Materialist Campaign Creation

Jon Peterson’s book The Elusive Shift captures some of the discussions around campaign creation that took place in the early days of the hobby. This serves as interesting context to Gary Gygax’s influential 1975 article on the topic. Something that stood out to me is its account of Tim Waddell’s division of campaign creation into 4 levels, also written in 1975.

Achieving the first 3 levels of Waddell’s levels is surprisingly simple. In fact, the quick start Gygax ’75 method blasts through all of them. Level 1 is just a dungeon, level 2 consists of a few dungeons and wilderness hexes; level 3 is having at least 1 town on top of this. Level 4 though is certainly a steep jump, consisting of:

“several completely mapped towns, plenty of interesting townspeople, rumours, legends, history, etc. A total fantasy world”

In 1976, Dan Pierson in Alarms and Excursions #14 was seeking something similar (p133):

“A campaign: thousands or millions of square miles of mountain, forest, plains, and ocean; the conflict of empires, or slow attempts to put small states back together after a great war … your characters struggling in a vast setting for power, wealth, good, evil or whatever.”

Waddell describes himself as having only achieved 2/3rds of level 4, showing it to be a vastly more challenging undertaking than the first 3 levels. He acknowledged that a level 4 world would require “hours upon hours of work by a ref with a reasonably fertile imagination”.

My previous post looked at features of the Gygax ’75 method that minimise the workload of campaign creation. This is partly through a free mix of gonzo elements, with only weakly conceived “external” connections between these. In my own campaign, I’d like some of the depth that comes with a developed “4th level” world. But I also want to have my cake and eat it, avoiding the accompanying explosion in workload.

It's nearly 50 years since these issues were raised, so perhaps the task has been “solved” already. What I’m going look at specifically is how a materialist perspective can contribute to efficient campaign creation. I’m posting this as part April’s blog carnival - hosted by Codex Anathema on the topic of On creating elements for a setting.

Setting as a Starting Point

The first step of the Gygax’s 1975 method concerns the setting, rather than jumping straight to the design of the starting dungeon. While Gary opted for a generic, abstract setting, he gives examples of more specific settings can be used such as Norse mythology or the Hyborean Age in Conan. What he doesn’t give is guidance on which elements of these settings need to be accounted for right off the bat.

Waddell’s article as summarised in Chapter 4 of The Elusive Shift (as I’ve not been able to track down the original) mentions the following setting elements he wanted laid out:

• A large map (“towns, rivers, mountains, castles, evil places”)
• A record of who lives in every residence within a town
• A biographical sketch of every individual
• Proprietors and inventory of every store
• Laws, religions
• Important citizens, with their secrets and dispositions
• Legendary people and places

This is quite the list. Obviously, we’d no longer be looking at a quick start method, were it all included in the first stage. Gygax ’75 method is kick most of this into the long grass, adding further elements organically as the campaign progresses. This becomes increasingly difficult though as a setting becomes more specific and less abstract gonzo.

In their 1976 article, Dan Pierson bemoaned there being effectively an “inter-dungeon portal system” in D&D. He meant this not literally, but that there was an absence of meaningful experiences whilst traveling between adventure locations (p133). Pierson wanted to learn about what the rest of the world is doing in these periods, which is hard to do if this world hasn’t even been sketched out yet.

Even Gygax’s quick start campaign has some travel between locations and time spent in town. To a degree, we need to know how the setting characterises these experiences soon after the dice start rolling. A Norse town is obviously going to be significantly different to an Anglo-Saxon settlement.

A further issue is that some campaign elements we’ve kicked into the long grass are difficult to later helicopter in. The piecemeal expansion of a setting map might mean that it eventually incorporates the capital of an empire. But it then become mysterious why the empire’s influence was not felt at the original town and dungeon, when the players were adventuring there at the start of the campaign.

Eliminate the External

In a recent post, I looked at internal and external relationships in adventure design. The external is the arbitrary, relationships that could as easily be otherwise without altering the identity of the subject they apply to. If players ask the shoe colour of a random tavern patron, this can easily be made up on the fly or generated on a table. But being the captain of the town watch would likely be a major part of their identity, affecting how the adventure unfolds.

Generally, I think the external elements of a setting are the chief candidates to put off thinking about when starting a campaign. These are equally the elements most readily generated through random tables or “outsourced” by taking them from published campaign material.

It will not always be clear cut which elements of a campaign count as internal and external. This will partly depend on what you want the emphasis and purpose of the campaign to be. But looking at Waddell’s list of desiderata, some do seem to belong in the external camp more than others.

Far off sections of the setting map are an obvious candidate to treat as external, which is exactly what Gary does in his ’75 method. Whether there is a mountain range or swamp a week’s ride away is only going to marginally relevant to what happens in a starting adventure. Similarly, the inventory of the local store could be determined randomly on the fly. We might want to limit the total number of magic swords in a hex area, but the proportion of these in a particular store doesn’t need to be decided far in advance.

I would say that “legendary people and places” belong in the external category too. Whist the immediate history of a setting might directly impact its current events, for something to have “legendary” status places it further back in the mists of time. The lich of an ancient wizard might return to have a major impact on a campaign, but until this happens it can be easily filed with forgotten history. Not having timelines set in stone gives leeway for forgotten episodes to be rediscovered as adventurers delve the depths.

Historical Materialism

What then are the main internal relations in a setting, to consider before even drawing the first dungeon level? This is where I think historical materialism comes in handy. It provides a narrative of history that can be applied to a setting. This can serve as a rule of thumb to “grok” how the elements of a setting relate to each other, without plotting these out in advance.

Historical materialism views there being a dominant social process that drives change in a given period of history. There is a hint of this in Pierson’s desire for campaigns to show “the conflict of empires”. Imperialism, as the drive for empires to constantly expand, can be understood as part of a process. This is something that would likely influence a starting adventure location, and is hard to just helicopter in later.

For me, identifying the main process of change in a setting is one of the first steps of campaign creation. As the campaign unfolds, I will then think about how every character, institution and location introduced relates to this process. This probably sounds like a lot of work, but I find that it in fact saves you the far greater task of mapping how each atom of a setting relates to each other.

If the process is “the expansion of empire”, those who directly benefit from said expansion will often be a small proportion of society amongst the ruling class. There will also be agents providing valuable services to the ruling class, who indirectly benefit from this process (or aspire to). On the other hand, there are those who are subjugated as part of the empire’s expansion. There are also those within the empire who are exploited in the process of carrying out its expansion.

When introducing a new element in such a setting, I would think about whether it falls strongly or weakly into one of the above categories. This then gives me a good sense of their disposition towards other characters, in respect to the camp they fall into. There is similarity here with good ol’ D&D alignment, but with far more nuance, as the relationship is not always oppositional. Dispositions may also evolve. An exploited soldier might start off in opposition to those being subjugated, but later come to identify with them.

Something I’m trying to avoid are those charts of faction relationships you often get in published RPG settings. Obviously, this saves you the work of mapping such things out yourself, but I don’t think they are quite the solution they appear to be. While GMs often benefit from handy charts to track these relations, the players don’t have this reference. It can be a struggle to recall a network of idiosyncratic relationships between sessions. Far better I think is a simple basis from which to understand the forces that shape relationships, which players begin to intuit as they explore the setting.

Summary

I’ve been thinking here about a minimalist approach to incorporating Waddell’s 4th level into campaign creation. Elements I am calling external might still require some forethought, but don’t need to be locked in from day one. But starting with a simple central process can greatly help to orientate the “internal” relationships in a setting. I’ve looked at how this can work with physical processes in a prior post (“Winter is coming” in Game of Thrones is another example of this). But I think identifying the dominant social process has more general application and far-reaching connection to the goings on in a setting. I hope to expand on this with examples in future posts.