Vault of the Drow
Domains, Demography, and Warcraft of the Dark Elves
I recently completed a thorough review of the TSR D-series modules — D1: Descent into the Depths of the Earth, D2: Shrine of the Kuo-Toa, and the superb D3: Vault of the Drow. Playing through that series was one of my very first experiences upon being introduced to D&D, so it’s close to my heart, and one of the great classics. I have never had the opportunity to run it, but a close reading seemed a good prerequisite to doing so, and I’ve planned for some time to start seeding rumors pointing to the G-series modules in my ongoing Shattered Lands campaign so I’m optimistic to see the D-series on the table in the moderate future (primarily depending on how soon my players stop conquering cities and go looking for dungeons, though if left unchecked then I may get to instead play out drow raiding forces plundering the surface . . .).
Treasure
Naturally a first step in conversion was to sum up the XP and gp total across the three modules:
D1 contains 111,526 XP of monsters, 313,090gp of treasure, gp:XP ratio 2.8,
D2 contains only 38,073 XP of monsters, 552,118gp of treasure, gp:XP ratio 14.5,
D3 contains a stunning 484,752 XP of monsters (not even including the additional 170k XP of the 8000+ inhabitants of the city merely alluded to), an equally impressive 1,931,419gp of treasure, for a gp:XP ratio of 4.0, right what it should be for ACKS.
If a group works through all three modules and gets about 85% of the total XP, that's close to an even 3M XP. If that group has six PCs of around 10th level, plus a dozen henchmen or so, that's enough to just about level everyone twice. A very hearty haul. D1 is treasure light by ACKS standards, and D2 treasure heavy, but across the three modules the ratio is 4.4 gp per monster XP, very reasonable. Running the modules in ACKS, the treasure should need no adjustment.
Domain Structure
Per D3, the chief economic activities of the drow seem to be control of trade in the Underdark, and the cultivation of rare fungi. Control of trade attaches itself naturally to large settlements. Checking the mushroom farming rules for domains in By This Axe, two more significant requirements emerge: first, mycoculture requires dwarves. That is easily enough expanded to permit other Underdark-dwelling races. The second requirement is perhaps more challenging: a vast quantity of wood is needed as an input to be composted by the fungi.
To that end, I propose that perhaps different varietals of mushroom rely on different food sources, and that while the most convenient food source near the surface is wood, within the depths of the Underdark can be found other strains that feed upon the strange radiations that permeate the deep caverns where drow have chosen to settle. Similar varieties might exploit geothermal and magmatic vents, or lakes of strange chemicals. This has the interesting consequence for domains in the Underdark that most land there effectively cannot be settled — instead, settlements must cluster tightly around the scattered points where food can be produced, or must have a very secure means of importing food in bulk. This very aptly fits the lore.
It is not mentioned in D3 directly, but it seems very fitting as well for some drow to oversee slave-mines amidst the Underdark as well. I plan to treat these as outposts that are supplied centrally from parent-settlements, and privately owned by the great houses of the noble families as their major sources of wealth. Provision of food produced in abundance by fungal farms is perhaps a similar basic lever used to control the beastman clanhold-vassals that surround and pay tribute to the drow.
Deep dwellers in large settlements called vaults, mycoculture, mining — does all this together step too much on dwarven toes? I think not, for two principle reasons. First, I think it makes sense that different societies trying to survive in the same basic ecological-economic context will find similar solutions. The “dwarf package” is largely just the “underground-dweller package”, contextualized for the only race that commonly does so near the surface. If this is dwarf-exclusive, it’s hard to work out how any other Underdark dwellers can function.
Second, dwarves still retain some unique advantages in this system — dwarf families can be dynamically reassigned between different functions, which lends them rather more flexibility. They can also work as miners even if not enslaved, and thus benefit in morale. Drow, by comparison, seem to view manual labor as beneath them and will use slaves for both mining and mycoculture, requiring both more laborers and suffering worse morale when doing so. They also cannot delve deeply — that dwarf subsystem seems best to keep dwarf-exclusive, to reflect their superior stoneworking. Last, to secure that dwarf niche, it is worth noting that dwarves remain the only race that operates on both the surface and in the depths. As a species, they have a remarkable adaptability and flexibility, to operate in either environment and tune their allocation of labor to the environment, which effectively translates into the resilience and endurance that they’re known for.
Economics and Demographics
The challenges of demography are rather more difficult, but can potentially be resolved. A typical drow warrior is 2 HD, a 2nd level character. That could be explained by treating them like beastmen — they would be inherently 2 HD creatures, due to natural ferocity and ability. However, this largely precludes them being a playable race at that point (not awful but not ideal), and more importantly does not match how demihumans are treated more broadly — elven and dwarven combatants are 1st level characters, not 1 HD monsters. It’s also basically trivial in demographic terms, and thus unacceptably precludes the opportunity to play with spreadsheets.
The first option I considered was to simply build the realm as a demihuman clanhold, which would entitle it to abundant levelled characters, and seems broadly justified by drow raiding practices. But this encounters two core problems: first, clanholds don’t form large cities, and the domain in question is fundamentally built around a large city. Second, what would prove to be the most challenging issue, is the simple matter that a clanhold doesn’t explain how all of these 2-4 HD troops are being paid — they aren’t just raiding, but actively patrolling a fortified environment. Clanhold demographics also generally overcorrect: compared to default elven demographics, the quantity of 2-4 HD troops in D3 is only 14-31% greater than we would expect in a typical elven domain, not a huge margin, whereas a barbarian realm more than doubles those totals.
Using the as-yet-unreleased BR and wage formulae from the revised version of Domains at War: Battles (available to King-tier backers of the ACKS Patreon!), I calculated the monthly wages of 2 HD drow warriors at 120gp each, 3 HD veteran drow warriors at 240gp each, and 4 HD elite drow warriors at 480gp each. Counting only those drow of 2-4 HD actively serving in obvious garrison roles in the hex of the vault itself, I total 511 troops and 92,640gp in wages. Their mass combat stats as companies are given below.
120 Drow Warriors, LF, UHD 2 (uhp 16), MV 2/4/6, 4 Sword & Shield 8+ or 3 Crossbow 9+ with sleep poison, AC 7, SV E2, ML +2; magic resistance -5; BR 20
120 Veteran Drow Warriors, LF, UHD 3 (uhp 24), MV 2/4/6, 5 Sword & Shield 7+ or 4 Crossbow 8+ with sleep poison, AC 7, SV E3, ML +3; magic resistance -4; BR 40
120 Elite Drow Warriors, LF, UHD 4 (uhp 32), MV 2/4/6, 6 Sword & Shield 6+ or 4 Crossbow 7+ with sleep poison, AC 8, SV E4, ML +3; magic resistance -3; BR 80
All drow warriors additionally have lightless vision 90’, natural stealth, quickening, SLAs (illumination and tenebrosity 1/day each), sunlight sensitivity (-1 to hit, AC, morale, and saves vs fear in bright light and -2 in full daylight or equivalent). Females have MV 2/5/8 instead of 2/4/6.
The Domain
Of Erelhei-Cinlu, it is said:
Between 8,000 and 9,000 Drow live in the city, and double that number of half-casts, servants, and slaves. To this permanent population can be added a thousand or so creatures visiting for purposes known only to themselves.
Let us be optimistic, and assume that Gygax was noting only adult drow, and further assume the maximum of the range offered. That yields some 4500 drow families, 6000 slave gangs, and 500 metics. Adding up the slaves detailed elsewhere yields an additional 197 slave gangs, not including those on the estates of merchant houses. Since I assumed the maximum of drow within that range, I have declined to inflate that number further by adding in those individuals detailed throughout the module (an additional 1729 families worth if they were to be added in).
I have run the numbers many different ways, and it’s quite a strain to make those numbers work. If we treat the drow and metics as urban families (exactly reaching a Class II market for a net income + garrison expense of 5gp/family), and the slaves as mycocultural peasants maximally exploited and repressed (4gp land value for mycoculture plus the standard 4gp in services and 6gp in greatly elevated taxes, less 2gp for maintenance, for a total of 13gp/family, with the -6 penalty to morale rolls that imposes for high taxes and the absence of tithes and liturgies offset by repressing the peasants with at least about 64,000gp of troops each month . . . which they neatly have on hand). Having different tax rates for peasants and urbanites requires splitting this into two separate domains, but for various reasons, we want to do that anyway (the racial difference is part of that, but it also permits us to model the drow city with potentially good morale while the peasants are uniformly miserable under the crushing heel of generationally perfected exploitation).
Two incidental concerns: first, greater income is theoretically possible via special strains of mushrooms, and more justifiable than with dwarves (since the drow have access to ritual magic that makes it rather easier to successful experiments). That said, it still relies on experimentation, which I find highly dubious as an assumption, and moreover I think it poor niche protection for the drow to be better at mushroom experimentation than dwarves (I don’t know that I would forbid it outright, though I very well might, but they should be clearly worse at it). Second, land improvements might appear useful but close reading reveals they are not compatible with mycoculture (which is good and sensible, given how self-evidently broken they would be if they could be used together).
Taken all together, the combined income and garrison expenses of the entire city comes to 107,264gp monthly. There is virtually nothing more that can be done to further increase revenues, this is a functional upper bound on income. And the problem we immediately encounter is that just paying the wages of those drow who are obviously being used as a garrison consumes 86% of this income. That leaves too little for anyone else to subsist on, not to mention that does not include provision for the wages for the noble household guards, temple guards, or the militant orders! It’s a reasonable objection to argue that perhaps the nobles and temples cover those wages themselves, and there are some things they can do to help, but the bulk of noble and temple income is already accounted for in the city’s revenues (the merchant clans do have independent income sources, so I exclude them and their troops from these numbers). We can see that just as a matter of income, solutions that involve paying troops their normal wages seem clearly nonviable.
The above analysis came about as I attempted to instead make some variation on the (also a recent Patreon release) republican city state rules fit, which I found intriguingly well-suited thematically but fiscally implausible for these circumstances. Those rules would be better suited to a circumstance where there were simply too few combatants in the society and a larger proportion of the total manpower was needed to be mobilized, but that’s simply not an issue the drow face: their problem is one of economics, not demographics. There are 2208 drow soldiers under arms; a domain of 12,475 families can call up 1248 conscripts and 2495 militia, so while we’re near the upper end of standard mobilization rates, these numbers are ultimately reasonable even if we neglect Erelhei-Cinlu’s position as the sovereign over surrounding drow settlements that provide a larger population to pull from.
This is what ultimately persuaded me to return to Helgeran’s excellent Knightly Realms rules that he has very kindly shared on the ACKS Discord, which I had initially rejected as not quite fitting what I was modelling. They are an excellent system for supporting a large number of 2nd level knights spread across the countryside, by making them rulers who can be called up for military service, providing them with campaign XP qualifying income and only needing to meet their wages as levelled characters (rather than as troops) to prove reasonable (they can be paid less in essence because their pay is qualitatively superior, being under their own control in perpetuity, an XP source, and with the measure of autonomy gained by being rulers). For the 2-4 HD troops here in question, such wages are only about 40% of their wages as mercenaries, a reduction that makes it possible for us to get a working model while only minimally adjusting what Gygax wrote.
The trick here is that Helgeran’s model applies to knightly rulers of small domains, but even compared to that, there are a huge number of warriors to account for here relative to population. The solution I have found is to use the oligarchy to create a tiered “democracy” of sorts. Erelhei-Cinlu is not a single domain, but rather a set of oligarchic domains.
The Garrison
The soldier-caste is an oligarchy of all soldiers for the city’s defense, supported by 3000 enslaved work gangs formally allegiant to the city but in practice overseen and managed by the soldiery. (3 HD drow have double shares; 4 HD drow have quadruple shares). All soldier-oligarchs have an obligation to serve in the garrison and fight in the city’s defense (since they are rulers, they can be mustered by a Call to Council, though removing more than half from their position will lead to rampant slave revolts or leave the city without garrison). The oligarchy earns 39,000gp per month, split across with 772 shares, which works out to just over 50gp/share — enough to cover the upkeep expenses of all soldier-oligarchs, and get them to their respective levels via campaign XP, but no higher.
The Orders Militant
The militant orders can be provided for by similar means, as independent oligarchies, with some slight adjustment. Of the female society, we are told:
This military sisterhood supposedly counters all other factions in direct service of Lolth. . . . The nobles furnish recruits, equip them, and pay all costs. The sorority is only nominally answerable to the noble house, but the noble family is responsible for any misconduct or misdeeds of the group it sponsors. Considerable rivalry does exist between the sororities, and there is intrigue and politics within the society.
And of the males likewise:
Just as the female organization is a neutral instrument of the drow clerical hierarchy, so too is the fraternity of male fighters an instrument to counter any destructive feuding or power seizure by one or more of the merchant clans at the expense of the others. Of course, the society also furnishes troops for guard duty, patrols, etc.
From this we can conclude that there seem to be direct payments from the noble houses and the merchant clans to the support of the female and male societies, respectively. Given that the military sisterhood is “in direct service to Lolth”, we can tentatively assign the noble support in coin sent their way as the tithes from the city (5000gp), plus the patronage of its officers (that is, we can set aside the wages due the 5th+ level leaders of the warrior sorority as being paid by the nobility to retain them in their entourages, reinforcing those allegiances and deepening the intrigue there). With that and a further 1300 slave gangs whose services they are granted by the nobility, they can support themselves as an oligarchy in the same model as the garrison, earning 21,000gp per month split across 438 shares. The slave “barracks” noted in the module can instead be characterized as holding cells for unruly slaves, which seems reasonable given that the 62 enslaved occupants are locked and chained in separate rooms while under guard. Incessant slave rebellion is inevitable given the ruthlessness of the drow, but with such a large and capable force to repress them it is unlikely to develop.
The male society is harder to account for because we have only very limited means by which to assess the wealth of the merchant clans. However, I have previously worked out approximate numbers for the volume of trade in a Class II city such as this, and here we will make use of them. I assessed about 715,072gp of goods passing through the city each month in trade; split across 16 merchant clans, that works out to each having control of 29,807gp of goods and a profit of 6033gp each month, assuming the clans control roughly two-thirds of local trade (they almost certainly have trade and profits elsewhere, but they will also have officers, troops, and other expenses in those places as well). Those numbers are supported by tallying the average value of goods on each of their properties: 28,655gp, within about 4% of my estimate.
Each merchant clan is led by a 7th level clanlord, aided by two 5th level officers in overseeing various troops under them with wages of 5340gp (they are paying full wages to their troops, as their troops do not hold the benefits of rule). If the clanlord personally manages 75% of that and his two officers split the remainder, each can cover his monthly wage and secure sufficient campaign XP to reach the appropriate level, as well as retaining in entourage their share of the officers of the fraternity (575gp per clan). That leaves each clan 3414gp in profits each month, a reasonable sum. If we work back from the total number of slaves we determined originally, 1897 slaves are left for the male order, which would necessitate a monthly tribute of 700gp per merchant clan in order for the warrior fraternity to have appropriate income. That is well within their means, and a very efficient way for them to purchase political influence (which might in turn be leveraged to acquire or deny monopolies, which could transform my assumed even distribution into something considerably lopsided; such is politics).
The Sovereigns
This leaves the noble house guards and the temple guards, which I would have to make rather expansive estimations about tribute outside of what is given in the text to afford. Some tribute is reasonable: given the troglodyte and bugbear guards fielded by each house, we can reverse engineer that each holds in vassalage 1182 troglodyte families and 940 bugbear families (with those household guards serving via a call to arms as the free duty owed). This yields an expected tribute of about 4473gp per month to each house. D1 implies a network of additional outposts, mines, and other vassals that will also send in tribute, though the scope is difficult to assess and unlikely to be a huge further income source. But while valuable, all this is insufficient to cover the myriad expenses tallied by the great houses of the drow.
However, it is very much worth noting that all of the noble houses and the temple guards are led by divine spellcasters of 9th+ level. To borrow a mechanic from ACKS 1e, if each 9th+ level divine spellcaster is granted roughly a platoon of fanatic followers who need only be paid basic upkeep, then these numbers work out very neatly. That particular mechanic was cut from ACKS II, but priestesses do receive unpaid attendants (and in considerably greater numbers), so contrasting the single platoon here against the six or so that the updated divine spellcaster would receive, it seems reasonable. A matriarch whose household guard is slain will of course find them irreplaceable, and be promptly assassinated so that her heir can attain 9th level in her stead to replenish them.
Given the 5000 drow urban families, the revenues of ruling the city total to 25,000gp per month. The tributes mentioned above add another 35,787gp monthly, bringing the total to 60,784gp. While by no means an insignificant sum, split among the eight noble houses and four high officers of the temple, that covers only about half of the upkeep of the least of the houses, and only a paltry fraction of that of the highest among them. That entire sum could support any single house, but hardly all of them together. On the upside, at 5063gp per oligarch and an average of 10th level, they have +1 Personal Authority for an expected base domain morale of +2 (because they will all have Leadership, and an even or positive distribution of Charisma). The Vault of the Drow is not in danger of an uprising among its lower classes, but rather of treachery among the great houses.
The upkeep expenses of the drow matriarchs themselves total 92,750gp per month (save the High Priestess of Lolth herself). Their male consorts, officers, and such add another 87,050gp. The female officer and priest wages add a further 70,850gp. All together that totals to 250,650gp, well above the total revenues of the entire domain, but not unreasonable at the scale of the city — since the revenues of rule are so divided, the revenues of other city officials must be as well. Magistrate positions, magic research, and passive investments into the merchant clans and businesses within the city can fully close this gap; it is quite a sizable city they rule and well able to support that level of investment. If they intended to make up that entire margin of 189,866gp via passive investments, that would imply a principal of about 6,328,867gp value in the domain; a rough approximation from the JJ suggests the value of the assets across the domain is closer to 30Mgp.
Theocracy
To return to the questions of demographics, the large number of high level divine casters is noteworthy but can be explained by a rotation of the rights to divine power among the oligarchic rulers, with each of the eight matriarchs and the four 9th+ level priestesses entitled to the full divine power of the city for one month each year (with them appointed chaplain to the proximate vassal oligarchies of the soldiers and the orders, to draw on their DP as well). The 21,915gp of DP that typically grants is enough, combined with their long lifespans, for them to quickly reach 9th-10th level, and after a very long time eventually reach 11th level, which is the distribution that we do indeed see in the city. It also highlights the disruption caused by the ongoing religious schism in the city: if some fraction of the populace has followed the Eilservs in their conversion to the worship of Elemental Evil, the DP available will be split, and neither faction will have a sufficient amount to advance (House Noquar in particular is likely to be furious over this, given their ruler is presently only 8th level).
The exception is the presence of the 14th level High Priestess, whose station cannot readily be explained. Demihuman realms don’t expect to probably have a 14th level character until they’re around a million families. Barbarian realms expect to have one around 55,000 families, which is closer, but still doesn’t predict one in this domain. Given that this is the High Priestess of Lolth, and there are other drow vaults alluded to, if we interpret her as having supreme authority across a multitude of vaults then her presence here can be justified as being the overarching divine spellcaster across what is effectively a large realm. Perhaps she also receives her own tribute payments in that capacity, or perhaps she simply lives at the upkeep of an 11th level character rather than a 14th level one.
Composite Domains
If one wanted a more generalizable template for drow domains, I might suggest the following:
Drow rule composite domains. A composite domain has two distinct components: a citadel of the drow, and a cavern of their slaves. Each will have separate domain morale. Because of their reliance on subterranean mycoculture, composite domains can only be established in caverns bathed in the luminous radiation upon which such mushrooms feed in the furthest depths of the world.
The citadel yields normal urban settlement revenues to the ruler of the composite domain, but does not have to pay for a garrison. The domain’s steward, chaplain, and munerator receive their typical wages based on the citadel’s revenues and expenses.
Large citadels typically fracture between competing factions into oligarchic rule: one noble house or grand priestess will emerge for every 500 drow families. If not given a share in rule, they will agitate against the ruler: each faction not participating in the oligarchy imposes a -1 penalty to domain morale rolls for the citadel and a -1 penalty to vassal Loyalty rolls. Any bandits or NPC leaders who emerge as a result of low domain morale should be treated as additional factions fighting to seize seats on the oligarchic council.
The slave-cavern contains the work gangs whose labor feeds and enriches their tyrannical masters. It has a base morale score of -4 rather than 0, takes a -6 penalty to domain morale rolls, and does not generate revenue. Instead, it provides a free garrison force that the ruler of the composite domain can deploy as desired (usually, it is left to repress the slave-cavern). This force will contain one 2 HD drow warrior per 10 work gangs, one 3 HD veteran drow warrior per 15 work gangs, and one 4 HD elite drow warrior per 150 work gangs, and is under the command of the composite domain’s captain of the guard (who earns 12.5% of their collective wages as his salary, as normal; if the composite domain is under oligarchic rule, multiple captains can be appointed, splitting this wage among them). If the number of work gangs decreases, any excess drow warriors will become raiders who plunder the weak nearby and aim to found their own composite domains or secure oligarchic rule.
The maximum number of work gangs in the slave-cavern is limited to one work gang per 20,000 cubic feet of space appropriate for mycoculture. If there are additional work gangs, the ruler of the composite domain can use them to excavate additional space (100 work gangs can dig out an additional 20,000 cubic feet in a month, as can a 1000gp construction project), establish a mine, etc.
In a typical composite domain, there will be the same number of drow families in the citadel settlement as there are work gangs in the slave-cavern. If there is not, for every full one hundred drow families in excess of the number of work gangs in the slave-cavern at the end of a month, the citadel loses 1d10 drow families. Likewise, for every full one hundred work gangs in excess of the number of drow families in the citadel at the end of a month, the citadel gains 1d10 drow families.
Any drow families that arrive as followers, via immigration from prestige, or from high domain morale will bring one enslaved work gang to the domain as well. The citadel settlement is treated as an elven domain for purposes of immigration from prestige.
Conclusions
I am very pleased to have found a way to make this model work; I quite literally have gone through half a dozen major iterations that almost work, or that work if I make a few unwise rulings on edge case mechanics, but I think this is actually broadly reasonable. If a player wanted to rule a composite domain as above, I’d be fine with that (though that doesn’t mean it’s easy to secure such a position).
Erelhei-Cinlu’s defenders total about 564 BR across 20 companies. Being all loose foot makes them somewhat fragile, and definitely superior on the offensive; their stealth and extreme sunlight sensitivity reinforce that. Cavalry can’t really be used in the Underdark, but on the surface they’ll suffer greatly if beaten on the field and subject to pursuit. If they took half their force on campaign, they’d be in the same league as a prince after accounting for their heroes, but with so few units and their deep internal tensions they could potentially be withstood by a strong duke with a band of PCs if he fought hard and smart — characters right about 10th level, which is what the module is recommended for. Very nice.
Incidentally, I found it interesting that the stats given for Lolth herself make her roughly equivalent to a speaking cacodemon of the eighth rank (a “devil”, with 16****** HD). A good reminder that with appropriate minions, support, and environment, reasonable threats can remain potent opposition well into the late game.
If anyone has thoughts on further refining this, or considers some of these jumps unreasonable, or would be interested in specific analysis of the array of monsters and treasures, I’d be interested to hear it. I am still working through all of the implications of it myself, but it fits the text and seems well-suited to creating the chaos, intrigue, and ruthless treachery that drow society has been depicted with. Let me know what you think!