N
K	Copper Dragon		Too Large to list here.	Probably?
L	Sahuagin	N	NT	Y
M	Rust Monster	Q	2 Gems	Y
N	Water Weird	IOPY		Y
O	Giants	E	3 tusks, 2 cloaks, Magic Boots,	N

I've only done level 1 here but level 2 will give similar results.  We can see that in 1976 Gygax pretty clearly is not strictly following the Treasure Types tables, nor is he following the treasure by dungeon level table found in U&WA.  And yet, the treasure he does list are "close" in values and quantity to those the Treasure Types tables generate, as if he is winging it with those tables in mind.  It's certainly not the sort of "gimped" treasures found in his later published works.  It suggests that Gygax was willing to be guided at this point by the tables, even if he didn't always stick to them.

4) Reasoning From Practice
Let's simply look at what it take to employ our two competing methods.

To stock a single room by DM Fiat:

Imagine a monster and pick their numbers.
Make up their treasure.

To stock a single room by the book:

Roll 1d6 for occupancy
roll d6 for monster determination level
roll d10 on monster level table
roll dice as required on for #appearing in room
roll on Treasure Types table as follows
Roll % for copper then roll d# for amount
roll % for silver then roll d# for amount
roll % for gold then roll d# for amount
roll % for gems 
If gems are present roll % dice first then roll a d6 to determine value
If jewels are present roll % first then roll a d6 or a d10 to determine value
roll % for the chance to roll on the Maps or magic table
For each positive, roll on the table described
For "any", roll % on maps or magic table.
If maps, roll % on maps table, then roll either magic or treasure as prescribed
If treasure map roll d8, then roll dice for prescribed treasure

For any step above that results in magic, roll % for type  
then roll on the tables prescribed to determine magic item, often requiring multiple rolls.  Magic swords may require more than a dozen separate rolls.   

Repeat this 30 or 40 times for one level.  

I stocked a 10 level dungeon using this method.  A single room can take dozens and sometimes scores of rolls and chart lookups.  It was exhausting and took me about 8 months off and on.  No kidding.

Now let's consider the quote advised by Gygax:
""The determination of just where monsters should be placed, and whether or not they will be guarding treasure, and how much of the latter if they are guarding something, can become burdensome when faced with several levels to do at one time." U&WA p6

Given the difference in effort between ease DM fiat and the chore of rolling BTB as detailed above, I can't see any basis in the slightest for those who would claim that Gygax meant to refer to dungeon stocking by DM fiat as the more "Burdonsome" of the two approaches.  It's frankly absurd, and yet typically that is exactly what people have argued it is supposed to mean.

We can definitively clear up the matter simply by careful reading of the rest of the passage, "It is a good idea to thoughtfully place several of the most important treasures, with or without monsterous guardians, and then switch to random determination for the balance of the level. Naturally, the more important treasures will consist of various magical items and large amounts of wealth in the form of gems and jewelry. Once these have been secreted in out-of-the-way locations, a random distribution using a six-sided die can be made as follows:.." U&WA p6

Note in particular the phrase "several of the most important".  That clearly does not say "most of the treasures."  For there to be "several" "most important treasures" there must conversely be even more treasures not as important.  There is really no way to read that except for it to mean that the "carefully placed" treasures are a minority of the treasures on a level, with the majority of the treasure " for the balance of the level" being generated with dice.

Thus the very passage many cite to "prove" DM fiat is the intended method for dungeon stocking, actually established the exact opposite for all but a few exceptional rooms on a level.

Let's look at one other aspect of that passage, the use of the word "Determination".  As author, Gygax is certainly aware that he uses this very word in the title of the stocking tables, such as the "Monster Determination &. Level of Monster Matrix" the table "To determine the kind of treasure" (p7), the "Magic/Maps Determination Table" (U&WA 23), and so on.   I don't find this to be coincidental.  The determination of monsters in the text is harkening back to the determination tables.

Surely by now you are convinced.  No?  

Objection 1: - M&T says the Treasure Type is for lairs, and lairs are only in the wilderness.

Even if true, and mind you I'm arguing it's not, it still makes no difference.  Page 7 of U&WA gives a dungeon level treasure table increasing in wealth and magic as the dungeon deepens.  I think this table may be intended for unguarded treasures, but anyone who would argue against using Treasure Type tables in the dungeon would still have this table as their alternative, with the understanding they may end up with even bigger treasure totals by using it instead of Treasure Types.

Having said that, there are a number of things I could point to illustrating that "lairs" are not restricted to the wilderness, but that's a whole other discussion.

Objection 2: Gygax didn't use the treasure tables in his dungeons!

I honestly think this is the weakest of all objections.  Let me ask a simple true false question.  Which sentence is true:

a) Gygax is well known for being a stickler for the rules he wrote, insisting that he follows them to the letter and you should follow his example.
b) Gygax is notorious for writing rules he himself ignored, never using them in his own games, often preferring to lean on DM fiat as a one of chief architects of the game.

If b is correct, then what Gygax did in his dungeons is not a guide to what he intended in the rules.  I think it is really interesting to know what Gygax did in practice and how his handling of the rules changed over time.  Trying to suss out and copycat his methods where they differ from the rules could be fun and maybe an argument can be made that in some cases he made improvements that we should emulate.  However I think we have to admit that what he's doing with treasure in practice is a separate thing from what he designed the tables to do.

Objection 3: The Random Treasure tables are Broken
I think this all depends on how games are run.  Original D&D expected Referees to take logistic and encumbrance more seriously (how are you going to carry 10,000 sp?).  There was also an expectation that more treasure would be siphoned off through taxes and expenses and that a greater number of players would be sitting around the average table, so that treasure shares would be smaller.  In any case, it's apparent that within a year or so of D&D's publication, Gygax wanted to slow down the speed of his players progress.  So, Gygax, who may never have actually followed the treasure rules he designed for publication, came to downplay the role of the treasure tables, and implied they were too generous in wealth for common use.   This, I think, led to the current confusion and lack of use among Referees.  

Objection 4: Okay, maybe that was the way it was back in 1974, but Gygax changed his mind, and like him, I don't like that old system.  I prefer DM fiat.

<shrug> It's your fun dude, have it however you like.  Just be aware and transparent so everybody who encounters your game knows what you have done.