Osteomancy: Design With Me

Part of the Magical Practices Project

Part of the Osteomancy series

This is a little bit of a weird post to be writing for a few reasons. First, I don't know what I'm actually writing for this – I don't have a specific end design in mind as I did with the Spider Broodmother, where I had the basics of the stat block I wanted down before writing. I've been struggling with actually building out the Osteomancer rules for a while, so I'm going to walk you through my current frustrations and maybe, just maybe, I'll have a breakthrough in the process of designing and we'll end up with an actual product. Or maybe we'll end up with a pile of junk and you'll just see an important lesson about homebrewing: "a lot of what you write is garbage."

Second, this is weird because it's part of my secret project that I referenced back in the Osteomancy history post. I am in the slow process of moving away from D&D and migrating to a homebrewed system using a mashup of the Sparked by Resistance framework and the Year Zero Engine, which I haven't dedicated an article to yet. I will! But I want to get at least the bones of my system in place before sharing my thoughts on the framework or too many details of my system based on it.

Third, this post won't end up with a final product for download. This is a "thought process" post. Whatever gets designed here today will go into the larger binder of my new homebrew game system to be revealed at some later point. Sorry about that. Feel free to skip this article if it completely does not interest you.

If you're still with me, without further ado, here are my thoughts on actually building an Osteomancer "class" for that secret homebrew system.

The Problem With Prophecy

There has been a lot of digital ink spilled about the difficulties of prophecy in RPGs. While they are often a staple trope of fantasy, putting a prophecy in your game can feel very limiting. Either, you're setting up a railroad – something that is going to happen, and your players have no impact on it – or, even well-meaning players might so radically diverge from your expectations that your prophecy goes unfulfilled.

Letting players see prophecies at will--such as a class feature--is even more difficult. Putting this power in the hands of the players can be fine for a story game or something heavily on the "narrativist" end of the RPG spectrum, but for GMs who want to craft a coherent narrative with secrets and reveals and pacing rather than having everything be just an emergent, player-driven sandbox, giving your players the power to predict the future is even more tricky.

Are you just going to reveal plot points? That's what I've done for a future-seeing character in my home game, and it is how D&D handles it with spells like Augury, where a player can ask the DM about the likely outcome of an event. This works most of the time but is still tricky. A player cannot ask "will our plan work?" or "will my fellow PC die?" because the DM doesn't know that; it depends on rolls. It also means that my plans have to be really solid; usually only for the session, since Augury is focused on the short-term, but if I plan on having a more improvised session, or if the players go against my initial plans and I'm winging the resolution, I can't provide accurate "prophecies" for the player who is scrying into the future.

D&D's other solution is Portents or similar effects – using dice manipulation. The idea is that, instead of a specific prophecy like "you will succeed at robbing the manor," you have certain effects that allow you to impact the results of future rolls. You can then flavor these effects as "oh, I saw us succeed in picking that lock." But to me, this does not feel very... prophetic. Mechanically, it is the simplest resolution, but for a class themed around seeing the future as our osteomancer will be, I want there to be something more focused on actually getting a glimpse of future plots, not just future rolls.

Lastly, there's the question of prophecy generation. In my ideal world, a prophetic character would be able to have some sort of mechanized prophecy feature. They shouldn't always have to go to me, the GM. While this works in my campaign currently, one of the complaints about a lot of D&D subclasses like illusionist is that they are so GM-dependent to work. I don't want my system to be easy to spoil by a GM who is unwilling to give a peek at their notes or who plays a more improvisation game than me; I want the player to be able to generate their own prophecies, in a way, independent of the GM, like dice manipulation does. However, I also want to be able to do that without turning into a mess where the player says "the duke will betray us" as their prophecy, but I have in my notes that the duke is exceptionally loyal and trustworthy, and all of a sudden there is a conflict between the character's identity and the prophecy a player made up.

Curse of Strahd

I have a lot of problems with D&D's Curse of Strahd adventure, primarily around the racist implications of it – more on that also to come later – but one thing that they do exceptionally well is foreshadowing using prophecy.

For those unfamiliar, one of Curse of Strahd's rather racist caricatures is of the "Vistani", a Romani stereotype of strange travelers who ward people off with the evil eye. But the Vistani play a very important role in the setup of the adventure. The players have their fortune told by a Vistani using a "Tarokka deck" – basically, a fantasy tarot.

But instead of simply being fluff, the tarokka reading has a massive impact on the campaign. The story of Curse of Strahd involves seeking out three artifacts to defeat Strahd, a vampire lord. Three of the cards in the tarokka reading determine the location of these artifacts, while another determines which NPC is a secret, powerful ally, and the last determines the location for the final showdown with Strahd.

In practice, this means that the card reading near the start of the game can actually feel like telling the future. The reading can provide instructions for the players, it automatically makes each playthrough different which is cool on its own, and it makes the reading actually "tell the future" in a way because it directly shapes the future.

That is the feeling that I want to capture for my osteomancy class. The only problem is that it can't be generalized. To replicate for a campaign not set to the narrative of Curse of Strahd, I'd need to come up with campaign-specific meanings for each card, for any given reading. If I want my osteomancer to be able to take a reading about any question they have as a real fortune teller would, I'd need them to pose the question, take a week off for me to generate 50+ prophecy options answering the question, come back and have them draw and then only use 3 of those options. That's not efficient or practical. So, how can I adapt this for regular play and use, and also make it bones-themed to fit one of my osteomancer traditions?

Basic Concept

So here's what I'm thinking. I think I want to pull mostly from throwing bones, the African tradition I described in the history article rather than the European deterministic dice (less versatile) or the practice of scapulimancy (harder to imagine how it connects to a dice-based TTRPG).

Our class will have a core ability of foresight, where you roll a collection of bones and they will tell you something about the future. The Phansi Museum writes that "four principal bones which represent family and a number of supplementary bones each representing a part of society" are used in bone throwing, so I think we can borrow that as our basis with supplementary bones added on.

Something that I did for my current prophecy-seeking character in my home game was classify three tiers of visions. This character could receive nothing, portents/dice manipulation, a minor vision (augury), or a major vision (me revealing a scene from my notes). This sort of ranking, I think, works well as a mode of making sure that visions feel distinct and at different power levels. In fact, I've liked it so well--and the three-tier system is central to the Year Zero Engine--that I'm making it central to how abilities work in my new homebrew system.

So, for our Osteomancer, we can similarly create 3 different tiers of omens.

Tier 1 Omens

Our lowest tier of prophecy is some form of dice manipulation.

When trying to research how other people had done prophecy in their games, I came across Apothecary Press's Oracle class for D&D, available for free on their website and linked directly here. The Foresight feature, which they describe as "portents but different" offers a sort of option for a dice manipulation feature.

My homebrew system is going to be a dice pool system, rather than something with modifiers, so I will have to adapt the Oracle class's features. Spellcasting in the system--in its current form--will involve rolling to see whether the spell is effectively cast or not and falls back on the system's core mechanic, so for this design, we're assuming that the Foresight has already been successfully activated. There's no need to roll to see if we can foresee anything.

Here's my gut instinct for a base foresight feature:

Foresight: When you activate this trait, you throw your collection of bones and gain a momentary glimpse of the future. Choose a target character for the bones to focus on. Roll 4d6 and consult the following chart. Then, you must choose a situation that your glimpse applies to related to that character.

I made the following chart. It skews positive (more likely to turn 2 failures into 2 successes than 2 successes into 2 failures), but I think that is ok – after all, the character already had to succeed on the roll to activate the spell.

Then, the player chooses a scenario for this effect to apply to. This can be a particular skill--"when Alice next makes a Sword attack"--or a more general scenario: "when the dragon attacks Alice, its first hostile action."

I need a restriction to avoid this being spammed. I don't think it makes sense to have this just expire based on an amount of time, as that feels like visions will often not come up in time, which will rob the osteomancer of feeling like they actually foresaw something. Instead, I think it should just be: "any character can only be under the influence of one Foresight at a time." So Alice can't have a bonus to her sword attack from one foresight and her own magic from another; only whichever the most recent foresight was will apply.

Ok! I like this. It's basically D&D's portents ability but adapted to my new rules system and with enough of an extra condition (naming the character/skill or situation ahead of time) that it will feel a little more prophetic.

Tier 2 Omens

I want the second tier of throwing the bones to be something more significant--more on the lines of an Augury. It should be something that will likely be relevant fairly soon, and which will provide guidance for the party in determining how to act.

Honestly, I haven't hated D&D's augury when my players have used it. Yes, it involves some amount of prep work from the GM, but with the right questions, it does not involve that much extra prep burden. My main issue with augury is the limitations that it must be “about a course of action that the party is about to take.” Generalizing this to allow questions about NPCs is easier for the GM, I’ve found, because it allows me to answer questions about which I have more control.

I still do want some sort of "throw the bones" element to this tier as well. Maybe there’s a way to tie the questions asked to a dice roll?

So, tweaking augury with these two things in mind (a dice throw and easier questions for the GM), here's what I'm thinking:

Augury. Roll 4d6 to seek out answers in how the bones fall. For every 1-2, you get a question with the answer “yes.” For every 3-4, you get a question with any answer. For every 5-6, you get a question with the answer “no.” Then, you may ask the GM a series of yes-or-no questions. For each response, mark off one of your available responses from your roll. When you are out of either pool, you can no longer ask questions. You get a minimum of 1 question.

This gives you a maximum of four questions to ask, which is not an unreasonable number, and requires you to pick good questions to not run out of either answer. In practice, if you do not roll all 3s and 4s, it is unlikely you’ll be able to use all 4 questions.

This has not been playtested in this form at all. If you try it out, let me know how it goes! But I think that these rules accomplish an Augury sort of ability, something that allows you to get information from the GM. The broader type of question, to any yes-or-no question, can make it easier for the GM because they don’t necessarily have to predict the future course of the game. And it ties the mechanic back to rolling 4d6, like with the tier one omen, to replicate our idea about “throwing bones” to discern the future.

Tier 3 Omens

The most powerful omens are things that should shape a character's arc. This should be almost Curse of Strahd-esque, where there are multiple facets to the prophecy being generated with different elements being there to inform the players of a direction for the campaign to take. Plus, continuing the same "picking a character to target" idea from the previous tiers, it fits into my established worldbuilding well--the goblins, who use this type of magic, use it partly to name their children for some event that the child will accomplish in their life. Major life-shaping omen territory.

For this, the GM and the player must work collaboratively to generate such a major prophecy. If it were just the player, then the GM might accidentally (or on purpose, depending on the GM) never have the party encounter such an important situation, completely neutering it being a valuable feature.

Here's what I'm currently noodling with. I don't love it – if you have advice on how to improve this, please please please make note of it in the comments! (Anyone who has signed up for the site should be able to comment). I'd love to hear your thoughts.

Prophecy: When you activate this trait, you throw your collection of bones and gain a momentary glimpse of the future. Roll 4d6, making sure that you identify the order you're rolling in. Each individual die result will have an effect.

For each die result, the player should consult with the GM to generate 6 options per question. I think, ideally, these should be generated before activating the trait.

The first die will determine the character of focus for the prophecy. So, for example, you might pick 1 to equal the caster; 2, 3, and 4 are the three other members of your party; 5 is that NPC you have your eye on; and 6 is the local ruler. These choices are up to the player, ultimately.

The second die is an action or activity. These can be detailed or general. For example, "fight a Large creature," "lie to someone," or "meet a dragon." This is the "if/when" portion of the prophecy. Again, the player generates 6 options--things that they expect should happen soon, potential courses of events--and then rolls to see which one comes up.

The third die is a risk. Examples for the options for this die could be "meet an enemy," "find a hostile beast," or "clash with nature." All of these should be challenges that the party would have to overcome somehow.

The fourth and final die is a fate. For this one, the player does not have to generate anything. This simply tells you whether the omen is good, portending success against the risk, or bad, portending failure. The values here are 1-2 = 1d4; 3 = 1d8; 4 = 1d12; 5-6 = 1d20.

Then, assemble the pieces. If [result from first roll] does [result from second roll], they will [result from third roll]. The DM's job is to interpret the prophecy and make it happen if the result of the first roll decides to do the thing from the second roll; it is not set in stone that this will come to pass. The fourth die is the bonus given – in the scenario described in the third roll, the target (from the first roll) adds an extra die to the pool equal to the result generated by the fourth roll.

So, as an example: If I (the spellcaster) lie to an ally, I will run across a hostile beast. The omens are bad (5), and so when that happens, I will have to add 1d20 to my dice pool for the duration of that ensuing situation (the fight with the hostile beast), increasing my odds of failure.

As a player, I still have agency. I can avoid lying to an ally so that the prophecy never comes true. As the GM, I would need to keep track of generated prophecy conditions, but then if the trigger happens, I can just make the risk happen as well.

The GM and the player should consult on the list before rolling anything to make sure that all the options are things that the GM thinks would fit into the plot fairly easily so that it does not require wild detours to make this feature work.

This mostly accomplishes my goals for the feature/class. It does not require a ton of pre-planning from the GM; instead, it can guide and inspire them. It provides guidance and agency to players in navigating their situation. It creates details of future plot events for the players to know about, essentially "foreshadowing" like Curse of Strahd. The one area it does not succeed in is that a poor GM (or one who is simply uninterested) could pretty easily avoid the scenario generated, but I don't see a way to accomplish that with a trait about discerning the future.

Wrapping Up

There are still several steps for this product. As the larger system develops, I need to make sure that it still works mechanically. I need to playtest this set of abilities, specifically, to make sure that it does not create too much burden on either the GM or the player (particularly with Tier 3, I’m worried about the amount of prep required to generate the tables for each roll). And I need to work on tightening the rules language to make sure that everything is clear without being super wordy. I already think some of these abilities are too wordy and unclear without the blog post explanation, so I need to figure out how to fit everything into a more simple description.

If you have any ideas on how to improve this–any of these three options, in any of the capacities I just mentioned—please put those ideas or reflections or opinions in the comments! How clear is this? This will likely need several revisions still before it is in a ready-to-go shape (especially as I tweak the core mechanics of this system).

Thank you as always for reading and supporting Veritas Tabletop! Be sure to subscribe to the blog for updates on my homebrew system as it develops, as well as for the normal homebrew, advice, and history posts.

Next week, I’m planning to adapt these three abilities into something D&D 5e-compatible. Working it out for this has actually helped a lot in figuring out how to make something like this work for D&D. Thanks again for your support!