I6 - RAVENLOFT - Review

OH THE CREATURES OF THE NIGHT...

I'm not sure if this is a review or a mediation on adventure design principles, but I've recently been thinking a bit about the horror genre and tabletop adventures - while running a location based wilderness point-crawl with a post-apocalyptic high fantasy setting.  My long running game set aboard the HMS Apollyon attempted to have horror elements, though perhaps failed in that regard, and with that in mind I find myself looking at the first really successful horror themed D&D module - I6 Ravenloft, -written in 1983 by Tracy and Laura Hickman, a few years after two other modules I've reviewed here (somewhat unfavorably): Rahasia (1979 republished/rewritten 1984) and Pharaoh (1980), but before the couple launched into Dragonlance.

To understand Ravenloft, and perhaps to give my critique of it a more appreciative cast I think it's first important to look at the Hickman's adventure design in general.  If Gygax's adventures can be identified by a certain actuarial abundance and callously fair mercilessness (The deadly but entirely rational traps of S1-Tomb of Horrors, or the highly detailed Keep in B2-Keep on the Borderlands), the Hickmans' adventures are about: storytelling and heroic narrative.  They may be the creators of this school of adventure design.  If one should thank Jennell Jaquays for interesting non-linear map design, the Hickmans stand tall in the annals of game design by offering an alternative to the location based adventure and approaching tabletop games with a novelistic eye rather then that of a wargammer.  That is to say that even Rahasia, written at the same time as Keep on the Borderlands, wants to tell a specific story where the player characters are heroes of an epic fantasy struggle and does so with set piece encounters that flow logically into one another.  Of course this "story path" style of play can lead to the sort of awful excesses that deny player choice and freedom to have a say in the story of their characters - the railroads, forced decisions, coercive encounters and  quantum ogres that mark 00's and 10's mainstream tabletop products. The Hickmans' own work isn't usually as bad as all this, it's not even as bad as a contemporary OSR influenced WOTC product like Lost Mines of Phandelver, and the saccharine vanilla fantasy flavor of most of the Hickmans' work is done earnestly, with a certain unique flair and in general far more forgivable in 1979 or 1983 than 2016. Ravenloft is likely the best of their adventure design work, and it contains a lot of interesting ideas that sometimes fall flat or are two big for the module they're in, but are generally not boring. It's a novel and useful effort at design - one that I personally wouldn't run expect under certain circumstances, but would be tempted to steal some ideas from.

Yup, That's Pretty Gonzo

THE ADVENTURE

A storm, a tavern and a note impugning the parties' honor (this hook would not work on my players - but isn't terrible as hooks go) leads through some scary woods and into the pocket domain of Barovia - sealed from the rest of the world by impenetrable mists and packs of wolves. In Barovia things are bad, sinister gypsies roam at will, but the townsfolk are besieged by the forces of their liege - Strahd, the vampire.  He's been more nasty of late - having plundered the holy symbol that protects the town, caused the death of the mayor and made efforts to steal the mayor's daughter for some undoubtedly sinister purpose.  After wandering about Barovia for a bit the adventures will discover it's a pretty tiny and empty place, with some frightened villagers, a gypsy camp and Strahd's menacing castle high above on a cliff.

The mayor's children need help, the holy symbol that protects them and the town has been stolen by Strahd, and the daughter bitten and drained.  The gypsies (while described as evil and in Strahd's service) are happy to assist the party by allowing their leader Madam Eva to read the characters' fortune and give clues to how Strahd can be defeated through discovering the magical treasures in his castle.

The castle itself is a dense dungeon with plenty of creepy encounters and ominous signs.  A few dangerous combats and deadly traps (mostly of the trick variety) block the adventurers from reaching their goal, combined with comically sinister encounters with Strahd's human servants.  It's a creepy castle that at its best feels a bit like Scooby Doo or  Abbot and Costello meet Frankenstien (though it seems to lack a Frankenstien monster) and at its worst sinks into bathetic 80's horror nonsense (which is fun in its own way).  That is to say that Castle Ravenloft is a bit of a gonzo funhouse, but it holds together well with the trappings of B-horror, far better then most funhouse dungeons.  The real conflict inside the castle is between the party and Strahd, who will hunt them and harass them with several attacks, putting pressure on the party as they hunt him down and possibly try to find the artifacts that will make killing him easier.

THE GOOD

Ravenloft is steeped in flavor - perhaps the campy flavor of Hammer Films Dracula, but that clear nod to obvious horror movie cliches and the gonzo feel that goes with it is both fun, and provides a good potential well of pop cultural knowledge for the GM to mine for description and improvisation.  Regardless of the cliched nature of the Ravenloft setting - foggy wolf haunted woods, evil gypsies (rather insensitive to the Roma people, given their history, but mercifully so damn cliched that it's obvious farce), vampire brides, and Strahd himself - it holds up and it's atmospheric.  I'm not usually a fan of the Hickmans' tendency towards melodramatic scene building with hoary pop cultural props, but in Ravenloft I can't really complain, because here the red satin interior of Strahd's cape allows allows the players to easily visualize and enjoy the change in setting.  Likewise the protective (or railroading) mists of Barovia would normally be a very nasty bit of GMing in my opinion (note that nothing traps players in Death Frost Doom or the Tomb of Horrors - characters can retreat anytime to find an adventure they like better), but here they work to enforce the idea that Ravenloft is a genre shift and to separate it from the regular D&D setting.  The NPC in Barovia seem to do this fairly well - the hollow zombie like tavern keeper for instance is a nice touch and the nightly procession of ghostly adventurers are quite effective.

I suspect a horror game must perhaps have this break, in the same way that the traditional break between overworld and underworld signifies that it's time for the players to become cautious and scrutinize the environment, the break between normal game setting and the horror setting - where danger comes from new sources and a ten foot pole, a flask of oil and a block of lard might not offer an optimal solution, is a useful signal.  Most D&D style fantasy games (especially of the era Ravenloft was written) are exploration based and focused around mechanical or diplomatic solutions to problems.  Yes, these solutions often involve the use of fantastic tools such as magic, but they also are fundamentally rational.  There is a logical puzzle element to most D&D adventures of the classic dungeoneering sort that is lacking in Ravenloft.  Strahd may be defeated by the clever use of tripwires and a pail of blinding caustics, but the adventure itself certainly doesn't countenance this kind of solution - the castle is to be explored, grim and forbidding atmosphere dripping from its walls and the clues of the old gyspy's card reading will reveal secrets that can aid the adventures in their final confrontation with the vampire - but not until he manages to make several hit and run attacks, some of Strahd's melodramatic story is revealed and the dangerous of the castle manifest themselves with a proper level of creepiness.  The players and characters are expected to experience the haunted mansion story of Ravenloft and unravel the story clues to find the artifact items within that will make defeating the vampire easier. While I have a strong dislike for cliched and hackneyed setting elements, the classic horror feel of most of Ravenloft is set off by enough D&Dism that it still feels more of a tongue in cheek gonzo setting adaption and avoids the dreariness of the default oppressively standardized world of D&D's pseudo-Tolkien vanilla fantasy.

While Ravenloft exists with the goal of evoking a certain kind of horror movie story and culminating in a heroic ending after a proper build up of dramatic tension, it's not purely about telling this story.  What saves Ravenloft as an adventure is that it tries to tell a type of story, not an entirely specific one -  it doesn't resort to the cheap tricks of that many later story based adventures do to make sure everything follows a specific plan to a certain ending.  The characters may be more or less forced into Strahd's castle, but once there events can unfold as a collaborative effort between GM and players - not a chain of predetermined encounters. It's not a linear adventure within the castle, and the randomly generated elements encourage a GM to improvise and fit together it's pieces organically rather then following a script. For example - while Strahd has certain set types of attack that he can use once, but he doesn't appear at obvious predetermined spots - except for his "lair" location determined randomly with the 'card reading' clues.  The artifacts that can destroy Strahd are likewise distributed randomly and clues provided - though without knowledge of the castle layout it's hard to see how these clues will help.  Still, even these artifacts aren't strictly necessary to kill Strahd, though with his vampire abilities he's plenty resilient.

Damn Nice Maps

Ravenloft has a complex map and a large number of keyed spaces in the castle itself, though most are minimally described despite the use of boxed text, and the actual "dungeon" portion of the adventure takes up only a modest portion of the adventure.  The remainder is the typical new monsters (though only a few, arguably less interesting new monsters are described), magical treasures and procedural generation tools that create the Madam Eva's fortune and impose the story elements on the adventure.  As mentioned above the randomness of these elements is a very positive way of creating the story here.  By providing a random set of several possible goals, strategies and locations for Strahd the GM must do the actual work of stringing together the story, constrained by the various elements the adventure provides.  This assures both that there is some improvisation, that the details of the adventure will flow more naturally from exploration of the castle location, and that the story will remain a vampire story.  Had Hickman just provided one possible story and set of possibilities I suspect Ravenloft would contain a lot more scripted encounters and situations with one correct answer or solution predetermined by the needs of story and the desire to make the story follow a proper narrative arc.  While Ravenloft assumes a certain story, where the party will hunt for relics and use them to defeat Strahd - ultimately hunting him down and killing him in his tomb, a clever group of players and GM can use the elements of Strahd and his castle to tell other stories - which is good.  It's worth noting that even many of Strahd's 'victory conditions' don't necessarily result in the death of the entire party and that there is plenty of room for Strahd to remain in the campaign if he isn't defeated. 

As a location based adventure, the individual encounters within Ravenloft are generally good (as they are in most of the Hickman's adventures) and include some great unique monsters (the gaze attack wielding portrait is one of my favorites), despite the use of box text to describe them.  This box text is generally harmless, though it has the problem of box text more generally that it encourages players to ignore room features, while offering a space for confusion and disconnect for the GM when the box text does a poor job of explaining the room, or the GM text afterward fails to follow up on apparently key room features within the box text. Ravenloft's descriptive text (box or otherwise) is rather sparse in most areas, but generally sufficient, and rarely slips into the room keying sin of telling all sorts of details that the players will never learn.  A bit of this kind of detail on NPCs and monsters is useful (and provided by Ravenloft), because it helps the GM decide what sort of roleplaying encounter will occur when the characters don't jump straight into combat with everything - though Ravenloft seems to default to combat encounters for most of the castle's denizens.

Beyond description, the map of the castle is excellent, a real 3d space and helpfully annotated with the distance down chimneys and open stair cases.  While this seems to assume that there will be plenty of pushing people off battlements or down the stairs, it's a useful set of annotations that implies that the space of the adventure really is open to player imagination and exploration - something that many maps could consider.  Ravenloft's writers (or map designers at least) have considered a party that wants to abseil through windows or shimmy up and down chimneys. Both map and description have the advantage of feeling like a 'real' space, and a mostly logical layout that helps define the uses of rooms and the potential locations of zones (The crypts and dungoens are downstairs, the library upstairs etc..).  This sort of organic adventure design is useful in a treasure hunt adventure like Ravenloft, and generally beneficial to exploration adventures as it encourages players to think about the space the game takes place in.

THE BAD

Most adventure designers have their little obsessions and tells that mark a design as their own coming through as funny mechanics, repeated tropes or even odd phrases.  Some of these are charming, some meaningless, but a huge number of them are exasperating.  The way Gygax likes to neuter certain spells or turning to make an area present the 'right' type of challenge for example.  The Hickman's have these as well - a Ravenloft has it's fair share of them.  From painfully bad verse and overwrought prose (the verse at least is limited in Ravenloft) to certain demands on the GM and players that exist solely so the adventure can have certain story moments. Everyone in Barovia is also at least a 2HD creature, with the local simpleton a 9th level fighter whose leather shirt is apparently the equivalent of +1 platemail.  This of course is to prevent players from taking their characters frustrations about being trapped in the adventure on the good citizens of Barovia, The first of these "Hickmanisms" is forgivable, the second laughable, but the last comes annoyingly close to one of the great mistakes of bad GMing - using the players to tell the GM's story rather then letting them tell their own.   In Ravenloft these little demands aren't overpowering, but they start at the very beginning of the adventure - with the demand that "At least one of the PCs must be a fighter with a longsword".  As demands go this is nothing like the plot protected pregenerated characters of Dragonlance - but it's the sort of thing that sets my teeth on edge, because it doesn't respect the GM's ability to adapt or the players ability to define their own characters. This demand isn't alone - there are others, such as the admonition that the quest giver Madam Eva can "is never fooled by adventurers". A little thing to complain about perhaps, but the need for a special PC to wield the Sun Sword is a marker of an attitude towards adventure design and play that puts the GM's desire to tell a specific story at the expense of the players. 

The Art has a Certain 80's Flare.

Ravenloft has many of the hallmarks of a railroad, but underneath it's not really, it may make some ugly efforts to trap the characters in Barovia and force them to explore Strahd's castle with the firm assumption that the only way to escape is to kill the vampire, but Ravenloft backs off at an appropriate point - trusting the narrative alchemy between a good GM and an engaged group of players to tell all but the highest level of story in the adventure.  Still there's something a bit unforgivable about trapping the party in Barovia with unavoidable magical mists - better to my mind to make Strahd a power, a growing faction that if ignored will have his way and make other efforts to muck up the game world.  Though Strahd's memorable nature and power present a danger themselves that Ravenloft seems to encourage.

One of the ways that Ravenloft creates a novel adventure is the inclusion of a villain who the plot and all the action focuses around.  In a way, like a horror movie, the adventure is about Strahd, not the players.  The adventure constantly asks the GM "What will Strahd do?" and encourages the GM to run Strahd like a player character.  This is both interesting and dangerous. 

It is interesting because it's somewhat novel and provides a unique adventure - its dangerous because it encourages adversarial play.  GMPCs, either Mary-Sues that exist so the GM can play along, special superpowered types that swing down ex machina to fix problems or explicate plot, or villains that must continue to prosper and thwart the players until the right climactic moment all detract heavily from the game because they remove player agency and attempt to force a specific story.  This isn't to say that a smart powerful villain like Strahd can't make for a fun game, but the moment he crosses that invisible line between an NPC and a GMPC the game becomes an adversarial mess where the power of the GM is aimed at thwarting the players rather then adjudicating their actions.  In Ravenloft's defense there are several mechanism included in the adventure that blunt some of the GMPC impluse with Strahd - his goals, location and schemes are controlled by randomly generated results, and there are certain rules for playing him designed to prevent him from being unstoppable. All of these limitations are good, I just wonder if they are enough - especially in light of other reviews of this module I've recently read and the admonition to treat Strahd like the GM's character.  For a bad or inexperienced GM Strahd is an excuse to run an adversarial game and feel a sense of accomplishment in wiping out the players' characters.

The way that Ravenloft encourages adversarial play with Strahd is a dangerous enough enticement towards "Killer GMing", but it's not alone in Ravenloft.  There's a certain thoughtlessness and callousness about the adventure's traps and trick.  The drawbridge for example is a low chance trap (automatic 5% with a Dexterity save) that causes automatic death - it provides no warning and no real discussion of how something that deadly (if inconstant) can be avoided.  Compare this with the traps in S1-Tomb of Horrors which are often lambasted as unfair and cruel.  The false entrance trap in Tomb of Horrors is easy to detect - investigating the ceiling will spot it, it's easy to trigger and avoid, and it's not lethal (It does a lot of damage, but an amount many 10th level PCs can survive, certainly not 100D6). Most importantly though, by clearly noting in the text how the Tomb of Horror traps can be detected, disarmed or avoided Gygax builds a wall against the careless or murderous GM. Moreover the traps of S1-Tomb of Horrors are the adventure's entire point, it's a puzzle dungeon full of traps - the Ravenloft bridge trap (and some of its other traps such as the teleport traps) are rather undetectable and appear without warning or seemingly much thought.  Some of them (the iron golem puzzle for example) are better, but there's a general lack of information provided in the adventure to really run any of these traps as proper puzzles rather then seemingly arbitrary ways to kill a character or two.  In most adventures this wouldn't be much of a problem, something a good GM should know how to handle, but with the way that Ravenloft encourages player and GM competition they present an easy stumbling block.  

In general there's a certain lack of information about the rooms in Ravenloft castle - the descriptions are often quite bare, and while the excellent maps help, and what description there is is solid many mechanics and the way things work are a bit fuzzy.  Perhaps this is a result of the location taking a subordinate role in the adventure to the conflict between Strahd and the party, but it seems like a lost opportunity.  Ravenloft feels like it's too big for the 40 or so pages it contains, and the writers have skimped on description and well thought out areas rather then reduce the scope of the adventure or size of the castle.  The huge scope of Ravenloft is either a noble decision to encourage GM creativity, or a failure to present an adventure that's playable without excessive work, though given its novelty and the era Ravenloft was written I'm inclined to root for the adventure and applaud its audacity

HOW I'D RUN RAVENLOFT

Ravenloft Box Set -The Ravenloft Setting did not Become Less Gonzo Over Time.

It seems that while Ravenloft would be hard to reskin without losing a lot of its charm (beyond a bit of atmospheric chicanery or a cultural gloss other then Eastern European - perhaps a hopping vampire kung-fu movie treatment instead of Hammer Horror?) because its so tied to the elements of Classic 1950's Vampire horror.  Classic Vampire horror is of course very silly - which means that Ravenloft, for all its gothic gloss is a gonzo adventure about running around a spooky castle.  In this light the bombast of the Hickmans' writing and world building becomes quite acceptable - and in Ravenloft I think they know it, including several set pieces that are meant to be silly and numerous terrible puns.

Who am I to deny this gonzo impulse, and it seems to me that gonzo absurdity is one of the best ways to run horror scenarios as a tabletop game.  Horror is meant to be taken seriously, to build tension that turns into fear, and then to offer release through action.  In a D&D game this means that the normal impulse to joke around will be emphasized if the game succeeds in building tension.  Trying to limit the humor and joviality of ones players will fail and likely result in the sort of mock seriousness that tends to make tabletop games even more absurd.  Better to embrace the gonzo humor and use it as the counterweight to gruesome description, sudden player death and the fundamental absurdity of telling a story about exploring a vampire's castle.  For these reasons I'd use Ravenloft in my favorite gonzo setting - Anomalous Subsurface Environment.  Ravenloft makes perfect sense in the Land of a Thousand Towers where insane semi-immortal wizards rule pocket domains through magic and superscience. 

Strahd is another crazy wizard - one that happens to be a vampire with a well established, but poorly run little domain somewhere that brings him into only occasional conflict with the city of Denethix.  Just placing Barovia on the map and leaving some clues in the vampire themed town of Lugosi should be enough of a hook to keep Ravenloft in the game. I'd blunt Strahd's nastiness a bit - making him more of a schemer and less of a dungeon dwelling monster.  Strahd's hand would start to appear in the form of adventuring parties sent into the ASE to find things Strahd wants, and eventually he might even invite the campaign's characters to his domain for lucrative work.  When the players eventually turn on Strahd the entire edifice of Ravenloft is there to build off of. 

I'd also turn the gypsies into anthrophagic ASE elves - no one needs ugly gypsy stereotypes in their game in 2016.

Beyond this light reskinning (more the finding a proper game world for Ravenloft then anything) I think the GM advice in the adventure is bad - it seems to me that most of the troubles I highlight could be fixed by cautioning the GM not to think of Strahd as the GM's character, but as a powerful foe to run well, noting not just his strengths but his weaknesses and failings.  Likewise a note about the castle traps being detectable and a general admonition to discourage adversarial play and encourage the idea of character competence (also assuming that characters have made the obvious precautions - especially in situations where the players lack specific technical or situational knowledge that the characters would have) would be far more helpful then the advice included.