Polyhedral Pantheons: Sanity Check

At one point I thought 32 gods as a pantheon was a lot — for an RPG setting, at least. Even just twelve ‘core gods’ seemed a little high.

I know that real world pantheists often had hundreds of gods to keep track of.

Also, four to six domains per god, and each domain being associated with four to six gods.  I decided to take a quick look at a published setting.

I didn’t have any large Forgotten Realms references handy, but Complete Divine from Wizards of the Coast shows a fair number of the Greyhawk pantheon… where ‘fair number’ is

  • 23 core gods
  • 24 secondary gods
  • 29 monstrous gods

So far, it looks like twelve, twenty, or thirty-two are all decent numbers to start, and still leave a fair bit of room for obscure gods.

Examining Gods

I decided on a bit of simple analysis — number of gods with each domain, number of domains per god.

Core gods

Complete Divine identifies 23 core gods.  They mostly have five domains each, with six having only four and having six.  The rank (lesser, intermediate, or greater — no demigods in this group) doesn’t seem to have any bearing on the number of domains.

The number of gods with each domain is much less regular.  Only one has the Fire domain (Obad-hai, if you’re curious), one has cold (Bahamut), one has Competition (Kord)… only one has Healing! (Pelor).

On the other hand, eight have Evil, seven have Good, seven have Law, and six have Chaos.  Six have Protection, eight have Trickery, five have War, five have Luck.

Clearly there is no real balance in how the domains are allocated here, nor necessarily any variety once you get there.  Not that balance in numbers is truly important, but I do like to see it.

Secondary Gods

I see 24 secondary gods.  They mostly have four domains, a couple of three, half a dozen have five.  Again, the rank doesn’t seem to have much to do with how many domains a god has.

The number of gods with each domain again varies pretty heavily.  Several domains are not represented at all, others (mostly alignment domains) many times.  Travel and Trickery both seem to show up quite a bit.

Again, I don’t see any particular balance in numbers here.

Monstrous Gods

There are 29 monstrous gods described (very briefly) here… I didn’t bother counting everything this time.

Visual inspection, however, suggests these are much more coherent than the other two groups.  I actually see common patterns to the domains granted (various pairings of domains evident).  Of the hundred or so domains assigned to these gods, only four or five appear to be non-core domains from outside the RSRD.  This appears to be just about right in line with what I get out of Polyhedral Pantheons.


All in all I think Polyhedral Pantheons will produce something workable.  The numbers of gods are in line I might have to run it two or three times to get a full set of gods.

It just occurred to me that I might want to do it more times than that, or pick three to five faces of the polyhedron (plus the points in or adjacent to them) for a limited pantheon.  If I take faces 1 through 3, plus points A, B, E, F, G, H, N, O I get 11 gods.  If I want to be even harsher, I could include only the domains from the selected faces and points I end up with gods with six domains (faces 2 and 3), five domains (face 1), four domains (point A), three domains (points B, E, and F, and two domains (points G, H, N, O).

I might still do that anyway — the orcs get this set of gods (quite limited), the elves get that set (not quite as limited, but still a subset), but the dwarves and humans learned to get along and thus share a bigger set.  There might be some shared domains around the edges, but there will be domains available to only one of the groups or another.  This has possibilities.

I think I will want to have more domains and subdomains to work with, though.  The Pathfinder Advanced Player’s Guide gives only 33 domains and about 60 subdomains (considering that some are shared) to work with, which isn’t quite enough to give all of the gods unique spell lists.  Unique spell lists aren’t important to me, quite, but I do like them to be varied.  Also, I like to have more choices to work with when building the base domain lists.  I think a combination of raiding third-party domain lists (Mercenaries from AEG collects a bunch of them from the AEG books, as I recall, and I know I can find other domains elsewhere) and reskinning spells can go a long way toward what I want here.

I distract myself.

Closing Comments

The numbers look like they’ll work.  The pantheon presented in Complete Divine for Greyhawk suggests that 12/20/32 are all reasonable numbers of gods for the start of a pantheon.  Examination of the domain assignments suggests that four to six is a good range of domains for a god.  Looking over the domain assignments I suspect that the really infrequent domains might work best as subdomains of more common ones.  For instance, Law and Domination are often associated with each other, so (without examining the spell lists or domain powers) it might be worth making Domination a subdomain of Law (ignoring also for the moment that Domination seems to be assigned quite a few times).  ‘Domination gods’ are still associated with Law, but their powers are not the same as all the other ‘Law gods’.

Good enough, plan looks sane.  Carrying on with the next part — fleshing out the fledgling pantheon.

No related content found.


  1. Nik

    One complication there: The Greyhawk gods are not one pantheon. Greyhawk deities are actually several pantheons, one per major ethnic/cultural group, and in many cases there are two or more gods with nearly-identical portfolios. In both Greyhawk and Forgotten Realms, which gods are “core” and which are “secondary” shifts with the individual writer, and demihuman gods are often treated like monstrous gods.

    • This is true enough, and it could indeed affect the comparison.

      However, thinking about it I’m not convinced it would invalidate it. If anything I think it might indicate that Polyhedral Pantheon might make for a conceptually stronger pantheon than the set of gods presented in Complete Divine.

      I might also make for an unusually homogeneous one, given that the domain sets are more coherent between gods (greater overlap in domains with some gods — each god shares no less than two domains with each of its neighbors, face to face and point to point). This might be a limitation or undesirable effect. It might in turn be countered by building ‘less complete’ pantheons as described in this post — orcs have no awareness of Goodness, or at least no gods that include that domain, but lots of Evil, Death, and Destruction; elves have goodness and light and inexpensive cookies.

      Thanks for mentioning it, it may affect how the god described are applied. I’m already seeing divisions and subsets of the gods here, where different subsets might be openly followed by different groups of mortals. I think this might encourage divisions, while still supporting a broad range of gods covering a large number of domains.

      More to consider in application.

Leave a Reply

Your email address will not be published. Required fields are marked *