Ravenloft has long been my favorite official D&D setting, but that doesn't mean that I don't think it has problems or doesn't have areas where it could be improved. Over the years I've suggested a number of alterations that I would make to Ravenloft, were I to be placed in charge of it; consider this the second post in a "best of" edition of my Ravenloft Remix. The first post in the series is here.
2) Create cohesion
It has been rightly pointed out that Ravenloft's core feels more like a patchwork of ill-fitting domains than a cohesive setting. Originally, this was by design. The Ravenloft setting was built around the idea that it was a place that D&D characters would visit for a session or two, and then return to more prosaic settings such as the Forgotten Realms or Greyhawk. The intention was never to create a setting with organic unity; rather, each domain was more or less a separate playground oriented toward a specific Gothic trope that had little connection to a larger, encompassing conception of the setting.
As it turned out, fans of the setting actually did want Ravenloft to feel like a self-contained setting in its own right. TSR shifted its presentation of the setting to accommodate this perspective, but the results were mixed. From the Domains of Dread hardcover onward, TSR changed the focus from characters sucked into Ravenloft through the mists for a couple of horrific adventures before returning to the vanilla fantasy grind to characters who were born and bred in Ravenloft, and thus had actual reasons to fight against the evils haunting those lands.
The 3x versions of the setting published by Arthaus followed that path, but unfortunately added a lot of questionable detail while also having to work under some unfortunate restrictions, such as a Soth-less Sithicus. 4e floated some mediocre "domains of dread" articles, then the planned Ravenloft product got shelved. (I'd love to see a draft of it, but frankly it sounded awful.) 5e has walked back from the notion of Ravenloft as a "native setting"; the Curse of Strahd adventure assumes that the player characters are abductees from the Sword Coast because of the edition's Realms-centric approach.
If I were rebuilding Ravenloft, I would refocus on making it a cohesive setting with creating characters native to the core as the assumption--with the option of visiting characters entering through the Mists. Here is where I would apply my efforts:
- Reorganize the geography based on theme. Ravenloft's geography is already malleable; the shape, arrangement, and inclusion of domains varies wildly over the setting's 2e-era history alone. We could take advantage of that by remixing the placement of the core's domains to make a more coherent cultural landscape. Since so many of the domains are based on themes drawn from the real world, that could be a helpful guide when it comes to placement.
- Expand on the themes already extant. There are great themes built into Ravenloft's domains, but most are woefully underutilized. For example, Bluetspur was clearly meant to be a place to bring in the possibility of Gothic Lovecraftian adventures, but in execution it is totally barren and really only functions as the setting for a single adventure module. (Thoughts of Darkness, which also happens to be a pretty terrible adventure.) What Bluetsput really needs to capitalize on the promise of Gothic Lovecraftiana is Innsmouth-style fishing villages, weird cults, bloodlines infected by the kuo-toa, and an Arkham analog where scholars have been collecting horrid knowledge of the Far Realm. Every domain could benefit from expanding on its thematic elements in a similar way.
- Populate the setting. It is a common sentiment that Ravenloft is far too depopulated to sustain a variety of scenario types. My solution is to add more high population density areas. The approach often discussed on the Fraternity of Shadows board works well: multiply the population figures for the larger towns by ten and make them into cities. Mix urban and rural areas for a greater variety of adventuring possibility.
- Add more interconnection. There should be more movement between domains in Ravenloft, and a greater sense that the domain borders are most often permeable to the land's residents. Make trade and commerce between domains more prevalent. And definitely add reasons to travel by sea! Ravenloft has always badly wanted to feature Gothic Nautical adventure, but the current geography makes it seem insane that anyone would build a masted ship just to trade with a far-off island with a few hundred inhabitants. One way to do this would be to include closely guarded secret paths through the sea Mists that are sought after to secure important nautical trade routes.
- The supernatural is not hidden. The point of Ravenloft is to mix D&D with Gothic conventions, but this sometimes gets lost in the weird restrictions intended to enforce the Gothic side of that genre admixture. (Which doesn't work for the Gothic anyway because it is such a migratory and mutational mode.) For example, those narrow lists of appropriate monsters? Gone. Use any monsters you want. If you can't figure out how to describe a green dragon as a primeval horror of the eldritch forest, you need to learn more words. Strahd is a vampire and everybody knows it; none of this silly "the peasants don't really think about why he doesn't age and is never seen during the day" business. In a world with clerics and wizards, vampires are not going to be something unheard of--lean into the rich supernaturalism of folklore, not away from it.