[Review] Maze of the Blue Medusa Follow Up – A Comparison with Various Megadungeons

Given the responses, the buzz and the page views on my, admittedly flowery but ultimately well-founded criticisms of Maze of the Blue Medusa, I thought it might be worthwhile to do a follow up to elaborate on some of the points that were made or some of the points I did not quite cover. It strikes me that the first review was actually fairly lenient, as we judged the work purely by the standards of the author. We did not get into a comparison with its contemporaries.

One of my points was that Maze lacks, for the most part, the higher level organization that separates good Megadungeons from just big dungeons. This is all the more amusing if juxtaposed with the claims of being ‘madly innovative’, as structurally Maze is more primitive then some of its predecessors.

Megadungeons go back all the way to the description in Book III (The Underworld & Wilderness Adventures) of the little brown books and originally the Dungeon as Campaign was the assumed standard but paradoxically no true examples were available in print. Dungeon creation technology is ancient. Or in the words of the Gygaxter:

You tell em’ Gary

The actual largest period of megadungeon innovation likely took place during the resurgence of interest in megadungeons in the OSR around the early 2010s that saw the release of Barrowmaze, Stonehell, Anomalous Subsurface environment and probably culminated in the ill-fated but massively successful Dwimmermount Kickstarter. Before this there was of course the legendary Rappan Athuk, which likely carries over some baggage from the D20 era, 2e Undermountain, which is verbose if I recall, Temple of Elemental Evil, which has its own problems and might be a bit too short and the hated Castle Greyhawk knockoff, to name but a few examples. Today we have the critically acclaimed Castle Xyntillian (which I have yet to review), Günderhofen, Highfell and the brobdignagian Halls of Arden Vull, to name only the major examples.

In terms of High Level Organization Stonehell and Barrowmaze utilize several mechanisms to accommodate or even take advantage of the lengthier campaign format. Most of these are 1970s technology, others are applied principles that were known but seldom applied properly. The point of many of these is not that they require a super-high level understanding but that anyone who has actually run a megadungeon campaign or has done more then skim-read the little brown books or any of the dungeons in question should be able to apply these or at least consider them to some degree.   

* Progression. Stonehell is divided into levels. I know what you are saying, levels are old Prince!, duh, but hear me out. Dividing areas of the dungeon into different sections, providing opportunities to ‘skip ahead’ and using gradually increasing (and sometimes rapidly increasing) risk/reward ratios introduces the ability to make interesting decisions. Do we go deeper or do we slum it on the second level? It even becomes possible to make areas with sub-optimal risk/reward rations that are best avoided altogether or that appear as a nasty surprise to the players, something that Stonehell, because of its immense size, can easily pull off. What about discovering exits to the surface on lower levels so the PCs can skip entire areas, what is a better reward then that?

Barrowmaze does this to a lesser extent, as there is again a clear difference between the lethality of different areas of the tomb, particularly in the random encounters. The whole organization serves a common sense goal, to provide an increasing level of challenge and reward as one progresses further on in the dungeon. Barrowmaze is also king of multiple entrances, and immediately provides not just an initial entrance, but an entire overland map filled with barrows (of decidedly more various threat levels) to explore, some of which connect to the dungeon proper. The moment you enter the field you are SPOILED for choice, but you are restricted too, since the village is hours away and the dungeons are so perilous you cannot stay there. This time element, and the undead encounters getting more hazardous with night-fall, is utilized beautifully 

In Maze this organization is largely absent, or rather, the areas have been organized by theme more then by threat level. Areas are differentiated only in the frequency of the random encounters. Monsters of varying abilities are scattered haphazardly across the map. The Chantry and the Cells with their high frequency of Caretaker/Moon Man encounters might be counted as ‘more dangerous’ but distinctions are blurry, particularly because of the often ambiguous nature of Maze’s unique inhabitants.

The vertical component also introduces another element. The dungeon is now of unknown, potentially infinite size. Sub-levels can be introduced passageways, dungeon levels can lead all the way to the Hollow Earth, the Underdark or Hell. This speaks to the imagination.

Conversely Maze, for all its pretensions to Non-Euclidean geometry and use of what are essentially teleporters, is organized on a single 1000’ by 1000’ square. Vertical components are illusionary i.e. they exist in the text but can be ignored in mapping.

Think also of elements like foreshadowing, or building-up towards a climax, or even having a goal. In Barrowmaze there are conditions that must be met, certain artifacts that must be gathered to fight the final boss (the Keeper), with a (possible) suprise reveal that there is yet another, second final boss (the dracolich) all with the end goal of destroying the Tablet of Chaos, the sinister cause behind the undead rising in the Barrowmaze.

Gary on monsters: Book IV

Stonehell, if anything, does it better even without a clear end goal, with the dungeon starting off almost perfectly vanilla and the party being hired on one of several pretenses and the weirdness gradually increasing, new races get introduced as the party delves deeper. You only figure out how fucked shit is as you go deeper and deeper. Every level starts throwing vicious curveballs. Reality becomes mutable, more and more completely unknown creatures show up, until the final level, a surreal death-march through a psychedelic hellscape populated with insane horrors, and a slugging match with an almost lovecraftian horror (the Nixthisis), whose presence has been foreshadowed levels before, and who can be killed only by fighting him on two planes simultaneously. That’s perfect. That SCREAMS epic.

In Maze of the Blue Medusa, the encounters dazzle with creative energy but there is no climax, no build-up. There is plenty of behind the scenes lore to discover but it does not spur action. The Maze exists in effective stasis, removed from causality, and the players are the disruptive agency. Everything else is hampered from action for one reason or another. This might be the reason why it feels so oddly static. This is exemplified by the Medusa herself, whose goals involve only the removal of minor nuisances and otherwise becoming emotionally reconnected or (if we are less charitable in reading the text) to have gay sex with the three minor powers of the Labyrinth. All her major opponents are already bound, and can only be released by the players, to a net detriment. What happens if she achieves her goals, we ask ourselves? What are the stakes? What about the three Sisters? One wants to futilely start aging (despite the fact that all three of them are immortal) and to talk to the Medusa, one serves as Gaoler for a type of Arch-lich and wants to escape but already knows what her presence would do to the world outside so she does not resist her imprisonment (why not have the Lich planeshift you to one of the higher planes, like Nirvana, where your peaceful aura cannot harm anyone?) and Zamia Thorn, who inspires obsessive love in everyone that looks upon her, is much the same, to the point that she would rather stay a ghost. Their miserable predicament seems irresoluble.

We have a very memorable climax if the Medusa is ever killed because this flash-unpetrifies everyone in the cells and immediately triggers at least one direct attack but why do we even want to kill the Medusa, she seems like a benevolent person. Is it just so we can help ourselves to ONE MILLION DOLLARS in treasure? Are there hidden weaknesses of hers that we can discern elsewhere? Is there some weapon or hated enemy, hidden somewhere in the maze, that we can use against her? Maybe we can lure her into some sort of trap? I’d be more inclined to give Maze some slack in this department if I had the idea these sorts of gameplay related questions were even considered. Instead we must grapple with ‘themes’, experience ‘wonder’ and sort of nod slack-jawed at the humanity of it all, all the while ingesting bite-sized chunks of designer-DnD.

Pictured: Dungeon Dressing Technology from the Golden Age

* Continuity (via Restocking or Otherwise). A simple, yet useful component to dungeons above size XL. Tables to restock areas exist in both Barrowmaze and Stonehell and go all the way back to ToEE and the advice in the little brown books. The rationale is simple and serves multiple purposes.

Versimilitude: In a huge underground construction of unknown but vast size containing multiple tribes of humanoids it makes sense that areas suitable for habitation become re-inhabited as the dungeon inhabitants move around the dungeon.

Uncertainty/Unpredictability: In drawn out campaigns things must be kept fresh and occasionally that means throwing a wrench in a well-oiled machine. Areas that were previously considered safe can become hazardous again, forcing detours or precious resources. While in the dungeon, even in areas that were previously cleared out, the party must always be on their guard.

Time/Pressure: Say it with me. Pressure. Pressure. Pressure. Put their soles to the flame. Dungeons that are gradually restocked must be tackled in bold, sweeping strokes rather then the cautious inching forward that the all too cautious and meticulous of players can fall into. DnD is generally good about making you pay for each wasted turn and this principle is merely an extension of that.

This is not to say that you can’t have a good dungeon without restocking, there are many ways that lead to Rome. Again, in the words of the big guy:

With Maze this is ignored and procedures for restocking are absent. If anything, the Dungeon becomes considerably more desolate and monotonous as time progresses because of the nature of its random encounter tables. As previously mentioned, maze uses the same Base Random Encounter table for each area and varying 30% to have different encounters of that same table depending on which area you visit. The interesting part is that many of the monsters on the table are unique, and killing them will cause the entry to be replaced with Chameleon Women. 9 out of the 15 different entries are unique or can be permanently destroyed. These Chameleon Women show up in groups of 1d6 of which one can cast all spells up to 5th level, increase in 1 HD each time they are defeated and have an 87.17% chance of attacking unawares when the party is either weak or already engaged with something else. Because of this increasing ability curve and ability to attack with surprise, it is actually wisest to leave all the unique monsters alone and run from the Chameleon Women at every opportunity. I am stumped again, this could almost be a good idea. It is certainly difficult and it requires learning the rules of the dungeon, albeit the rules that govern the arbitration of dungeon, and not the rules of any in game reality since these modified encounter frequencies don’t have an in universe rationale. If this was some sort of Borg-like entity or represented some sort of Maze defender that gets more powerful as it accumulates knowledge against you this would actually be great design, in fact I think I postulated something like that in this post here.

Indeed, the old random encounter model in OD&D is quite interesting. As one can observe the dice would allow for a broad range of monsters to be encountered even on the upper levels. Most megadungeons use a far more narrow range.

Of Dungeon Alteration. Much ado is made about Altering the Dungeon but Maze seldom stops and asks itself ‘what does this do?’ or rather ‘how does this affect gameplay.’ You can get rid of the Vines in the Gardens, does that do anything? No. You can burn down the archives, does that do anything? Not answered (the king dies but he lets you pass anyway). You can flood the entire dungeon to 5 ft depth. Okay and now what? You can steal golden balls from a fountain and make the Reptile Archive more aggressive…that is an example of the Good change because it alters HOW THE PLAYERS INTERACT WITH THE DUNGEON.

In fact, if your dungeon is any good whatsoever, playing regular D&D in it should to some extent ‘alter’ it, that is to say, you clear pathways so they are safe (at least temporarily), you rid the level of dangerous inhabitants, you unlock or break open closed off areas, you can barricade or spike open doors, suspend ropes or erect planks over chasms, set fire to things (repeatedly). I think spotting interesting possibilities for this sort of sweeping change is a good thing but I am not convinced setting it apart rather then incorporating it in the dungeon design naturally as one goes along is the superior way.

Comparisons with Cha’alt. Of all the Megadungeons I have covered the one dungeon that has most in common with Maze of the Blue Medusa is Venger Satanis’s Cha’alt, both in terms of advantages and disadvantages. Versimilitude is mostly done away with, normal mapping procedure is eschewed in exchange for something strange and creative and the dungeon proper is built up from single encounters that have interrelations but do not neccessarily form a coherent whole. Lavishly illustrated, minimalist statts.

Brief Aside: More Gary Dungeon Tricks

But compare the different approaches; Cha’alt toys around with the idea whether or not there is a point behind it all, provides ample hooks, rumor table, random encounters 1/hour or whenever the PCs fuck around too much, minimal backstory because you expect this to be some sort of insane funhouse nonsense that you can explore and be dazzled by, and lots of deadly traps. Every encounter has been fearfully written, as though some wrothful consultant was peering over the author’s shoulder, yelling whenever an encounter did not potentially initiate some sort of PC interaction. Final boss is whatever TPKs you because wandered into the wrong room. Cthulhu, the Devil that you lost a rockoff against, Zargon ripoff, Ara’ak Zuul, Szar’Cholek or the button you pressed that destroyed Cha’alt. Its fun, its wonderful, it’s nuts, its a guilty pleasure and with just over 100 rooms it doesn’t outstay its welcome.

Maze: THICC convoluted backstory that ultimately serves little purpose because of dream-logic and lack of consistency, underlying themes that you can observe or experience but not truly resolve, rooms that are little more then art pieces or creative writing to be observed. You have all this power, all this writing talent at your disposal, and it’s misapplied, because the aim is not to create something that is good but to create something that is ‘innovative’. We must be visionaries! Tried and true dungeon staples are cast out because they have been ‘done before’ but the result is something that, for all its creativity, lacks the range of more conservative Megadungeons. All the innovation burns out after a while.

One of these is in harmony with what it is and tries to be. The other tries to be something more then a mere DnD megadungeon but ends up being less.

I conclude with the sincere question: For all its claims of being innovative, and all the praises laid at its feet, for all its undeniable market penetration, is there anything Maze has done that has ended up in subsequent megadungeons? Anything in terms of organization, encounters, gameplay, mapping? Has the way we do megadungeons changed since 2016 and if so, has it changed for the better? If so then it has earned its monniker and it can claim brilliance or at least excellence of some minor or major sort. If not then we are forced to conclude that what it has done is not innovation.


Also if you want something that might not be madly innovative, but sure as hell is madly good D&D, check out the winners No Artpunk Contest here.


18 thoughts on “[Review] Maze of the Blue Medusa Follow Up – A Comparison with Various Megadungeons

    1. I agree.

      Three reviewers. 10ft – prince – melan, have the review domain in control but IMO they have poor a understanding even of basics like Jaquays or Bledsaw.


      1. Edgecases, but probably the closest of the old stuff upon a cursory inspection. Dark Tower is two submerged towers with 5 levels each, with occasional connecting passages. Thracia is 117 encounters over 4 levels with 2 sub-levels, with a size of about 15-22 each?

        Melan has always stated that his work is more JG then Gygax. I have read only Wilderlands (should re-review) and Tegel (also probably), which one might also lump in with these other Kilodungeons.


      2. I’d call Caverns about as close to a published Megadungeon one gets from TSR era D&D. 4 floors is small in comparison to some but respectable. Stonehell definitely pulls from and innovates on Caverns in terms of “each level archeologically drags the players back further in time”, with certain later levels showing the once great history of the guys you were slaughtering earlier. Stonehell’s more open later levels also pull from the verticallity created by the crevasses on level 1 of Caverns.

        Dark Tower I only scratched the surface of at a convention a couple years ago, so I can’t say. Looking at it’s maps, it’s got the levels for it, but being mostly set in two towers might hamper that.


  1. Mordenkainen’s Fantastic Adventure – terrible name, sure – is the closest thing to an OD&D megadungeon TSR put out, since it is drawn mostly from Robert Kuntz’s El Raja Key, and has definite OD&D sensibilities.

    Caverns of Thracia qualifies in a sense, but Dark Tower, while a masterwork, does not. It is a very large dungeon, but it is a dungeon centred on a mission, not an open sandbox environment which you can send repeated expeditions into over the span of a campaign.

    Liked by 1 person

  2. While I will defend Maze of the Blue Medusa as the only truly avantgarde form-innovation that the hobby has seen for the Dungeon format in a long time, I will not defend it as a Megadungeon. It simply is not one. It is a large avant-garde poem-dungeon (maybe even a dungeon-poem?), first and best of its kind, yes.
    But a Megadungeon, no.


    1. I’ve listened to your excellent podcast interview with Prince at least twice (thrice?), and know you appreciate the artistry of MotBM’s “dungeon poetry.” That doesn’t make it a worthy adventure, megadungeon or not.

      Everyone likes what they like. I like Dwellers of the Forbidden City. I1 is a flawed adventure in MANY ways, but I find it an inspiring one…so much so that I’ve run it multiple times without any addition, modification, or attempts to “flesh it out.” Doesn’t change the fact that the execution of the design could have been better. I just “like” it.

      Liking Blue Medusa isn’t a crime. Liking its artistry isn’t a crime. But defending its value as an adventure based on “artistic merits” is bullshit. Lots of people can create giant, sprawling dungeons – with or without original monsters – and add NPC personalities, fictional histories, fantastical backgrounds. That shit is NOT hard; anyone can do it, given enough time (and scratch paper) to ruminate and brainstorm.

      Good artwork IS harder…but in the final analysis, D&D is not art to be observed and “taken in.” Rather, D&D is a game to be played and experienced. The adventure must be designed in a way that facilitates such in a coherent fashion, whether it is a megadungeon or the 5-room variety. That’s the only real measure of an adventure’s quality.

      Everything I’ve read about this adventure indicates a tremendous amount of work went into it. Everything I’ve read indicates it is incredibly artistic. Everything I’ve read indicates it is a humongous piece of crap as far as design. Coupled with the amount of work, artistry, and page count, that all reads “narcissistic” and “self-aggrandizing” to me.

      I have a pretty good idea of the limits of funhouse dungeon design. White Plume Mountain strains the edges of it at a bit less than 30 encounters. Adventure sites (“dungeons”) divorced from their settings or a larger meaning/role in the campaign grow stale rather quickly and (in my experience) become all too tedious for the players when exceeding more than a couple/three dozen encounter areas.

      Based solely on the reviews, I find it difficult to fathom any group of players wanting to experience MoBM in its entirety, as written. And that makes it a poor adventure for me, regardless of any “artistic value” it may have.


      1. Not really a fan of the art in MotBM. Whereas much of Patrick’s prose I do enjoy as reading material. I thought it was a personal hipster failing of my own that I could never run it when I first purchased it (amid the raving reviews) of 2016.


  3. I sense that Prince is definitively putting the critical smackdown on MotBM, leaving no squirm room to evade the relentless, if not concise, erudite analysis. But it was seemingly worded in a way that would evade inviting excess mockery, which Prince is most capable of. Would it be presumptuous to think that Zak could take this as meant, add some good notes and addendum or even change the text to allow for a few modest improvements? It would be nice to see.


  4. And MotBM ain’t a Megadungeon . Unless the Chameleon Women’s society gets anal-retentively Fluffed up into content all their own? Maybe MotBM should not try to be so unwieldy as to want to be a Megadungeon.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s