Monday, August 9, 2021

Spectral Interogatories II - Castle of Mirrors


Castle of Mirrors is a 39 page adventure written for house ruled 5th edition Dungeons & Dragons by Runehammer Games, presumably by its lead creator Hankerin Ferinale. It’s an interesting product in that it represents the work of a clearly talented designer attempting to wrestle with the contemporary play-style of 5th edition and the impulse to create dungeon crawl style adventures without system support. I don’t know anything about Runehammer or Mr. Ferinale, but I attempted a similar project for over a year, only to abandon it, finding 5th edition's lack of exploration mechanics frustrating in many ways. Unlike my efforts to reintroduce classic exploration mechanics, Castle of Mirrors tries to distill 5th edition's fundamentally cinematic and encounter based design down to its components: combat, DC based skill checks, and abstracted exploration. It’s a clever concession to a system where combat is the primary locus of play and intensely time consuming, though it’s also confusing to someone used to Classic play and dungeon design.  I haven't played Castle of Mirrors, and don't intend to, but it's interesting all the same for what its design choices say about mega-dungeons and Contemporary Traditional play.


Each region in the adventure is composed of several ‘Scenes’ that require a skill check to move between. For example the first,  ground level region is composed of ten Scenes, each of which has a skill check (DC 12 to 18) to move to without an incident (meaning a combat encounter). Scenes are lightly sketched (up to a ½ page of large print) general description that lists "Zones" within, such as “The Gatehouse”, “The Causeway”, “The Rubble Pile” and “Muddy Pond”. There’s almost no detail for these Zones, though each Scene may have (unique) monsters listed.  This makes the Zones feel less like locations and more like a backdrop without distinction from other parts of the Scene beyond the descriptive name and maybe a single mechanic -- movement is slowed by the Muddy Pond for example.

Scenes also have several “Actions” within that appear to be a sort of substitute for a location key or an individual obstacle within a key. Most are combat encounters, though there’s some stat check based obstacles, potential role playing encounters, and descriptions scattered within. It’s unclear how well this works. Interactivity is low, primarily the adventure will consist in a string of combat encounters and general description -- this is of course effectively the common 5th Edition design and play style, but for players trying to move away from this style, or even break with it momentarily, Castle of Mirrors offers few tools to replace combat with other kinds of play. The major tool that defines the adventure's scope is not description or even complex tactical spaces where the environment plays a large part in the encounter, but hooks for each Scene. These hooks are extrinsic character motivations or missions, provided for each region of the adventure and presumably they act as goals for individual sessions, drawing the party to a specific region, Scene and Zone of the dungeon for a specific purpose.

This structure isn’t new to Castle of Mirrors, notably it’s similar to the design of 13th Age’s “Eyes of the Stone Thief” a 2013 mega-dungeon for a similar style of play which is far more complex, of a larger scope, and has greater imaginative punch. Castle of Mirrors is more spare then Eyes of the Stone Thief (and much, much shorter), pared down to the elements necessary for its chosen play-style, and dispensing even with the sorts of descriptive (rather than procedurally substantive), very attractive, maps found in Stone Thief.


The content of the Castle of Mirrors, its aesthetics, themes, and descriptions are simple -- the most generalized contemporary vernacular fantasy. The enormous influence of Gygax’s ideas about setting are evident of course, but it’s a more cartoon sort of fantasy where the cliches and standards of Dungeons & Dragons have been fed back through forty years of popular culture: anime, video games, and superhero aesthetics all accrete atop and modify the concepts of fantasy adventure that Dungeons & Dragons originally distilled from war-gaming minutia, Tolkien, the Western genre, and pulp fantasy novels. Practically this means that like those “old school” adventures which rely heavily on nostalgia and players who have internalized the implied setting of early editions of D&D, Castle of Mirrors depends on the players and referee filling in details with a shared set of assumptions and cliches. For example, the first likely set of opponents in Castle of Mirrors are “The Kromsmoorim Barbarians” from across the Eastern border (of course - Barbarians are always from an exotic and untamed East...) The adventure offers little to describe them: their leader is named White Wolf, they determine leadership through arena combat, they are known as “The Pack”, and fight to the death. We assume they are all pantherish and mighty thewed, with minds of “of great mirth and great melancholy” beneath wild manes of raven black hair. The only aspect of these barbarians that isn’t a pastiche is their use of giant carnivorous vines to break the castle walls, but that’s not developed beyond stat lines for the vines, and a suggestion that the characters can learn the trick. The vampires, demons, ghosts, “deep dwarves”, and dragon that make up Castle of Mirrors other foes are similar, the lightest sketches filled out largely through stat blocks and implied cliches.

Just as Gygaxian vernacular fantasy helps fill in details and save space in minimalist “old school” adventures, Castle of Mirrors uses the well worn cliches of modern vernacular fantasy to fill the vast gulfs of description, characterization, and detail that one would otherwise need in an adventure of its scope. The question when reading through it is how well do this space saving technique and the abstracted navigation mechanics work?
Typical of the product and its art


Castle of Mirrors is ambitious, and its scope is enormously broad. There are at least two major antagonists, numerous factions, and agenda driven NPCs, many varied locations and a profound number of possible hooks in 38 pages. It's a 5th edition product (using a set of modified house rules) so the adventure is story-line, rather than location or level based, and each region or Scene within the Castle offers several possible stories. While sketched rather then detailed there’s almost as much adventure within Castle of Mirrors as one would find in the 250 plus page campaign tomes that Wizards of the Coast Produces, partially because of the deeply cliched setting, but largely due to the efficiency of the writing found its zonal approach to adventure design.

One can’t help but get the sense that Castle of Mirrors is set up like a videogame, possibly one of those gorgeously drawn Konami fantasy beat ‘em ups from the 1990’s, but still an adventure where there’s a foreground for combat and the occasion alternative challenge using the same mechanics (smash the car with a pipe for extra points) in front of a lusciously painted background that shows a change in scenes. Alternatively perhaps it’s written like a mid-1970’s superhero comic, where each month the protagonist encounters a different supervillain and they fight, all for very serious, but often incomprehensible reasons. This design formula evidences a pretty solid understanding of how 5th edition D&D is commonly played, as a series of moments, which largely involve heroic tactical combat. To its credit Castle of Mirrors doesn’t shy away from this play-style or insist that it’s something else by hanging vestigial mechanics and design about. The characters will move through the dungeon, shifting background with a few key descriptive words, and encounter challenges as individual moments and set combats. There will be bosses of both the mini and mega variety linked to many of the ‘Scene’ based areas within the adventure who represent a goal or mission.

This description may upset some fans of older style play, maybe even some who enjoy 5th edition, who insist that exploration is an inherent and necessary part of D&D or even RPG play. However, Castle of Mirrors is based on a different understanding of RPGs and what seems an astute insight into 5th edition play -- it’s an experiment that offers an alternative method of running a dungeon adventure for a play-style focused on cinematic moments within a more guided story. The abstracted nature of exploration within Castle of Mirrors is refreshing, and it certainly seems like it would provide a quick adventure of the heroic variety, acting as an arena to show individual characters powers and spotlight their stories and interests with a variety of hooks and details that lead to the various places within the Castle.

The way Castle of Mirrors builds relationship between its factions is also a successful adaption of classic adventure, it largely depends on combat as an obstacle, but has a significant number of factions and presents the possibility of alliance and double cross in a more nuanced and less scripted way then one might expect from the contemporary traditional play style. Faction play is at the core of a good dungeon crawl, and while Castle of Mirror isn’t and doesn’t intend to be one, offering this aspect of play is a sign that Runehammer Games understands that there’s more to a good fantasy RPG than tactical combat, even when that's the primary locus of play.


While Castle of Mirrors offers an experiment with an almost map free dungeon, even a map free mega-dungeon, it’s a less successful experiment then it might be. The side view map of the Castle and its environs is decent enough, evocative of the similar map in the 1981 Moldvay Basic Dungeons & Dragons book, and the zonal maps are likewise sufficient enough to provide some vague sense of how the areas within the adventure are arranged.  Yet these maps aren’t especially useful -- movement between zones within the Castle of Mirrors may be accomplished through a DC check, but better spatial sense, especially as to how the sub-zone “actions” are connected, would be valuable. RPG maps aren’t only about managing spatial puzzles, navigation, or providing additional detail, they are a way to condense information so that the referee can understand the relationship between areas at a glance. The maps in Castle of Mirrors don’t do a good job of this, and worse, because they have some (unused) detail gives the impression of the spaces and rooms*, encourage a referee or player to read them as spatial description -- incomprehensible spatial description that confuses the adventures basic conceit of abstract movement between zones. Maps with meaningful spatial content and imagery that showed the linkages between regions, Scenes and Zones without implying specific spatial relationship to exploit would greatly benefit Castle of Mirrors, even if it’s not an adventure about procedural, turn by turn navigation of a fictional space.

These are the maps provided, pretty at a glance, largely useless

This isn't to argue that it needs a set of top down, 10' by 10' square maps in the traditional style, but a visual representation about how its zones and sub-zones are arranged would be helpful, especially when players are still likely to state their goals for movement in directions as often as they will in vague concepts. A map could better show how the overgrown gardens patrolled by lions relate to the surrounding towers and ruined walls, as well as how the muddy pond (with negative movement effects) and tangled trees within are spaced.  Even these vague relationship would give the players better a understanding of the environment that might allow it to exist as part of play rather then mere background.  For example players could decide if they wanted to try to escape the ravening lions by climbing up into the towers or dashing into the pond. As it is written, Castle of Mirrors doesn't provide much indication of how its spaces are connected, making player decisions hard to offer.

A second complaint is that Castle of Mirrors may suggest a variety of non-combat solutions through negotiation and trickery, but it offers little support for them. If it was an adventure for another game this would be less of a problem, some NPCs have scripts for a "roleplaying encounter", many more do not. In a play-style focused on combat encounters one needs to do more work to set off the possibility of non-combat resolutions. Players expect and have been taught that every problem is in need of a cinematic fight sequence, and stepping away from that will take some work.  Here, even when parley is suggested there seemingly no reaction mechanics that would ingrain it in the game and no morale to end combat with survivors.  Runehammer should do more to emphasize non-combat resolutions both narratively and mechanically if it wants them to be an actual part of its adventures.  With only a couple lines of dialogue, but no meaningful description or personality, NPCs and monsters are rarely fit for negotiation and role playing encounters. Without these opportunities the varied factions of Castle of Mirrors threaten to just become "sprite changes" in the waves of enemies that the adventure throws at the players.

Both these issues point to the way a lack of detail plagues Castle of Mirrors more generally. Despite relying on genre knowledge, Castle of Mirrors pushes the boundaries of how much can be accomplished with the aid of cliche, trope and genre expectations.  Denser detail (and I mean sentences, not paragraphs) would help elevate the setting itself from background to something more interactive.  Even if zones act primarily as a backdrop for specific encounters they could be much more detailed then this:

"The giant doors of the upper castle reveal a cavernous room. At center, a table set with candles and decadence that could seat hundreds.Three grim figures sit at the far end."

This is the description for a major scene in the adventure, the space where the vampire disciples of the castle's master are encountered.  The site of a cinematic confrontation. Even if one expects the referee and players to be familiar with the great hall of a vampire castle it's worth providing a few more details: smells, sounds, and a couple of specific objects that help set a scene.  Perhaps the brilliant agate of the floor tiles, the wax from 10,000 nights of candles running from the table to the floor in thick petrified flows, or the cobweb shrouded banners of defeated foes motionless even in the cold drafts that whip through the hall. Something. To belabor the side scrolling 90's arcade game metaphor, even background matters.

This is a background from one of the Samurai Showdown games of the mid 1990's depicting a fantasy Versailles:

There's a lot for the eye to take in and grab onto: the glittering cascades of crystals on the chandeliers, uniformed footmen, the reflective marble checkerboard of the floor, red rugs, curtains lit from huge windows, opulent gold trim, and of course a monumental painting in the background.  This may be a cliched space, but the pixel art gives detail to the vague sense of baroque opulence that one associates with the palace of Versailles. Without significant illustrations*, an adventure need not provide this level of detail with words, but it should offer something to help the players visualize the specifics of the space or else every encounter risks feeling like a battle in a bare white room. Even when we create minimal and trope filled settings, detail is a useful addition that saves the attention budget of the referee.  It may be easy enough to fill out cliches and genre standards implied by a few words, but an active referee has other things to manage besides pulling details from the unimaginative mass of vernacular fantasy to give players some reference points for a location.


Obviously Castle of Mirrors is a very different adventure, for a very different play-style, then a Classic mega-dungeon, but it's still a mega-dungeon in scope.  The numerous hooks and missions within, as well as the overarching secret of its mirrors and vampire master are likely to draw a party back many times. While Castle of Mirrors won't offer even a quarter of the sessions that Stonehell would, it's also a far smaller book and despite a smaller size it feels expansive (perhaps too much so given its lack of detail). This combination of enormous scope and truncated 'key' offers a possible alternative to the sprawling hundred key mega-dungeon, one that can both save player time and designer effort, while perhaps still offering some of the same experience and thrills of adventure within a monumental space. I haven't played Castle of Mirrors, and my 5th edition experience is limited to minimal success attempting to play it in a more classic manner using tools like the exploration die, strict encumbrance and random encounters, so I can't say how well exactly Castle of Mirrors' abstractions function using a typical 5th edition play style but they appear like effective simplifications from my read through. Additionally this isn't a review of 5th Edition play or even contemporary traditional play more generally -- Spectral Interrogatories looks at adventures for usability as a Classic Dungeon Crawl.  Castle of Mirrors isn't this, in any way, but it's intriguing.

For the Classic play enthusiast, the implied space of the zonal dungeon may be shocking, and can be rejected quickly as a total replacement to the keyed map -- it simply doesn't allow the primary form of Classic adventure: the movement by movement, step by step, torch by torch procedural dungeon crawl.  However, while Castle of Mirrors might at first glance appear antithetical to a this style of play, its efforts to pare the adventure down to the minimum it needs for the play experience it wants to deliver is a laudable example for mega-dungeon design, where space is especially important.  Maximalist mega-dungeons exist, but generally the form depends on some degree of minimalism and most often this is accomplished by following player expectations about setting closely.  In this way the Classic mega-dungeon (with some exceptions -- hmmm. What should I review next?) shares design principles with Castle of Mirrors, which strips many standard ideas about adventure design away and in doing so offers two potentially useful tools for Classic mega-dungeon designer.

First, consider the disconcerting adventure by zone. Castle of Mirrors effectively dispenses with discrete keys and spaces which makes perfect sense for a scenario where decisions about how to move through the fantasy space are relatively unimportant. An aberration, maybe an abomination, to the Classic style perhaps, especially with its dependence on an abstracted die roll mechanic (admittedly Castle of Mirrors could offer better ways to modify this roll), but terribly convenient and concise for filling in space.  Even in a Classic dungeon environment not every space needs to be measured or explored with exactitude. Zones can exist in a more standard dungeon environment, filling largely deserted and empty liminal spaces between more detailed, densely keyed nodes. Imagine a ruined city adventure where the streets and tumbled down neighborhoods use zonal keying, but a few larger buildings: temples, palaces, museums, and the grand market, act as node style dungeons.  This would create the feel of a sprawling mega-dungeon while page length.  Like procedural generation, a zone and node approach could be advantageous for filling in large, less interactive areas.

Second, Castle of Mirrors, borrowing from a more elaborate system in 13th Age, focuses on hooks as the primary narrative tool that pulls players towards specific spaces within the mega-dungeon.  Castle of Mirrors zones are not linearly connected, and some can be accessed directly.  This is a hallmark of good mega-dungeon design, perhaps even a necessity, multiple means of access and egress to allow repeated expeditions without wasting too much time traversing the same areas. Tying multiple hooks to almost every zone allows for a greater player drive to access various regions of the dungeon and to move through it in an intentional mission oriented way.  In an adventure without granular and cohesive space, these hooks provide the majority of the adventure's structure, a sense of progress and accomplishment without navigation.  Again however, for a more traditional adventure, especially a large one, distinct hooks for various nodes, factions or locations within the larger hole encourage exploration and searching out the best ways to access various sections of the adventure directly. Such directed exploration presents the opportunity to encourage navigation and reward it with narrative successes that can better motivate players then mere treasure. Predetermined distinct goals become something that characters bring to the dungeon, and almost by necessity relate back to the larger setting, making them sometimes difficult to structure, and requiring at least a selection of interested above ground factions to offer them, but offering far greater player engagement with setting and dungeon.


Castle of Mirrors is simply not an adventure I would run, but reading through it offers a glimpse at a novel and ambitious effort to use the default Contemporary Traditional play-style to tell a more classical exploration adventure.  As with many ambitious projects its seems like only a partial success.  Space saving changes to basic adventure structure haven't been used to expand parts of the adventure that could aid a referee in running Castle of Mirrors chosen play-style, but instead Runehammer has chosen to simply to broaden the scope and keep the adventure short.  More detail, more description, more complex set piece encounters could all add a great deal to the basic design of Castle of Mirrors, perhaps transforming quotidian cliche into something more complex, where tropes are used to tell a novel story, changed slightly, or described well so that the designer is offering an adventure that contains something more then any referee could produce if they'd simply had time to.  It doesn't take much, a few novel details act as tinder for the imagination, a few complexities and description can add wonder to even a thoroughly unexceptional setting. Castle of Mirrors doesn't do this, instead making some bold structural and mechanical decisions but stopping, leaving an intriguing, well-built skeleton without the flesh of memorable locations, NPCs, or monsters.  At best Castle of Mirrors encourages one to borrow some of its interesting mechanics and use them to tell a more interesting story, while at worst it leads to a plodding predictable adventure that a obscures its good ideas and insight into the nature of Contemporary Traditional play.


*A Note on Maps, Public Domain Art and Design. 
Castle of Mirrors uses public domain art, old woodcuts and etchings that have been collaged together a bit and vaguely follow its action.  This is an increasingly popular style of illustration for small press adventures, and it often works.  It's far cheaper then hiring artists, a compromise between vision and resources.  However, in an adventure with a fundamental lack of description, where the description that does exist suggests a pop fantasy world, early 20th century black line of vikings just doesn't work very well to supplement the text.  At best the illustrations are filler and at worst they are contradictory. Using public domain art may make sense, but it comes at a cost.

Using public domain maps however (and its unclear if the maps in Castle of Mirrors are borrowed from elsewhere) is worse.  Maps are something tailored to an adventure and setting, they are the item that referees will likely spend the most time looking at during play and should help evoke the mood of the location as well as make it easier to describe movement through it.  Maps that aren't well incorporated with an adventure (such as those in Castle of Mirrors) risk the same confusion of aesthetic as public domain art, but also add to it a confusion of description.  The map in Castle of Mirrors is hard to read, but appears to include both side and top down views of certain parts of the castle, a castle that has no keyed locations and which contrary to the map can't be navigated through describing how rooms connect, but only through skill checks.  A map may be largely superfluous to this style of play, but it shouldn't actively work against the referee's understanding.  If you are doing experiments with maps, one should take the time to produce a map designed for the experiment rather then borrow one from elsewhere that works only when ignored in the right way.


  1. Thanks for the thoughtful review! It was quite interesting to read given that I've been watching some of Hank's recent youtube stuff and since this adventure was published, he's switched to running OSE and is running a campaign that features a lot of just-in-time creation and procedural generation. I think he's moved on from a lot of these 5e-isms too.

    1. I did read Runehammer's OSE hack and I was struck that it was sort of an effort to snap the system off its Classic/exploration play roots and turn it into a combat and heroics driven Contemporary Traditional ultralight.

      That is of course fine, Ferinale seems to understand and enjoy this style of play, and it's a damn popular one. It will be interesting to see how much Runehammer evolves "from" 5Eisms vs. how much it evolves OSE towards 5e's Contemporary Traditional play style.


Ludic Dreams III - My Body is a Cage

At long last, I turn my Ludic Dreams series to what I probably should have been doing all along: reviewing games and adventures about dreams...