Monday, 22 August 2011

The Relationship Hexmap

Following on from yesterday's post, and drawing from ckutalik's interesting sequence on Top Secret and network linkages from earlier this year, I did some thinking last night on how to come up with a conceptual hexmap for non-geographical sandbox play. This post is just a summary of my preliminary thoughts - it's basically some rough ideas.

What I want, and what I envisage, is a hexmap revolving around interpersonal space rather than physical space. Instead of the players exploring a geographical area hex by hex, with encounters and events taking place as they go, in the Relationship Hexmap they explore a network of interpersonal relationships instead.

The key to this, as I see it, is that it has to retain the elements of what makes a hexmap work when you're using it for a D&D-esque sandbox. It has to be loose, it has to allow for random events, and (crucially) the GM shouldn't have any preconceptions about what will happen.

This last point is probably the most important. What I like about a fantasy sandbox is that, although the DM has set up the basic framework and scattered the thing with interesting locales, it's as much a voyage of discovery for him as it is for the players. This means that, although there needs to be some pre-existing information on the contents of each hex, it should be as vague and nebulous as possible until the players discover it.

Maybe it's best to give a physical example of what I'm talking about. Here is a small relationship hexmap that I came up with this morning when I should have been working. (Something about drawing it using paint's "crayon" feature appeals to the inner child in me.)

First off, an explanation of what all the pretty colours mean:

  • Blue blobs are actual sentient entities that have agency in the game world. Each could represent an individual, a gang, a corporation, a cartel...really anything that could forge communicative or personal links with another blue blob. For the sake of this hexmap, I'll call them "people", but remember this could refer to one person or many people in one unit.
  • Red lines represent a strong, positive link between two people - basically, where an alliance, agreement, or mutually benefecial relationship exists. It could be love, it could be a military alliance, it could be a free trade agreement - anything you can think of.
  • Orange lines represent contact. Simply, two people linked by an orange line have had some sort of communication with one another, whether regular or not. It could be they've just had one interaction in the distant past, or it could be that they're in regular email contact but the relationship is not positive enough or close enough to be "red".
  • Yellow lines simply represent awareness. The two people are aware that the counterparty exists.
  • Arrows are used to indicate whether contact, or awareness, is one way. Yellow and orange lines usually work in both directions, but in some relationships one party is always contacted by the other and not the other way around (a spy who recieves communiques via encrypted SMS messages, for example), and sometimes a person can know about another person without the same being true in reverse. Arrows indicate this.
  • Black lines represent hostility. They are the opposite of red lines. Here, an antagonistic relationship (open or covert) exists.

So how does all this work in actual play? The answer is that I'm not sure yet, but my basic thoughts on how it might break down are as follows:

  • First, the GM draws up the hexmap in a random(ish) way, as I did this morning. I haven't decided on who any of the blue blobs represent yet, or why their relationships are the way they are. I just had fun with crayons.
  • Second, the GM picks a number of the blue blobs and gives them a personality, history, and description. It doesn't have to be all of them, and in fact, it shouldn't be. Around half will do. This is the equivalent to the way in which a D&D DM draws up a hexmap but only details what's in some of the hexes - the rest will come as and when it is necessary.
  • Third, a handful of the blue blobs/people are selected as the locus points at which the PCs could "start" the campaign. As with a fantasy sandbox, there needs to be a beginning point, but it would be too restrictive to select one blue blob/person and say "You know this guy and he tells you [x]" (although the Mr. Johnson routine is fine if you're into that sort of thing). I imagine that most Cyberpunk sandbox campaigns begin with the fixer-types using their streetdeal skills to find contacts and rustle up jobs. This could give them a number of different "leads" (each being a separate blue blob). 
  • Fourth, the immediate relationships (orange, red, yellow, black) around those locus points are fleshed out in a bit more detail. 
  • Fifth, the GM puts in some "other interesting information" relating to each hex. I imagine this introducing some sort of temporal element to the process. For instance, you might write that on Day 5 of the campaign, Person 0402 will attempt to assassinate the head of Company 0503. Or, on Day 8 of the campaign, Person 0307 will be contacted by Person 0402 in relation to Gang 0607.
  • Sixth (and this is hazy) there need to be random tables changing the nature of the relationships on the map. At various points in time, the GM needs to be rolling dice to find out what is going on on the map irrespective of what the players are up to. I'm picturing a table of stuff like "relationship deteriorates because of X" or "communication breaks down because of Y".
  • Seventh, profit.
I imagine it working especially well for investigative play - for instance, if the player characters are police, engaged in corporate espionage, that sort of thing. This is fitting, since I cribbed a lot of the ideas from ckutalik, who in turn cribbed ideas from Top Secret.

Anyway, more on this to follow as I ruminate some more.


  1. You might find this of interest. It deals with very similar topics:

  2. In some ways I like this even better than the Top Secret style nodes, better visual organization--I also like how you can represent relationships stance (hostility or affinity).

    Very nice.

  3. I also like the phrase "non-geographical sandbox" I had a hell of a time trying to think up a phrase to cover the concept.

  4. The immediate things I see here are:

    -It's possible you could just leave them and not change the relationships at intervals because PC actions would change the relationships for you.

    -Geographical relationships (i.e. how far town a is from town b and in what direction) don't usually change, no matter what the PCs do, whereas these presumably do (and are more interesting if they do). This means the conceptuaization of a normal hexmap and the visualization of this interpersonal space might need some essential underlying difference in order to work properly or in the most interesting way.

  5. I'm really interested. I keep wanting to push the landscape metaphor. So, getting more familiar/tight with an entity means getting closer on the map and there are certain features on the map that are triggered when players become that close. Have a speaking relationship with two folks from cartel A, say: you're in some rough territory in regards to how cartel B feels about you. Sort of like a flowchart, but not binary.

  6. Interesting concept, and it does seem like a logical development of the Top Secret node map.

    Thoughts: How does the relationships relate to a map? Do the hexes represent distance in a relationship, with the closer hexes being more personally related, either as enemies or friends? If you go that path, then Stars Without Number has a stellar map generator that could be helpful to randomly create distances.

    To build from Telecanter's concept, could you have different map scales, where the "large hex" is the organization/faction, major NPCs etc, and then "zooming in" shows the web of more personal relationships, minor NPCs etc.

    Also, why not represent the PCs on the map as well? Their starting hex is then connected with the others by the original contact lines provided by their mentor/sponsor etc.

  7. -C: Did you ever do the follow-up posts as promised in the comments?

    ckutalik: Mostly this sprang from playing around with the new paint program on Windows 7. The crayon tool is my favourite thing ever.

    Zak: I agree. As it stands it's sort of a useful way of mapping NPC relationships, but not particularly interesting. It needs something else. I'll come up with something.

    Telecanter: Yeah, see above, basically. I'm not sure it's possible to map that in 2D space, though? Ideally as well as a horizontal axis there would be a vertical one too.

    lasgunpacker: Hexes don't indicate distance, at least as it stands currently. Again, I'm not sure how you could really map closeness of relationship on a 2D space along with everything else you need to map.

    I do like the idea of "nested" mapping though, to represent factions and then factions within factions and then individuals within those.

  8. the conceptuaization of a normal hexmap and the visualization of this interpersonal space might need some essential underlying difference in order to work properly or in the most interesting way.

    I ran a session of my White Sandbox campaign using the artifact flowchart from Gamma World. While it did provide some structure to the process of conducting guerilla raids within giant-controlled territory, the thing that made it less than satisfactory was the way I conceptualized the territory. In Gamma World, a single dice outcome can move you closer to disaster. Because the planning for these guerilla raids was done with a bird's-eye view of the region, we all tended to think of the threats in the region spatially - enemies are known to be over here, the invisible elf scouted this farmhouse and thinks it's safe.

    This conflicted with the way I used the idea of being at risk to guide the way I called for rolls to move along the flowchart nodes, because a bad roll could mean danger becomes manifest and I had to distort my spatial thinking to bring it onstage.

    The beauty of a relationship map is that it makes it clear when you roll to move along a node: only when that relationship is affected. And we can intuit a lot about what affects a relationship (e.g. leaving dead giant bodies to be found in someone's demesne), so often it's really easy to visualize what happens. Just mapping the relationships is worthwhile, the chart doesn't have to do as much of the work of resolution as the Gamma World artifact chart does.

    The things I think would make this totally & immediately useful:
    - a random table for generating who is represented by each note, maybe specific to some color-coded faction: red fire giants are linked to dragons on a 1-4, but slaves only on a 1.

    - a system for rolling the hit points of the relationship, maybe with a value for friendly (it's really easy to bribe these giants) and unfriendly (they are really hard to beat in a fight).

    - Tavis

  9. Tavis: We'll hash this out by email, but yeah, random generators are sorely needed to make this sort of thing work.

  10. As I look at this idea I'm reminded of my old programming flow charts, which I think I've used for plot lines more than I ever used to program with. These allowed me to display possible paths to the "big boss." Then again it also reminds of the complicated "you're sleeping with everyone you partner has ever slept with charts" that were popular a few years ago.

    Have you considered that this might be better modeled with a molecular model set?

    The classic law enforcement model used to imagine a monolithic underworld, a king pin, a Don, yet the reality of many crime cases reveals that it's a much messier set of relationships. That is you can move laterally from Whitey Bulger to a crooked FBI cop.

    How about...

    Adam is a thief. Adam knows Bob who deals in guns, Barry who deals in Data, and Bishop who sells drugs.

    Bob, as a gun dealer knows a number of Adams (Amos, Arnie, Allison) who all specialize in different types of thievery. He also knows Camerons (Hitmen), Davids (weapon makers), Eds (smugglers), and Franks(Enforcers).

    Bishop as a drug dealer would know Franks, Eds, as well as Georges (club owners), Ivans (celebrities), etc..

    Yet each new NPC might circle back to an old contact or a new one. So you calculate a probability of "inbreeding" or outbreeding, maybe base this upon some sort of modified "alignment" of law to chaos, representing the NPC's personal mindset.

    Start be defining the possible NPCs, and what are their probable connections. You might run into a cartel of smugglers, each with his own contacts. So when the party infiltrates/contacts/kills one they might move through the cartel/ring till they find the one they need or want to follow up on.

    Geography would come into play, perhaps as a second factor of each NPC. A smuggler is likely to have a probability of a higher number of contacts outside the locality, a thief a low probability. If it's a randomly generated number, then the thief with an 18, would be the international jewel thief, rather than the back alley thug.

    So let's see that is
    Law vs Chaos
    Introvert vs. Extrovert (local vs non local)
    Networking (covers the number of contacts he has, a random number coupled with profession)

    Then like most sandboxes, you use a smaller number of actual scenarios to give the illusion of the giant sandbox.