Tuesday, August 23, 2022

Spectral Interrogatories V - Dwarrowdeep

Dwarrowdeep has Good Art
I don’t know Greg Gillespie, I don’t have any intense feeling for his work or his design history, as while I have looked at some of previous adventures they haven’t made much impression on me. I suspect he was on G+ and I may have spoken to him there, but I don’t remember anything specific from those long ago days. I purchased the Dwarrowdeep PDF (which costs as much as a good bottle of booze or nice steak - but cost isn’t a criteria I use to judge adventures) specifically to read it for this review and because of my overall interest in megadungeon design. I have not played it and don’t intend to. As always I will take a look at it for what it offers as a playable product and from my perspective as someone interested in dungeon and megadungeon design. I feel somewhat guilty about this review and don’t expect Mr. Gillespie will appreciate it much, but after spending a week reading Dwarrowdeep - time I could have used for paying work or on my own hobby projects - I can't play nice for the sake of comity, even if I'm not trying to be cruel. Darrowdeep is an abject failure, but in that offers a useful example of how not to design a megadungeon and asks questions about the limits of dungeon design.

Dwarrowdeep is the new (May 2022) 336 page megadungeon by Greg Gillespie, known as the writer of Barrowmaze and a few other big dungeons. This has made Gillespie a notable figure and dungeon designer from the early and middle years of the OSR. What Dwarrowdeep offers is a brand new megadungeon that I had hopes would highlighting the state of design in 2022 and the evolution of the form since the early days of the OSR or at least the 2010’s when Gillespie published Barrowmaze ... if it does, it only shows a loss of basic knowledge and the decline of OSR design and imagination over the past decade.

In light of its recent publication and Gillespie’s long career as a megadungeon designer I was hopeful that Dwarrowdeep would be something special, offering new ideas, or better utility - in short I was hopeful that it might have something to teach about megadungeon design. I was gravely disappointed. When I write reviews I try to be charitable and understand the author's goals rather than focus only on the work's failings, but Dwarrowdeep’s positive aspects are largely limited to the excellent art within, an audacious scope, and occasional moments when a decent idea shines through the mediocrity.

Dwarrowdeep fails as a megadungeon in three interrelated and key ways: variety, interactivity, and usability. I suspect this is the result of both ambitiously excessive scope and the poisonous idea that nostalgia alone is sufficient to produce good work. What I mean is that Dwarrowdeep doesn’t just try to provide a nostalgic aesthetic or feel, it goes deeper, with nostalgic layout choices, nostalgic key design, and nostalgic approach to setting (generally emulating the early 1980’s BECMI era TSR adventures). This fails, partially on its own merits, but partially because it’s so insistent on cleverly aping a particular, possibly imaginary, past that it ignore the work of other designers, both since the early 80’s and before. Dwarrowdeep drowns because it chooses to submerge itself in the nostalgia for a design that wasn't optimal even for the 30-page BECMI modules where it first appeared.

Dwarrowdeep begins by making the claim that it is “the largest dwarven-themed sandbox dungeon ever created”. It’s likely right, at least among published RPG products. The first thing I know that comes close is the 1993 TSR boxed set “Dragon Mountain'' which might also be described as a “the largest kobold-themed adventure path and sandbox dungeon ever created”; so I can’t quibble with Dwarrowdeep’s claims about scope. Another option might be the 1984 Middle Earth Role Playing "Moria - The Dwarven City", which uses far better design principles including Stonehell style one page regions and has very good maps - but is only 75 pages (the 1994 edition is 175 pages but that's down to the wordy style of the mid 90's) and is mostly inspirational reading. Dwarrowdeep is almost certianly the biggest faux-Moria ever published.

Dwarrowdeep isn't just dungoen, it contains a town, region, and that ruined dwarven city in the style of Tolkien’s Mines of Moria. In two important ways it fails all of these possible predecessors, it lacks the mood and intensity of Tolkien, despite slavish emulation (understandable), but also it fails to provide the basic elements of a functional megadungeon or even the clumsy utility of Dragon Mountain (incomprehensible). Dwarrowdeep bites off more than its author’s imagination can chew. The least of Dwarrowdeep’s problems is that it's dripping with cliche - while this annoys me, so many players love fantasy cliche so I won't even claim this is an issue for its aesthetics.  Vanilla or vernacular fantasy is a helpful and standard practice (though Anomalous Subsurface Environment proves it's not necessary) in megadungeon design because using a world where the implied setting of the rulebooks and the genre expectations of most players and referees can easily fill in blanks dramatically helps to shorten keys and streamline the overall adventure. Unfortunately Dwarrowdeep’s other design issues mean that it can’t or doesn’t take advantage of its choice to use the an exceptionally trite version of the lost dwarfhold cliche. It also errs in extending cliches to its physical design and mechanics -- which is a huge mistake for a work that isn't really a classic keyed megadungeon. 

The City Stealing Kind
of Dwarf
The city of Gundgathol (taken from Tolkien, another name for Moria and likely translated to something like “earth home/fortress") has been overrun and desecrated by evil dwarves from deep below - the duergar. Not the puckish creatures of Northern English folklore, but the gray dwarves of Dungeons & Dragons’ underdark. Like almost everything else in Dwarrowdeep’s setting, both its name and this background are generalizations, deeply tied to the expectations of Gygaxian fantasy and with an obvious relation back to Tolkien. There are genre standards working hard here to make Dwarrowdeep “Moria the Megadungeon”, with evil dwarves replacing some of the goblins and the Balrog. Importantly though, and despite some emulation, Dwarrowdeep does breaks with novelistic conventions to a degree, and is designed with some attention to the needs of dungeon crawling. Dwarrowdeep is largely the arena for room by room exploration that the classic dungeon crawl requires. This is also likely why the dungeon (unlike the literary Moria) is supported by a small region with some lightly sketched overland factions and a nearby home base for rest, hooks, rumors, new characters, and supplies. Unfortunately the other possibilities of this region, namely large scale battles and retaking Dwarrowdeep aren't supported. 


So at first glance Dwarrowdeep is very much a classic megadungeon, set on its regional map and designed to enable an “extended campaign lasting months or years” that rarely leaves the megadungeon. Surprisingly, it isn't a “world dungeon” -- a setting where there is only the megadungeon with no overworld -- and lacking that pretense it very much risks being quickly set aside for adventures on the regional map. Even with less tedious megadungeons, players often decide to venture into the wider world for various reasons.  With Dwarrowodeep this danger is much greater as its contents are so monotonous I can’t see any party (no matter how enthused by dwarves and Gygaxian fantasy) bothering with the dungeon for more than a few sessions. Finally, however classic or standard Darrowdeep’s physical design, it takes tentative steps towards the currently popular method of using procedural generation to fill in large areas of the dungeon map.

This isn’t to say it’s a “depth crawl” or a toolkit, it's procedural generation mostly uses techniques drawn from Gygax's D1 - Descent into the Depths of the Earth. Darrowdeep is a megadungeon in the early OSR style, akin to Stonehell, but far less innovative or enjoyable. 151 of Dwarrowdeep’s 233 page are keyed dungeon locations, densely packed onto two column pages in an unbroken torrent of mostly minimalist dungeon keys. Even with hundreds of keys, the majority of the dungeon Dwarrowdeep presents is unkeyed. Miles of passages are traversed using a system again drawn from Gygax’s Descent into the Depths of the Earth: randomly generated obstacles and encounters at a near wilderness travel scale that link the dungeon's few key regions and a larger number of geomorphic nodes that the referee must populate from a dozen or so random tables.

The content of the dungeon is precisely what anyone familiar with Moria would expect: assorted evil humanoids forted up in a ruined dwarven city of mines, tombs, defenses, and forges. There is some variation, and occasional glimpses of enjoyable novelty, but in general Dwarrowdeep remains aggressively nostalgic, supremely repetitive, locked into a vision of fantasy that is constrained, conservative in its aesthetics, and stifles design, usability, and playability. It's not the cliched nature of these locations that make them bland, its a larger lack of variation.  There are no overgrown fungus caverns, giant crumbling machines, towering apartment blocks, vast quarries, or flooded ore docks on the edge of underground seas -- only tightly packed, often largely symmetrical dwarf halls and fortresses with almost no elevation differences and winding mines or caverns of 10' passages. Variety is key to a megadungeon, and there's very little in Dwarrowdeep. 
The scope of Dwarrowdeep is its most notable aspect, an adventure with hundreds of pages of keys for large regions or nodes (many areas are over 100 keys), but Dwarrowdeep is so enormous that these spaces make up a relatively small portion of the mountain. Like Tolkien's Moria, Dwarrowdeep is largely empty. In what seems a poor choice, rather than creating a set of subsystems for moving quickly through its abandoned spaces, Dwarrowdeep choses to procedurally generate them. This choice seems especially odd given Dwarrowdeep’s enormous, impossible scope, because it falls short both in of the limited variety of content on the generation tables and because the idea of spending most of a campaign navigating empty and vaguely described spaces is singularly unappealing. I don't think the solution is simply a bigger set of keys - 900 pages of keyed Dwarrowdeep would add little and it’s perhaps a mercy that the Geomorphs and procedural generation are instead asked to do a great deal of work (though with them so is the referee).

All this makes the functionality, variety, and appeal of this likely necessary procedural generation a special concern. Parties will spend days of travel (at an unclear exploration scale on an undergorund hex map) between its keyed nodes on paths that generally pass through multiple geomorph generated spaces requiring room by room exploration. Campaign players will interact with the randomized contents of a few tables at the back of Dwarrowdeep far more often and for many more hours of play than they do with the keyed locations.


As weak as its mechanics and keys sometimes are, and I’ll take more time to look at them below, Dwarrowdeep is a polished product with workmanlike overall production, editing, and writing. Despite some emphasis on the currently popular practice of procedural dungeon generation, Dwarrowdeep’s production as well as its aesthetics, setting, and theme all feel extremely nostalgic. Considering this though, it’s very nicely produced for its chosen design style, a reminder of how far expectations for RPG layout and especially cartography (thanks Dyson) have come in the past ten years. Dwarrowdeep has solid, slightly stodgy, overall design that while not especially notable or remotely innovative … is a shockingly and obviously more attractive product than TSR’s box set “Dragon Mountain” (1993), which I noted above as Dwarrowdeep’s clear TSR predecessor and which was made by a larger team with the backing of a large publisher.

Dwarrowdeep’s overall production and design is pleasantly competent for higher quality indie RPGs in 2022, though its PDF lacks usability tools such as links and its organization choices are far from ideal. Dwarrowdeep's art however is of very high quality and very generous, with almost every page containing at least a few flourishes. While, like the rest of the adventure, the art tends towards predictable and generic fantasy images, it’s often extremely imaginative within those constraints. There are numerous artists, with many well known TSR-era illustrators providing at least a few pieces. Their professionalism and long association with fantasy art shows through.

In contrast to many large works with numerous artists Dwarrowdeep's art is rarely jarring, despite distinctive styles the overall quality of the pieces and the welcome limitations of black and white line means the work all comes together with a sort of classic fantasy RPG art sensibility. The art has a similar charm to the original Dungeon Master's Guide, and the single most enjoyable part of reading it for this review was seeing what the (consistent and provided as pregens) party of adventurers depicted on most pages would get into next. Dwarrowdeeps best art is much higher quality then these little vignettes, as the regional overview includes a full page piece by Darlene: a densely and finely inked woodland that feels fantastical because of its precise, crisp illustration and detail rather than its specific content. Darlene manages to evoke art nouveau illustration without being an obvious copy of any one artist. Equally impressive are a set of wonderful, often full page, illustrations of empty spaces within the ruined dwarf home: vast halls, crumbling lines of columns, vertiginous stairways, and bone scattered catacombs. All of these are done in bleak, heavily shadowed style and drawn with a real attention to line weight and use of negative spaces. These illustrations are at times nearly architectural, but also have a strong feel of Pirenesi’s Caceri — and like the 16th century prints, represent an inspiring creation of monumental fantastic space through perspective and line.

Sadly art alone isn’t the adventure, and alone can’t make thousands of very similar locations, poor usability, random generation tools insufficient for the adventure’s scale, or a toxic level of nostalgia into a functional adventure.

Dwarrowdeep's Best Art Isn't Available
Online. This Cyclopsman is still good

I can’t think of a good dwarven ruin-themed megadungeon, despite the importance of the Mines of Moria to the concept of the RPG dungeon (which is not a prison). The only one that comes to mind is the 1993 2nd edition TSR box set “Dragon Mountain”, which while well reviewed at the time, is a product of the post - Dragonlance era of TSR design -- meaning it is largely an adventure path and has not retained its popularity. This suggests that the ambition behind Dwarrowdeep isn’t entirely misplaced, it’s a creation aimed squarely at one of the principle literary sources of the “dungeon” and one that hasn’t been addressed in a definitive way. One struggle is to see how Dwarrowdeep’s somewhat glib embrace of the foundational dungeon cliche -- a generalized application of the themes and tropes of the dwarven ruin, could make for something more memorable than Dragon Mountain, let alone Moria.

Comparing Dwarrowdeep to its most obvious aesthetic source, Tolkien’s description of Moria, is illustrative. The Fellowship of the Ring’s Moria is a series of disjointed scenes, which works well as a fictional device to capture the discombobulation, vastness, and ominous threat of a 40 mile trip through a pitch black, demon haunted ruin. Obviously this sort of technique doesn’t make for good classic dungeon design. First, despite its impact, Tolkien’s Moria is minimally described. All the reader learns of know of it are its magic door, branching tunnels, an ominous deep well, rusty chains, flooded halls below, a great hall for Gimili’s lament, Balin’s tomb, and the Bridge of Khazad-dum. A dungeon crawl needs to allow for exploring a fantastical space, not just provide a few evocative images.

Second, most of the power of the two chapters that document Moria come from the stories that the fellowship tells or discovers about the place. Tolkien He spends as much time describing an orc chief that stabs Frodo as he does on Moria’s great hall of pillars and glassy black walls. Instead we get song and metaphor as well as the grim account of Balin’s failed expedition. It’s all powerful stuff for fiction, building threat and oppressive darkness -- but again dungeon crawling RPG adventures aren’t about using scenes to tell a story and characterization (Gandalf’s obvious fear, Boromir’s foolish and ineffectual bravery, Gimili moved to poetry by sadness, or Pippin’s realization that his shenanigans threaten horrific consequences) to create an emotional response. RPG dungeon crawls are a game of navigating fictional space through interacting with the referee's (and designer’s) clear description. Tolkien may offer a set of feelings and a few pithy descriptions - but it's up to Greg Gillespie to turn them into usable adventure content.

Instead, the focus on fictional Moria makes Dwarrowdeep stumble as a usable adventure. Tolkien’s Moria is a journey through an enormous and largely abandoned ruin. The Fellowships sneaks through the dark above the flooded, orc infested, and Balrog ruled halls below. Tolkien’s journey is a series of scenes, impressions, and fragments that use the vast empty darkness and the power of the novel to skip or compress most of the dungeon into a few words without concern that the novel’s characters will want to investigate something or wander the wrong direction. This doesn’t work in a classic dungeon crawl, at least not without significant mechanics for faster travel through empty halls.

As a referee you have a few hours to play a game about exploring a location, discovering its secrets and interacting with its inhabitants, but you lack a novelist’s tools to skip time and direct exploration. Turnkeeping, navigation, and player choice are primary elements of classic dungeon adventures and they often work against narrative story-telling. For a Contemporary Traditional game where exploration and navigation are a function of a few stat checks, one could follow Tolkien’s concept of Moria as an empty vastness, and even his narrative structure, more directly.  5E design allows for a series of exciting scenes linked by quick referee narratives because it mostly depends on set-piece combat encounters where 5E’s player choice largely occurs. Since this understanding of play is not a part of the classic dungeon crawl, certainly not as Dwarrowdeep understands it, silent and empty halls stretching for days beneath the mountain offer only monotony not Tolkien's forbidding. 

The gameable content of Dwarrowdeep is no more novel than its overall theme. At the start Dwarrowdeep's introduction certainly isn’t memorable or inspiring, but at least it’s only six pages.

“The/some/these Dwarves were chased from their mountain city by exiled bad dwarves”. Presumably this makes them different than their neighbors who suffered a balrog infestation, or was it a dragon - dwarves are irresponsible property owners. Now they want it back. They don’t have the army to take it. They will have to depend on heavily armed tomb robbers and desperadoes… That’s your overall hook.

There is some history, but it doesn’t matter much; it’s certainly neither complex or unexpected enough to act as the basis of puzzles and secrets in the larger adventure: it's the opposite of what we see in Tolkien’s poetic descriptions of Moria and something that is generally quite important to maintaining freshness in a megadungeon. Otherwise Dwarrowdeep’s regional setting has everything one would expect from Northern European based generic fantasy that clings closely to Gygax and Tolkien's pants legs. The ‘Deep’ is located in a feudal type pocket polity surrounded by mountains and beset by humanoid raiders. Knights are implied, merchants mentioned, there’s a free city, and a land full of battle sites, ruined forts, and trade routes. Everything is of strategic importance and/or relates to some past conflict - though most aren’t especially important to the megadungeon itself. They are perfectly acceptable as a small fantasy region, but also deeply and sadly predictable. I will dub these lands the “Dells of Lost Opportunity” Dwarrowdeep calls them something less memorable.

I tend to be snide about this sort of standard Gygaxian vernacular fantasy setting, and where Dwarrowdeep deviates from cliche it does manage to offer something moderately memorable, such as its band of otherworldly cyclops led by a blind cleric who can travel between megalithic stone circles. This is an interesting piece of regional world building, offering a race of potential characters, a regional faction and a means of ‘fast travel’ to discover. There are few other aspects of interesting subversion or transformation of the obvious cliches, but overall Dwarrowdeep is bland and more disappointing. Its overworld factions, even when theoretically attached to the megadungeon (the dwarven exiles), don't have much in the way of specific connections or goals.

Again this first section or impression is typical of almost every aspect of Dwarrowdeep, it manages to include interesting or useful elements but just leaves them floating, without support or connection to the adventure — marooned in a sea of dwarf based cliches. Remember how I mentioned Tolkien’s Moria is described in two short chapters, maybe a 100th the length of Dwarrowdeep? Yet, while the literary Moria lacks physical description, architecture, or design, it still contains plots, secrets, and stories. Most are implied and unanswered in Tolkien, for example Durin’s crown lost in the flooded deeps, one star among a subterranean nebula of jewels (or eyes), sounds like the basis of an adventure hook. If only some dwarf followed Gimili’s example and related this mystery to the characters in Dwarrowdeep. The story of a powerful lost artifact has potential.

Dwarrowdeep may even have a lost magic artifact, but there are no such allegorical rumors in Dwarrowdeep, and this is where one begins to notice that the generic nature of the adventure is less an issue than the fact that this blandness is piled high to conceal a lack of basic coherence and the complexity necessary for a megadungeon. I may dislike generic fantasy settings, and they do produce problems because they reduce player engagement and must adhere to player expectations to avoid feeling unfair, but even the most generic of settings can still use hooks, rumors, and NPCs effectively.

“But still the sunken stars appear
In dark and windless Mirrormere;
There lies his crown in water deep,
Till Durin wakes again from sleep

This is the last line of “The Song of Durin” that I’ve been talking about, the lament Gimili sings when the Fellowship enters the black mirror polished great hall of Moria. It’s a damn effective rumor.

“The Eternal Mountains are known as Thaneduhr’s Throne to the dwarves.”

This is a rumor in Dwarrowdeep. I’m not going to demand that RPG rumors have the poetics of Tolkien, but the verse offers more usable, gameable content than the Dwarrowdeep rumor. Give the players the name of a location within the dungeon “Mirrormere”, a magic crown and vague implication that its recovery will awaken a slumbering Demi-god. Don’t tell them a random useless factoid. Dwarrowdeep's rumor is especially infuriating because it is one of exactly seven rumors (on a table of 20) that relate even tangentially to a 1,000 key megadungeon and none mention the lost artifact. Nor do any of the NPCs offered provide much in the way of hooks.

There is no need for this sort of cursory, blithe design - it’s not as if Dwarrowdeep lacks the page length to offer decent rumors or functional hooks, it’s that a conscious choice has been made to spend about ten pages on Dwarf lore and town description but provide almost nothing playable or inspiring: Dwarves value their beards, they like drinking, they use runes and have Scots accents, they honor clan and family etc. One wonders, what's the purpose of pages of vernacular fantasy pastiche, especially if it lacks directly gameable content, and why offer it instead of actionable hooks and rumors? This is the real problem with Dwarrowdeep’s nostalgic aesthetic, it’s not used to fit more content into a smaller space, but rather to pad space because there’s not enough creative content to match the adventure’s pretense of scope.

John Blanche Drew These Dwarves
in 1985. They have flavor.
The keyed areas of the adventure are little better, they have exactly one story to tell - the same predictable tale of a dwarven city seized from below in a war with evil deep dwarves and now home to tribes of slaver orcs and other humanoids. Large keyed maps are provided for the entrance areas and several key nodes within the mountain but the keying is repetitive even within the context of the setting.

There is minimal variation, a frozen tower entrance high in the mountains that is the lair of a powerful white dragon. For the most part though, even once the party delves deep into the earth, Dwarrowdeep’s keyed areas consist of humanoid outposts and lairs: orcs, stronger black orcs, derro, kuo-toa, and dark dwarves. In each, themes repeat in unimaginative ways - guard rooms, and ruined temples, rooms with rot grub infested corpses whose stench causes vomiting, false tombs and slightly better hidden true tombs of various dwarven heroes clutching magic weapons and piled runestones which contain meaningless inscriptions. Little changes between these locations except the complexion and size of the evil not-people one is fighting. Again, Dwarrowdeep is huge, the entire dungeon might be a five year campaign from levels 1 to 10. It would be an excruciating one, something close to playing one of the “gold box” Dungeons & Dragons videogames of the 90’s where every area was a series of barely distinguishable corridors and intermittent combat encounters. This moves beyond simply a bland aesthetic into the larger issues that Dwarrowdeep has - even where it’s keyed and mapped it’s not designed as a useful megadungeon.

What makes a decent megadungeon? There are a number of answers to this: maps, levels, hundreds of keys … reading through years of attempts though I think megadungeons work when they can hold interest over many sessions. What’s the point of a huge dungeon if the players refuse to continue adventuring in it after a few weeks?

Comparing Dwarrowdeep to the best (almost) megadungeon I’ve reviewed here: Caverns of Thracia, it’s obvious that it contains several things that Dwarrowdeep profoundly lacks, despite Dwarrowdeep’s far grander scale. Variation, secrets, and faction Intrigue. There’s only so many rooms of fallen stones, orc outposts, and dwarf skeletons that a party can explore. Reading through Dwarrowdeep it’s clear that nostalgic embrace of aesthetic or setting alone is insufficient for Dwarrowdeep’s scope - technique is needed.

This is why attempting to emulate the feel of Tolkien’s Moria without access to the novelist’s tools creates grave problems. Beyond an unconsidered embrace of Tolkien, a greater failure of Dwarrowdeep’s design is that it seems to refuse or reject the idea it must replace Fellowship of the Ring’s narrative tools with something appropriate to its medium and play style. Instead, Dwarrowdeep doubles down on nostalgia -- not only as setting and aesthetic, but in its adventure design. Nostalgia has risks, it elevates the design or methods of the past in search of an idealized feeling or representation of memory. Part of nostalgia becoming more than a personal inspiration is a claim that artifacts or representations of the past actually represent the nostalgic ideal. If you insist something from the 1940’s is the perfect fantasy story or an adventure from 1982 the perfect dungeon design, you can’t change, adapt or modernize. Nostalgia, once it becomes ideological instead of a personal memory, traps the nostalgic. Worse is the way this sort of nostalgic worship of an iconic work of image prevents deep examination of that past. Once we declare something the ideal, other lines of thought and products of the same era must be ignored where they contradict or offer alternatives. Dwarrowdeep is trapped. Being trapped in recreating Moria might be great - Tolkien’s two chapters are quite inspirational still, but being trapped in trying to do this with the design choices of BECMI is a disaster.

Dwarrowdeep can’t or hasn’t learned the lessons that even Dragon Mountain has about megadungeon design. It’s firmly rooted in the design of early 1980’s TSR, just the time when the free form and experimentation of Gygax and other original designers began to drain away, but before the tyranny of the storypath became entirely dominant. Dwarrowdeep’s keying and structure greatly resemble an adventure like Horror on the Hill, or perhaps the drudgery of King’s Festival minus the box text. In picking this moment of dungeon design to emulate, Dwarrowdeep largely refuses to learn from either the earliest adventures, or the adventures such as Stonehell or Anomalous Subsurface Environment that were produced later by the OSR. Dwarrowdeep isn’t entirely limited to this design period, and Gillespie occasionally escapes from its gravity, especially when he attempts to apply the underdark travel of Gygax’s D series, but the keying style and adventure sensibilities of Mid-TSR remain dominant and detrimental

To put it another way, both Lost Caverns of Thracia and Stonehell are far more innovative than Dwarrowdeep, because they make informed and inspired attempts to produce a playable megadungeon rather than what seem like empty gestures in the direction of a megadungeon. A project of Dwarrowdeep’s scope is unlikely to succeed everywhere, but putting aside its excellent art it’s fascinating that Dwarrowdeep succeeds almost nowhere -- and largely to the degree that it cautiously apes the mechanics and style of past works. As a dungeon Dwarrowdeep fails to provide faction intrigue, useful keying, or the sorts of mysteries and larger secrets that would justify considering it for a long campaign.

The first and best way to maintain interest and challenge in a large or small dungeon environment is to offer memorable factions and NPCs. I don’t mean using funny voices, but including groups and powerful individuals in the dungeon whose interests align partially with the party and who are open to interaction other than combat. Dwarrowdeep notes that it contains “Two factions, very broadly defined”. The Duergar and their massive alliance of: orc, black orc, derro, kuo-toa, and trolls make up the major faction while the other consists of a few scattered individuals such as the white dragon and small groups like a pack of gem mining Xorns, or randomly encountered fungus people and troglodytes. The iproblem is that none of these lesser "factions" make for much in the way of allies against the overwhelming (and monotonously keyed) might of the duergar. The large number of foes in Dwarrowdeep mean that not only will battles be frequent, but they will be large and very time consuming. The non-allied groups are too small to matter much and ones that could be useful such as the dragon or Xorn aren’t provided with any sort of relationships or goals that would allow players to negotiate with them meaningfully. The party will be fighting these battles alone, regardless of who one characterizes the "factions" in Dwarrowdeep.

It's worth pointing out that the duergar alliance could easily be a potential source of faction intrigue, but it’s not. Despite a table showing that some of the elements within it are hostile to each other there’s nothing to indicate how its various parts can be pried apart or what the entire coalition will do when faced with effective opposition. What will it take to turn the trolls against the duergar? We don’t know. The two headed troll chieftain is just a stat-line without desires, goals, fears, schemes or a personality.
The Missing Exile Army?
Dwarf Wars - John Blanche Again

A third faction potentially exists in the form of the overland groups - especially the exile dwarves who wish to return home. Ultimately Dwarrowdeep abandons this plot line though. As noted above the town of Hamelet is exhaustively detailed, down to the predictable customs and trite “dwarfied” slang. Every major NOC in the town, and many minor ones are provided -- but like the rumor table there’s nothing there to motivate play, and no hooks to engage in faction intrigue.

Several times in the adventure notes that the exile dwarves send raiding patrols into Dwarrowdeep but besides a few dead and the occasional captive this doesn’t have an impact or a meaningful model in the adventure. Let the players earn the dwarves trust to lead these patrols, offer a way for the exile dwarves to gain confidence and start launching larger expeditions. A simple faction chart showing resources and a list of misisons/goals would do wonders here. Even the numerous dwarven slaves within the mountain lack the potential to become a meaningful opposition to the duergar. The only groups of powerful dwarven captives (4th - 6th level fighters - still not much against 4th level duergar hordes) are in the last few areas, the fortresses and storengholds - meaning there's no option for them to be freed and return to the mountain for vengence, properly equipped and healed. Dwarrowdeep fail to give sufficient information about the order of battle for any exile dwarven faction or provide them with goals and motivations, but even with these additions and the other's I've suggested, the hit dice and ability differences between the exile dwarves and duergar are simply too great for any exile force to have a meaningful chance of recovering or holding any part of the mountain. It's deeply unfortunate that not even the slightest attention was paid to designing Dwarrowdeep around the possibility that the exile Dwarves or perhaps some force within the dungeon could raise a significant armed group to challenge the duergar in battle.

This is a shame, because without faction intrigue, base building/reclamation, or larger scale battles the fundamentally repetitive nature of Dwarrowdeep’s design is even more clear - it can’t be avoided - every delve into the underhome must be more or less the same assault on a humanoid garrison by the party.  Again, one can’t really run a Moria campaign with Dwarrowdeep, because there’s no way to take and hold space within -- Balin’s expedition is impossible. The adventure has to be resolved (even as the party approaches 10th level) through skirmishes between the party and the humanoid groups within Dwarrowdeep. 

Some will push back here (and on the idea that the Hammer of the Dwarven Kings should have greater ability to resolve the megadungeon) out of the misguided idea that offering these possibility of a conclusion is inserting story or a railroad into a sand box adventure - it's not, it's providing possibilities for players to create a story - not setting up a series of inevitable scenes. This distinction makes all the difference. Like a regional setting, a megadungeon works best as a place full of coiled potential: tight balances of power and ways for the characters to push static situations into chaos and activity. The idea of faction intrigue as a secondary mode of play in large adventures is as old as the hobby, and that a campaign can culminate in mass combat is even older, despite mass combat having been set aside for much of the TSR era. 

In contrast to Dwarrowdeep, Dragon Mountain (which is again not an amazing megadungeon by any standard) the humanoid population is largely made up of varied color coded kobold bands - their territories explicitly marked on the map and their hostilities documented. The party can fight, negotiate, trick, and play the humanoid factions against each other to make their way to their dragon overlord, and this produces a greater variety of play with more options, despite Dragon Mountain’s somewhat more limited selection of monsters and smaller size. With adventures as old as Dragon Mountain, Caverns of Thracia and Vault of the Drow as examples Dwarrowdeeps static, unambitious groups of humanoids both in and outside the dungeon are a travesty. 


The reason factions are important is because they allow and encourage player choice and engagement with the adventure or setting. The larger the adventure, the more important it is to encourage player involvement, to provide both immediate and multi-session goals that function as small victories or moments of excitement and wonder. Dwarrowdeep’s conscious decision to use a well known aesthetic and cliched description almost prevents it from creating wonder as the players are very likely to already know exactly what everything looks like within the dungeon, but it leaves plenty of space for engaging players with excitement, and minor victories.

Goals are best when they are self-generated, players will get more enjoyment out of accomplishing some sort of quest or fulfilling some sort of scheme they set for themselves then completing one provided by an NPCs. Of course without information players can't develop their own plans so, hooks and rumors are used provide starting motivation. Dwarrowdeep doesn’t have effective hooks or rumors, just the vague idea that maybe the party could explore its enormous dungeon where there is treasure, and some exiled dwarves would be happy if they reclaimed it somehow. To repeat a theme in this review … it lacks the rumor of Durin’s magic crown drowned in the depths.

However, hooks and rumors aren’t the only way to produce engagement just a first step -- the dungeon itself can give clues and signs of larger mysteries or smaller secrets. At the most simplistic level this comes down to keying - clues to a secret treasure cache and the like (which Dwarrowdeep manages occasionally). Especially in an adventure the size of Dwarrowdeep, larger mysteries or potential events and escalations are beneficial, and Dwarrowdeep even offers one or two possibilities at the grandest scale - while providing almost no means to utilize them in play. For example Dwarrowdeep repeatedly notes that the duergar and exile dwarves are one people, separated by a religious schism, the players can even figure it out, and there's a leader among the exile dwarves who secretly worships the duergar god. Yet nothing is made of this. There’s no corresponding faction among the duergar who want to reconcile with the exile dwarves, or otherwise interact with the overworld. The secret relationship between the exiled dwarves and the duergar is empty “lore”, with no effect on play. Why not offer the possibility to the players that by brokering an alliance (and likely wiping out detractors) they can become the rich benefactors (or founders even if they are dwarves) of a mostly evil dwarf hold?  That's a big way to move to domain play.

A slightly more defined set of secrets and a long-term goal is the idea that the party should “Seek the Hammer of the Dwarvish Lords” - a phrase repeated throughout the adventure by various sources, including some of the more friendly ancestral spirits the party will discover.

It means almost nothing. 

The hammer and the Forge of Creation are within Dwarrowdeep, but they are simply a powerful weapon, they may improve the PCs abilities to wreak havoc on the Duergar, but the hammer doesn’t have any other secret to it, it offers no resolution or interesting shift in the relation of the party to the mountain, either group of dwarves or anything else. In Dwarrowdeep a hammer is just a hammer, no matter how much various dwarf ghosts insist otherwise.

Even At the final confrontation with the duergar - their arena and kingly halls, Dwarrowdeep provides no support and almost no possibility for interaction or climax beyond a series of ever more difficult combat encounters. The only results from finally discovering the duergar stronghold that the adventure supports are: its destruction, death of the party, or the party’s enslavement. No secret or mystery will give the party a great advantage or provide alternatives. As mentioned, the discovery of the hidden Dwarven Forge of Creation and Hammer of the Dwarvish Lords -- repeatedly held up as the goal of the campaign and solution to the duergar problem -- doesn’t offer that much.  The artifact is a powerful, but not overpowering, weapon and it will not really change the balance of power between a 10th level party and the armies of the dark dwarves: they won’t run from it, offer to trade it for the mountain home, or recognize the wielder as King, they’ll just take more damage from being hit by it.

At least the keying of final duregar redoubt considers the obvious possibility that characters may be captured by the duergar and most likely thrown into the arena as gladiators. Sadly while capture is suggested, the arena is only hinted at by implication and almost nothing is provided to support this possibility or that of escape from duergar captivity. 

Like so much of Dwarrowdeep, finding this artifact is a very empty sort of triumph.

It's not just the grand mysteries and campaign goals that are missing. Less important secrets are also omitted, though the basis for a few is provided in the indulgent (because it has almost no impact on play) introductory section. For example, pages are spent on dwarven culture - bland stuff of course, derivative and likely easy enough for anyone who has any association with the fantasy genre to make up. This information though could matter, it could be the source of solutions to puzzles and secrets within Dwarrowdeep, and at least in one case it is, but mostly it's just meaningless fluff.

The existing use of secrets does more to prove this argument then improve the adventure. Dwarrowdeep obviously knows how clues, secrets, and mysteries sort work, because it includes a set of dwarven mine markings that can be decoded or discovered and are occasionally noted in the text, most often warnings against upcoming hazards. It's a secret language the players can learn to help them explore - precisely the sort of little triumph that most players find exhilerating.

There's just that one though, and there are so many more similar possibilities, even in the sort of standard fantasy background Dwarrowdeep revels in. As one example: the introduction also spends ½ a page on dwarven beards and their cultural significance … it’s just not useful information.

It could be.

Dwarrowdeep could have its dwarven ghosts react differently based on character facial hair, making party grooming an important decision, or  it might include puzzles with solutions based on recognizing beard styles and their meaning in carvings, tomb effigies and such. "Pop the lid of the tomb with the forked bearded dwarf on it - it's the sign of the gem makers - avoid the one with the four plaits tied in knots, that dwarf was an apostate of the balrog possession cult."

These seemingly conscious design decisions: to pad the keys and setting of Dwarrowdeep without connecting its elements or incorporating its mysteries in an actionable way are inexplicable. The tools exist within the adventure, but Dwarrowdeep only provides secrets in the most basic forms: a key to a secret door ten rooms away or a vague suggestion that there might be treasure somewhere far away. These kinds of secrets aren’t really something to seek out, and information about the mountain's past, or its mysteries isn’t ever especially valuable or accessible to the players. Ultimately Dwarrowdeep is simply rooms filled with foes, the occasional ally or treasure, and traps. There is little connection for players to discover between its numerous humanoid lairs, and no point in doing so anyways because no advantage is gained, no mystery solved, curiosity and discovery just provide a few random bits of useless information that don’t change the course of the adventure. One might say Dwarrowdeep passively punishes intelligent play and engagement by making it a waste of time.

The individual keyed locations to Dwarrowdeep are likely its strongest traditional element, but they are still marked by 1980’s TSR design conventions, especially those of the early BECMI era, and they include some of its pecularities that work against utility. The keys of Dwarrowdeep, like those of the adventures it models, often stumble when it comes to interactivity even within the context of minimalist key design, and finally, at least partially because of the adventure’s scope and limited number of environments and challenges, the keys quickly become very repetitive - the same ideas are recycled almost every level or node of the dungeon.

Keying and interactivity is still an area of discussion in adventure design, but in the past fifteen years or so a list of (admittedly contested) maxims have solidified around what makes a good or bad key. At the simplest and hopefully non-controversial level a good key is one that is both easy to use in play and inspiring. Perhaps this can be described as the tension between brevity and evocative detail -- A dungeon should be the opposite of this review.

The basic structure of dungeon crawl style play (at least when it’s not aiming to be boardgame like) is: referee description of the environment, player questions about aspects of that description, referee response, and player decisions about how to act.

Dungeon keys are useful to this structure as they provide a basis for an environment's description by the referee and ideally details to answer the players’ most likely follow up questions. It’s best to do this efficiently, and largely Dwarrowdeep is often efficient at providing basic descriptions, but it starts to struggle in offering details that the players can interact with after that basic description.

There’s also typical minor issues with the keying that were especially prevalent in TSR products of the 1980’s.

For example there’re keys that include inaccessible information or history:

“Beholding this chamber during the Golden Age of Gundgathol would have been a majestic sight: dwarven guards at their posts, the hustle and bustle of mountainfolk going about their daily routine, and caravans slowly making their way to trade with Hamelet.”

“The stone door to this chamber lies in pieces on the ground. Centuries ago the duergar broke into this chamber with the hope of penetrating the inner crypt. Two sentinels destroyed many of their warriors before they were defeated. The cost was so great they cursed the place in Dworgrim’s name and refused to return.”

Similarly there are a few keys that ascribe emotions to the characters who enter the space:

“The Hall of Stone is sublime. The long vertical lines of the pillars stretch upward and disappear into the darkness while reducing the agency of those who behold it. To some the feeling of being dwarfed by the sheer enormity of the hall is akin to a spiritual experience - specifically feeling the presence of Thaneduhr or Dhurindain in Gundgathol.”

The use of inaccessible historical information and even keys that impose an emotional response on the characters (unless it’s a magic effect - and it’s somewhat unclear above) are minor sorts of errors, but ones very typical of BECMI era dungeon design, as is the major issue with Dwarrowdeep’s keying -- it’s designed almost entirely for combat encounters.

Most of the keys in the adventure are very short, almost in the laconic style of Gygax, except where they include either monster stats (or long lists of gemstone treasures) and, to Gillespie’s credit, monster tactics. Tactics and stats are are good, and it’s one of the few consistent elements in the adventure’s keying that’s a solid positive. Monster tactics promote more interactive, more complex, and more interesting combat because they offer the referee ideas on when and how various creatures fight.

Unfortunately the same rigor hasn’t been applied to non-combat obstacles, or even monster goals and personalities, meaning that just as with its aesthetics and the overall structure of its regions, Dwarrowdeep creates monotony. The specific issue that most of the encounters in Dwarrowdeep are almost inevitably combat encounters (or at least they appear designed with that expectation) is tied to its problems with faction intrigue, but its lack of variety and interactivity in non-combat experiences is a separate and unique issue.

Dwarrowdeep also has traps in it - admittedly not a surprise. These aren’t the complex puzzle traps of Tomb of Horrors, but either simple mechanical or unexplained magical traps. I call these sort of traps  “hallway traps” from the ones used by OD&D's Underworld & Wilderness Adventures.  They are effectively incidental taxes on character HP for moving through the dungeon or interacting with treasure, as almost universally they do damage when a PC passes a certain point, opens a door, or messes with chest. The party will encounter a lot of chests with poison needle traps, and a fair number of carved mouths (bearded of course - but not with a beard that notably denotes danger - see it could be so easy!) that breathe things or cast spells. Notably even these simple traps, and the secret doors of Dwarrowdeep, are not interactive in the sense of offering puzzles for player ingenuity to solve -- they don’t give meaningful clues about how to disarm or discover them beyond the rules based solutions of thieves’ skills or search rolls. This is unfortunate, because once again traps are interactive traps and puzzles are a way for Dwarrowdeep to offer more than combat with increasingly powerful funny looking not-people in stone corridors. Even hallway traps, are an opportunity to provide some small measure of interactivity for players, but to do so they have to offer clues that they exist and sufficent description for specific player solutions (cut the tripwire with a spear, jam the acid sprayer with a rock etc).  Again Dwarrowdeep choses not to bother, to just throw things back to the mechanics of BECMI and players exploring it will still have little to look forward to besides endlessly "rolling for initiative".

Moreover, the few more puzzle-like obstacles that a party going through more than one of Dwarrowdeep's regions will discover are repeated often: all important tombs are false tombs with a nearby haunted true tomb and almost every corpse has rot grubs in it waiting to punish the curious. At least the monsters’ captives aren’t all ready to betray the party in a parody of B2 Keep on the Borderlands (though more should betray the party given the number of captives the party can encounter, because after 250 years some of those captive dwarves have gone over to the duergar). This is emblematic of one of the major issues in Dwarrowdeep is meant to be played beyond a few sessions.  Dwarrowdeep doesn’t improve or change, it doesn’t get more complicated or add new challenges. The only real difference between its entrances and its urban core are that the foes in later humanoid forts have higher HD, more magic, and are sometimes better organized.  Despite these siege regions offering the only potential variation in challenge design, Dwarrowdeep still won't offer an order of battle and means that if the duergar, derro or fishmen (whatever they are called) sound the alarm it will be considerable referee effort to count up the surviving foes and decide how they act without assistance from the dungeon.

I’ve saved the most unfortunate, but most understandable, failure of Dwarrowdeep for last. While there is no excuse for a lack of factions, bloated and repetitive keying, or a lack of secrets and mysteries that create connectivity and player engagement -- certainly no excuse in 2022 -- difficulties in any attempt to use procedural generation for an enormous dungeon are understandable. Procedurally generating dungeons is a hard trick to pull off, and one that designers have been trying with very limited success since the 1970’s.

Gygax’s first large article on D&D in Strategic Review is a set of tables for producing dungeons aimed at solo play. It’s largely repeated (and expanded on) in the tables at the back of the 1st edition Dungeon Master’s Guide. I think these tables are some of Gygax’s best pure game design work, and they remains relevant and interesting as a model. Like most of Gygax’s design they have a level of unnecessary complexity, but they largely succeed and their excesses are entirely forgivable as they represents a bold experiment in a new sort of art form. Procedural generation has also been a major trend in more recent designs. Most notably the work of Emmy Allen, such as Garden’s of Ynn, which generate highly specific locations. There have been many other interesting experiments in procedural generation in the past few years, including the cavern and underground generation tools in Patrick Stuart’s Veins of the Earth, which are very applicable to Dwarrowdeep. Dwarrowdeep though opts for techniques that are far closer to the earliest attempts at procedural generation.

Randomization is an essential part of location based RPG design - the dice can fill in spaces when players inevitably ask questions or make decisions that the referee or designer haven’t written down or even thought of. Additionally the ”oracular power of dice” and faith in random tables is one of the cornerstones of the OSR play style -- the idea that randomly generated content will be more interesting and unexpected, or at least serve as inspiration through surprising juxtapositions. However, there are risks associated with procedural and random generation, from treasure tables to producing entire dungeons: waste and repetition, and both are something found in Dwarrowdeep.

If the basic idea behind procedural generation is to prepare a set of tables that alone or together fill in unknown or lightly sketched areas of the adventure and so save the designer physical and mental space, these two risks should be obvious. The random generation may not actually save space, or it may not adequately fill in unknowns. Waste is easier to point to and endemic in adventure design, especially when the designer doesn’t really grasp the play or design style of older games. Repetition is less of a risk and tends to afflict larger products such as megadungeons or the implied setting produced by rulebooks.

Designers often use random tables in an adventure when they don’t need to: the random table fills more space then the content it would generate, with no advantage. This kind of table abuse wastes valuable pages, forces the referee to stop and roll, and offers less opportunity to provide information or theming for the adventure. For example, in one well-known recent adventure there’s a six entry table for a single treasure that the party has a chance to discover if they dig up the floor of a cellar. Why? It’s not as if the players will be digging up multiple cellars or there are multiple treasures to find. It’s a waste of everyone’s time, and this treasure could be a much stronger part of play if it was a specific thing that related to something else nearby or provided some kind of clue to some other aspect of the adventure.

Dwarrowdeep doesn’t quite do this, it lacks the level of detail, but it does randomize things that don’t need to be randomized.  For example a keyed location's cursed altar has a random effect from a table in the back when a specific and thematic effect would be far more evocative and usable in play. The most serious and obvious example is Dwarrowdeep's large scale hex map. Covered in waterfalls, elevation changes, and bridges where the condition, type and danger of each is randomly determined. The map is two pages --big enough to include symbols denoting which of the handful of waterfall types exists at each elevation change with an icon, rather than forcing the referee to randomly generate one, with adjustments for elevation change and then mark it on the map in case the party uses the route again.

This sort of waste is largely an annoyance, at worst a sign of disinterest, or unthinking emulation of Gygax.  Yet Gygax's obsession with creating subsystems for use in an entire fantasy world rather than a specific adventure, it makes much more sense.

On the other hand though, a lack of content for procedural spaces presents a grievous problem for Dwarrowdeep. While there are hundreds of keys covering its six entrances and seven “primary encounter tables” the majority of the adventure is a “hex crawl”; alternatively given that it involves moving along specific passages between nodes it has a lot in common with a “point crawl” as well. This portion of the adventure is covered in somewhat less then twenty pages and includes not only miles of tunnels and rivers, but also 30 or so “Secondary Encounter Areas” which the referee must create with a set of geomorphs and stocks with tables provided in the back of the book. These regions are largely empty (noted above as a problem because they will take almost as much exploration time as populated, keyed locations). The issue is compounded and rest firmly on the procedural generation tools Dwarrowdeep offers  because it provides only a comparatively small set of tables to fill all of these large nodes -- some several levels tall and most containing at least 30 rooms.

The referee effort for this sort of procedural generation is considerable (and can be replicated by a random dungeon generator) and because of the very limited content on the tables, playing through these spaces would be deeply frustrating. The most common critique of classic style play is that it's empty room after empty room, broken up only by the occasional simple trap or an encounter with some kind of dangerous dungeon vermin. I strongly disagree with that characterization when it comes to a well made classic dungeon, but for most of Dwarrowdeep it will be literally true.

As some solace, there’s a 100 entry table with room dressing, but given that one will have to fill thousands of rooms this is a paltry offering. It’s not even that the procedural generation tables here are terrible - they aren’t, they would work well enough to create the odd, small, underused portion of Dwarrowdeep, but they are profoundly insufficient for the amount of times they will be asked to create adventure sites in the average Dwarrowdeep campaign.

Additionally, and relatedly, Dwarrowdeep has a serious organizational problem. Like its procedural generation woes, this is something that’s fairly understandable in such a large product and could easily be forgiven if the rest of its content were stronger, but to use Dwarrowdeep one needs to flip back and forth constantly or presumably print out and annotate maps. It’s unfortunate, because for reference breaking up its keyed areas into subregions with maps nearby (effectively following Stonehell’s example) allows the referee greater ability to understand the layout of the dungeon as a whole while preparing the areas that their players are likely to explore in a given session. Again, this isn't a problem becasue there aren't better options: MERP's Moria and Stonehell offer immediate models for improved layout and both long predate Dwarrowdeep.  I suspect this is another issue stemming from Dwarrowdeep’s decision to use a sort of BECMI era layout and design.  To repeat myself again, the techniques that offer functional, but not exceptional, usability in a 20 or 30 page adventure with detachable map cover are very different then those one needs for a 200 plus page megadungeon.

Dwarrowdeep isn’t a good adventure, but it does a few things quite well and includes enjoyable content. This is especially true of some of its additions and appendixes. Dwarven Clerical spells and magic runestones, while both overused in Dwarrowdeep (almost every Duergar leader is a cleric and most magic traps simply cast these spells) are interesting additions. Likewise Dwarrowdeep’s maps are generally functional, even if several are very similar -- much like the keying for these locations. The pregenerated characters are charming and the Dwarrowdeep themed character sheet is a pleasant addition. As previously mentioned the art is excellent, and Dwarrowdeep even adopts one of my personal favorite early D&D oddities - a booklet of player facing illustrations that show rooms and artifacts within the dungeon. Sadly n
one of these thoughtful inclusions make up for the mess of the dungeon itself. 

Some fans may point to these positives and say that the above criticism of Dwarrowdeep is unfair, that all of these issues can be solved by a dedicated referee.  Maybe they can, but I don't purchase published adventures to rewrite them almost entirely just to make them remotely functional at the table.  Refereeing will always require prep work, but the only reason to buy an adventure - especially one so lacking in new ideas is because it reduces that prep work. I firmly believe that any adventure you write and run will be better for your table then any adventure you buy, but published adventures can be usable and inspiring enough to play without a complete rewriting.

There are Few Dwarrowdeep Illustrations Online
Here's Some Proper Evil Dwarves
John Blanche Again - 1985


I try not to be this critical in my reviews, as I genuinely believe most people making adventures are doing so out of the same enjoyment that drives me to do it. Yet, I won’t be overly kind either. Spectral Interrogatories is aimed at giving helpful advice to designers, providing as objective a reading as I can offer within my own understanding of how to design and run dungeon adventures. Dwarrowdeep desreves harsh criticism. Dwarrowdeep is not the worst adventure I have read, not even close, but doesn’t appear to be a playable adventure either, especially not for its intended purpose as a five year, level 1 - 10+ campaign. It's disappointing given the obvious care went into some parts of it, and becasue I sincerely want a new and better classic style megadungeon.

I think I’ve detailed how Dwarrowdeep fails in nearly every possible way as an adventure, even accepting the fact that it’s a cliched dwarf ruin - even reviewing it was a slog, but interestingly its failure has led me to two questions.

First, why does Dwarrowdeep fail despite the author’s experience and the resources behind it?

Second, what is the maximum scope of a dungeon crawl?

I don’t have definitive answers for these questions. I suspect that ultimately they both come back to the issue of scope. Perhaps the creativity or time to spend imagining an adventure the size of Dwarrowdeep simply wasn’t there? Perhaps the author hasn't read other works in the megadungeon genre? Still, there’s something stunted and ossified about Dwarrowdeep that I suspect it comes from the strange idea that when an adventure looks exactly like something from the past it will be as good as that past thing.

Both Gygax and Tolkien were creative and groundbreaking in their chosen fields, and while there are critiques to be made of their work, any critique has to take into account that it was new and experimental. The same doesn’t apply for derivative works 40 or more years later, and anyways Dwarrowdeep doesn't even approach the quality of Tolkien's Moria or Gygax's D series. A fundamental issue with nostalgia and nostalgic works is that there’s no way back to the past. It’s not a deep concept - pure high school stoner philosophy - but you can’t step in the same river twice or recreate the works of past greats and expect their success. Of course space remains for reference, homage, and using the tried and true, but especially when one sets to making something on a different and grand scale, simple fidelity isn't enough. Even if it was, there is no great BECMI era dwarven ruin or published megadungeon to faithfully copy. Dragon Mountain is about as good as it gets, and it's not a name one hears frequently ... for good reason.

I don’t understand why Dwarrowdeep chose the path of blandly nostalgic mimicry, not just in aesthetic (which I do understand), but also in design. The adventure even notes or starts to offer many solutions to the problems I’ve listed in it’s text, but fails to use them, and most aren’t anything new: Caverns of Thracia, Stone Hell, Anomalous Subsurface Environment and even Dragon Mountain to a degree all manage to grasp the issues of faction, functional keying, and engaging secrets better than Dwarrowdeep. It has no excuse for most of its errors, and that’s really disappointing.

I want the luminaries of what was once the OSR to produce material that lives up to the promise the movement had in 2006 -- it’s been over fifteen years.

Of course there’s another possibility -- that Dwarrowdeep’s obvious weaknesses as an adventure are magnified simply because it reaches the limits of size and scope for a location based adventure.

How big can a dungeon be? How many pages of keys can you have before the place becomes too repetitive to play through, too complex to referee, or too varied to be considered a single location?

There are obvious improvements to Dwarrowdeep that one could make - stitching mysteries and secrets throughout the dungeon more effectively, giving NPCs goals and relationships, providing varied and interesting cave and underground environments as secondary areas - and reducing these lesser nodes in size to 10 or 20 keys. Even 30 one page dungeons would make for a better set of lesser locations, while a slightly more abstracted long distance travel system using watches on a more detailed map would provide a far better experience. Organization in general, or even the minimal effort of a providing a PDF with hyperlinked table of content and links to maps from the start of the location description would all go a long way to improve Dwarrowdeep and all they would require was a willingness to step away from complete fidelity to the BECMI design model.

 Would this be enough?

I’m not entirely sure, maybe no amount of usability efforts or smart, streamlined, megadungeon design could make something with Dwarrowdeep’s massive scope function, and maybe it’s not even worth trying? At some point to make a huge dungeon hold players interest it starts to become more than one dungeon.

This question also raises a second, uglier possibility … are megadungeons a stunt? Are they a sort of a boast by the designer where the goal isn’t to create a good adventure, but to create the biggest adventure?  I don't want to say they are, I like megadungeons, but something like Dwarrowdeep feels far more like a stunt then anything else.

I am left with sadness and questions at the end of Dwarrowdeep, but much like reading one of WotC’s 250 page adventure tomes, I don’t have a great deal of memory about the adventure itself because beyond what I've written.  It’s that empty a thing. Neither Dwarrowdeep or 5E's campaign books seem designed for play as much as they arre designed to represent some abstract ideal or promise of play, and I can’t fill them to usefulness with theoretical reflections on design, just like a referee running Dwarrowdeep couldn’t fill it with engaging, interactivity and evocative detail -- at least not without rewriting it.


  1. I note only that this work *can* be compared to a very specific work - MERP's Moria module, which is from precisely the era emulated (1984), which - for what it presents- was quite well done and evocative. At the end of the day, it's an abandoned mine with miles and miles of tunnels, now inhabited by evil humanoids and a big baddy. Obviously, acquiring a legal copy is now price-prohibitive (thank the Maker I discovered eBay before so many like-minded nerds!), but the problem remains: How does one re-make Moria?

  2. I actually updated the review in my edits this morning with a mention of the 1st MERP Moria. It's only 75 pages and it does have the high quality and density of ideas I associate with MERP. I don't really think it's entirely usable though - more a construction kit for a Moria - which is fine - it looks like it has enough flavor to inspire something good.

    I don't know how to write Moria as a classic dungeon crawl, which is why I took the plunge, bought and read Dwarrowdeep. I may not know how to do it, but I know this isn't it. Mostly I think you'd want to make the map less complex and provide more for the smaller nodes, multiple factions, interactive puzzles/keys/obstacles and add multiple overarching hooks: dwarves retaking it as a domain/military expedition, finding Durin's Crown to return the dwarf demigod, making a pact with or overthrowing the balrog to rule it for evil, hunting down whatever the big score within is to get filthy rich/powerful or even just getting from one side to the other as a shorter adventure.

    Still can't say it'd work - and if I ever do a megadungeon it's more likely to be a crashed "war moon" filled with vampire space elves dressed like David Bowie then a faux-Moira. I guess we won't know what works until someone else finally does it.

    1. "O don't really think it's entirely usable though - more a construction kit for a Moria - which is fine - it looks like it has enough flavor to inspire something good." I think that's fair, but I also think it's fallacy (or perhaps, better, a Quixotic endeavor) to "stat" Moria. Moria is a setting on it's own. Mapping out miles and miles of empty tunnels and old quarries, for the sake of mapping, is pointless. It can really only be presented as a vast overview with detailed points of interest as a party traverses through. "Clearing" Moria room-by-room, tunnel-by-tunnel as if it were a dungeon isn't really a gaming experience, I wouldn't think. There's some enjoyment to exploration of spaces, to be sure, but after the ninth or tenth access tunnel, it certainly seems like it would become a joyless grind. Does one "clear" Mirkwood?

      We know that megadungeons *can* be done, but they are difficult to do well in a different way than how "regular' dungeons are also difficult to do right but can be. And once one is trying to emulate possibly *the* ur-megadungeon, one is setting the bar for onesself incredibly high, in particular because the author is competing against fictional Moria, the Moria readers envisioned in their minds already, MERP's high-end and TSR-contemporaneous Moria and every Moria-esque pastiche since.

    2. Yup.

      That's pretty much what I'm getting at in the review, I hope it comes through.

      Dwarrowdeep is broken on so many levels and I think the primary reason its broken is that it appears to reject any lessons from other efforts to design something like it -- and there are so many attempts.

      I think the MERP one is good, but fairly different and I largely chose to look compare it to Stonehell and Thracia, both of which I've read, both of which are good dungeons, and both of which are written for effectively the same system. That they aren't specifically "Moria-likes" matters less because the failures in Dwarrowdeep are largely general mega dugneon failures. I also chose Dragon Mountain because it's so similar in theme and does interesting stuff with faction - plus I had the PDF laying around. This stuff is a good chunk of what I mean about "poisonous nostalgia" - distinct from the regular personal nostalgia for old games and adventures that most of us have.

      There seems to be this point where loving something someone made back in the day, or that the designer enjoyed playing back then, blinds the designer to that beloved products weaknesses or limits (such as not being built for the scope of a giant mega dungeon), and to other sources of design. You get something at the end that's held up as an embodiment of the old design and tries to claim authority from its sources. It bugs me partially because the actual history of these old designers and adventures shows so much flexibility and innovation,

      Imagine someone new comes along and runs Dwarrowdeep, experiencing all the bad design I've been mentioning and is left thinking that's what old games are like. That's what makes me sad about Dwarrowdeep.

  3. It seems like we're throwing out the "Megadungeon baby" with the bathwater, as many of these newer ones are regression, not progression, on the design which saw its pinnacle with the famed 90s megadungeon, Ruins of Undermountain.

    I'm still struggling through the Halls of Arden Vul, but in my estimation, most of the OSR (or modern) megadungeons are either multiple levels without any tie-in (a dungeon for a sake of a dungeon) or they are basically some sort of pale imitation of Castle Greyhawk/Xagyg, but missing the actual mechanics, cleverness and randomness that made that grandfather enjoyable. Not that it could be replicated, as most of the mechanics of the original were probably lost to time.

    ROUM sort of circumvents many of these issues, adds in the factions needed, allows the DM to create scenarios rather than dictating a plot, adds in randomness, adds in purpose to doing the megadungeon, and removed the factor of "kill is the best solution" with clever encounters which changed over time, powerful enemies which weren't balanced, and random magic and encounter which kept things fresh.

    I don't think megadungeons are a stunt.

    They've been actively stunted by not iteration on the better design for them, which in my estimation, is exactly why Dwarrowdeep fails.

    1. I don't actually have an answer to any question about megadungeons as a whole here - I cite three good ones here: Thracia, ASE and Stonehell. I've reviewed them all on this blog (though ASE deserves something longer), so I mainly stuck to them. I hear Gunderhoften is good and Gradient Descent didn't quite work for me, but has some great ideas in a very concise package.

      The issue is with Megadungeons is SIZE. More it's an issue with designers trying to show off how big their Megadungeon is. It's the motion of the ocean fellows.

      Dwarrowdeep tries to be very very big, and I think this scope is a significant part of its failure -- just part though, there's a lot of kludged work in it, basic things like: factions, interactive keying, usable design, and mysteries with consequence are missing.

      The interesting question as I see it though is if there's an upper limit on dungeon size? Obviously Dwarrowdeep could easily do far better, but even if it did would it have to shrink to function well?

      As to Ruins of Undermountain - I think I've looked at it briefly, but it wasn't memorable to me - the writing style of mid/late TSR tend to irk me though, so if it still has fans outside the Trad scene I might have to take another look.

    2. Size is also a curse for Undermountain. It's gigantic, underdetailed, and TSR tossed in endless empty rooms to make it bigger (Level 1 contains the entirety of Quasqueton from B1.) Ruins of Undermountain I is a toolkit that works okay, but it's nowhere near complete. Other products have been shattershot. Oddly enough the 5e version is the first complete top to bottom keying of the dungeon (by turning most floors into 20-room linear dungeons and removing most of the interesting design and setting building), and the only other thing that comes close is 3e's Expedition to Undermountain, which is a mostly linear jaunt through the larger levels but does at least have a point to it.
      Undermountain is a monster, and it's never really been done full justice. I can't say I'd recommend it to anyone except Realms fans at this point because it's such a lot of archaeological work to sort out the 10 products on it across four editions and create something playable.

  4. I'm a little reminded of the unfortunately named Judges Guild dungeon Glory Hole Dwarven Mine by Edward Mortimer. It's a large sort of dwarven ruin (7 levels) that has some interesting sublevels and factions. There's a good mix of keyed locations and procedural type tables. All the elements you describe.
    And I like the new monsters. It was produced in 1981 so probably one of the first big Dwarven dungeons produced. It is Judges Guild so a little shaggy and unfinished but better than most non-Jaquays works.

    1. I don't think I've ever read that one - mostly with this series of reviews I try to use examples that are well known or which I've previously reviewed. I'll take your word that The Dwarven Glory Hole is a decent adventure, but I'm unlikely to do a series of reviews on "moria-likes" at this point, because there are so many.

  5. It would appear a LOT of work and effort went into making this unplayable adventure…that, to me, is the saddest bit.

    I wonder how much play testing was done, and if/how the author dealt with the various issues you’ve outlined.

    1. It certainly did take a lot of effort, and I suspect a fair bit of play testing. I've heard the author was running it at some cons and I assume there was more. I don't know how much if any involved testing the thing as a whole in the context of a long campaign (seems impossible given the timelines) or if there was testing by others. I find with play testing that if you're running your own thing you miss a lot because you already understand all the unwritten things. Getting some other people to give it a try is important because they can spot stuff the author will miss.

      Likewise, testing one node at a time here would likely work fine especially in the one shot context. For a convention game "go burn down the orcs/derro/naughty dwarves outpost" sounds like a decent session. Dwarrowdeep errs between those sieges and raids and when it asks players to just repeat the same adventure in the same environment again and again.

      The oversights or failures of Dwarrowdeep strike me as intentional choices - decisions to emulate specific models and ignore a bunch of others, decisions about what's needed for an adventure and what's not. I think it gets all this stuff very wrong, but it's almost certainly a design ethos at work.

  6. From your section on factions: " A simple faction chart showing resources and a list of missions/goals would do wonders here." Do you have examples of products that do present factions well and accessibly, megadungeons or otherwise? People often talk about them but I find they're often buried in the text and difficult to grasp.

    1. I haven't seen anyone really focus on it, but Dragon Mountain mentioned above has about a page on each kobold faction at the start of the key for their (color coded on the map) region with a brief order of battle, allies, plans, enemies and such. Not perfect, but it's from some of the darkest days of TSR design and yet still manages to be more useful, inspirational, and interesting then Dwarrowdeep - I'm not demanding perfection, but some effort would have been nice.

  7. This might be short sighted, I've played in megadungeons much more than I've run them and am also but a sallow youth, but the structure of a megadungeon that seems most suitable for something such as Moria would be the node based style of Black City (Dreams in the Lichhouse) where only small areas are mapped out in the conventional way but longer transit passages are handled as lines between those nodes.

    1. I think a node based mega dungeon would work great, and Black City is a lovely example - I used to look forward to the play reports back in the day and wish that had gotten published. The major issue with Dwarrowdeep's nodal design isn't that it uses it, but that it doesn't learn from the various efforts to produce that sort of design, doesn't provide useful systems for travel between nodes, and that it fills the majority of its nodes with the SAME proc gen.

      The proc gen isn't even terrible, one could make a perfectly fine 30 room location from it, but it's in no way suitable for 30, 20 - 200 room nodes.

  8. Yeah, I have to admit this one was a pretty big disappointment. I own and have played parts of Gillespie's previous three mega-dungeons and they all have something to recommend them. Highfell and Archaia can both easily be mined for smaller locations. Barrowmaze, while it gets by far the most attention, is more of a slog and gets repetitive pretty fast. As for this one, I'm most bothered by all the tedious random generation. I don't pay $75 for a product to have a huge chunk of it be DIY. After randomly generating a couple small sub-levels, I just gave up. We haven't played it and likely won't. Another pet peeve is the fact that XP aren't included in monster stat blocks. I know this is a throwback to early TSR, but I can play Rappan Athuk and those values are there. Likewise, Matt Finch's Cyclopean Deeps is a far superior and more creative Underdark product for a long campaign, even if it's derivative of the D series. At least the locations and factions are more interesting.


Folie à Trois: Trophy Gold

Below is a shared review of Trophy Gold (2022) , a fantasy adventure game designed by Jesse Ross and published by The Gauntlet. Although it...