I've written at length about the social aspects of the cape-and-sword sandbox, that the swashbucklers' game-world is less focused on setting locations than it is relationships between player and non-player charactes. However, a cape-and-sword campaign still takes place in a place - in fact, many swashbuckling campaigns are set in a world both alien and familiar. That world is our own, or more specicially, the world of our own past.
Now here's the thing: the real-world is big friggin' place. It's not a lonely penal colony in a backwater star system. It's not a keep on the borderlands, surrounded by a howling, monster-haunted wilderness. It's a world filled with people: their numbers are great, their settlements are many, their impact is profound, and their presence is all but inescapable.
So, the first thing a referee running a game in the real-world must consider is, how much of this do I really need to describe, anyway? The easy answer is, only those areas which are of particular relevance to the adventurers. One could, for example, play a campaign of roguish highwaymen and detail little beyond a forest and its roads and tracks, a few villages, a nearby town, and perhaps some ruins or a cavern as a hideout, with the assumption that the adventurers will remain within the edges of the sandbox by dint of social contract, campaign premise, or whatever.
But what happens when the edges of the sandbox are less distinct? How does the referee handle a setting which attempts to encompass a whole world, particularly our own?
In setting up Le Ballet de l'Acier, my Flashing Blades campaign, I wrestled a bit with this. The 'core setting' for the campaign is 1625 France, a country encompassing tens of millions of hectares of land, millions of people, thousands of kilometers of navigable rivers, and on and on. My approach to running any sandbox setting is to 'prep to improvise,' to assemble in advance the tools I need during actual play to make stuff up in response to what the adventurers do, so that if the adventurers, say, set off for Nancy from Paris, I can maintain the verisimilitude of the setting and offer credible answers to reasonable questions which may arise - how does, say, a farm in Champagne differ from a farm in Provence? what wares might a merchant in Limousin be transporting compared to a merchant in Lyonnais? are common names in Brittany different from those in Gascony?
As you might guess from those examples, part of my preparation revolved around a basic unit of political division in ancien régime France, the province. Provinces feature prominently in the career rules for Flashing Blades, so it was important for me to know something about them. I gathered information on physical geography, population demographics, and economics, along with a bit of history. Focusing on the provinces allowed me to break down the whole of France in manageable chunks.
Even so, my treatment of the provinces in necessarily broad-brush. Looking at a period map, it's easy to see the geographic complexity of the countryside, with hundreds of villages and anything from a half-dozen to a score of sizable towns and cities. Even detailing the provinces left me with a very broad-brush treatment of France.
More importantly, cape-and-sword campaigns are less about tromping across fields and forests than they are engaging in intrigues, so I needed to pay close attention to the geographical milieu in which these intrigues take place, and that meant cities. Early on a made a list of five cities, which I then set out to develop in some depth. Paris was the first and most obvious choice - really, it would be pretty easy to run an entire swashbuckling campaign set in Paris if I was so inclined. The largest, most politically important city in France - still the seat of royal power in the time of my campaign - I spent the more time acquiring a feel for Paris than I did any other place in the campaign, learning about its palaces and gardens, churches and priories, markets and alleys, bridges and quays, colleges and courts, prisons and cemeteries. The core rules for Flashing Blades and the short-adventure collection Parisian Adventure both describe numerous landmarks around the city, as does GRUPS Swashbucklers and At Rapier's Point, the Rolemaster cape-and-sword supplement - beyond those I dug through histories and guidebooks to the city, a number of them written as far back as the eighteenth century, to develop my feel for the Early Modern life of the future City of Light. I've found that approaching a city like a tourist tends to clue me in on what is likely to be important to the adventurers in the course of the game, and to prepare accordingly, a trick I learned while running Traveller, where adventurers may travel to three different worlds in single game-day.
After Paris, I next considered another important aspect for swashbuckling games - what if the adventurers chose to turn pirate? That mean ports, so the next three cities on my list were Rouen, La Rochelle, and Marseille. Rouen in this period is the gateway to New France and, along with Saint-Malo, one of the most important trading centers with the rest of Europe and points beyond. La Rochelle is also a critical port city in this time, with extensive trade with the Protestant countries of northern Europe - England, the Provinces, Denmark, Sweden, and the Empire; should some or all of the adventurers be Huguenots, moreover, knowing details on the most important Protestant stronghold in France was essential. Marseille is the gateway to the Mediterranean, and particularly important give the southward tilt of my sandbox.
The last of the five cities on my list was Toulouse, the 'Paris of the south,' an important administrative center in Occitan-speaking France, a city with deep history and rich culture - if forced to choose just one city in which to run a cape-and-sword campaign, I'd pick Toulouse over Paris in a heartbeat.
So, these - Paris, Toulouse, Rouen, La Rochelle, and Marseille - became the campaign's 'points of light,' the geographic 'shiny objects' with which a sandbox setting is studded. For each city there are various webs of intrigue against which the adventurers may brush, through rumors and random encounters; a number of them reach across France to one another as well, such as the Toulousain vicomte de Bouvard, a royal courtier whom the adventurers met on a visit to the Paris horse market. Over the course of play, two more cities, Grenoble and Torino, in the duchy of Savoy, have received more detailed treatment as a direct result of the adventurers' actions, and long-term I'm considering fleshing out Bordeaux and Amiens as well, in anticipation of future events in the campaign, specifically France's entry into the Thirty Years War.
It would be very convenient for me to limit my sandbox to a city, or even a province, but it would also be limiting in ways that I wouldn't really enjoy, and it would deny the players the freedom I want to them to have as they approach the campaign. "Small, isolated points of light scattered across a big, dark, dangerous world" may make for an excellent setting for some games, but for my cape-and-sword campaign, the game-world is a sea of light with a few that burn a bit brighter than the rest.