This is a review of A Small Collection of Flowers and Entanglements by Luke Earl, and A Tangled Web by Benjamin McCown and Christian Stryffler. Both of the zines kickstarted on Zinequest 3. Both present novel systems for tracking NPCs, their relationships, and evolving schemes. I have read but not used the zines in play.
My interest in these zines arose from a felt need for two things. The first is what I call "technologies of memory", i.e. systems of note taking for both adventure prep and record keeping over the long haul. How does one "write up" NPCs and their relationships in a way that is geared to adventure? The second is an interest in systems that model advancing schemes of NPCs and factions, but in a way that is sensitive to player action (and inaction). I find that without a mechanic to model a world of evolving threats and opportunities that react to player action, my default is just to prep the next adventure location. I lose the sense of a dynamic world that is in motion and responding in active ways to what the players do.
So I’m in the market for these kinds of tools. Let’s see what’s on offer.
A Small Collection of Flowers and Entanglements
A Small Collection of Flowers and Entanglements is a 24 page PDF presenting two different systems, “Flowers” and “Entanglements”. “Flowers” is a system for randomizing evolving situations that are sensitive to player intervention. “Entanglements” are ways of representing networks of NPCs via their relationships and knowledge. The actual techniques are presented over 9 short pages. The PDF also presents attractive printable worksheets that allow you the DM to easily employ these techniques. The PDF is available on DTRPG here and itch.io here for $2.34 and £2 respectively.
Before I get into the techniques, let me get a couple of criticisms out of the way. The PDF has some typos and is written in a cheeky style I found a tad distracting. It could use some better or more fully fleshed out examples to illustrate the techniques. It also presents the techniques over 9 brief pages, and is mostly taken up with printable worksheets. However, the techniques are interesting and the worksheets are beautiful.
Flowers
Flowers is a variant of the hexflower technique developed by Goblin’s Henchmen. To tell you about the variant, I need to explain the original. The best introduction to this fascinating, supple, and highly customizable tool is The Hexflower Cookbook available here. Goblin’s Henchmen’s hexflower leverages the bell curve of 2d6 rolls to represent evolving situations where, although subject to chance, there is a direction events are most likely to go, and where what happens next depends on what’s going on now. You can use it to model almost any evolving situation including the weather, terrain traversed in overland travel, a trial by jury, or morale in combat.
The technique involves assembling hexes representing evolving conditions. For example, in the illustration above, the hexes each represent the day’s weather for purposes of overland travel. You begin at the center. From there each day you roll 2d6 to see where you move to next, consulting the hex key to the right to see which of six directions you move. Since 6-7 and 8-9 are the most likely results, probability will tend to move you down (mainly) and to the left (a little), although anything might happen. If you move off the map, you renter the hexflower at the opposite side of the same row, coming into the hex through the edge opposite the one you went off on.
Goblin’s Henchmen presents a couple techniques for making the procedure sensitive to player interventions. One technique involves giving the players action points they can use to turn the direction moved by one hex face per point spent. So, gaining a political ally in a revolution hexflower might give the revolutionary players 1 AP allowing the party to alter a roll by up to 1 face, from 8,9 say, to 10,11. Another technique, where the outcome towards which probability directs us (at the bottom) is something the players are trying to avoid, invovles flipping the key around when the players get a win, putting 6-7 at the top instead of the bottom of the hex key like this so that the probability trend is in the "good" direction, i.e. upward.
Suppose we use it to model the villain's schemes. At the very bottom is the villain’s scheme accomplished, surrounded by hexes that represent progress towards that goal. For example, perhaps the goal is “total domination of the city-state”. One result near to the bottom might be “villain infiltrates the government”. At the top is the ruination of the villain’s plan, surrounded by hexes representing setbacks, for example, “a key ally deserts the villain”.
The system uses the 5E advantage/disadvantage mechanic to model player intervention. If the players accomplish something that sets back the interest of the villain, then they roll with advantage, giving them a 75% of rolling a 4, 5, or 6 and moving in one of the good upwards directions. If the players ignore the scheme or suffer a defeat, then they roll 2d6 with disadvantage, giving a 75% of rolling 1, 2, or 3 and moving in downwards. Although Earl doesn’t do this, one could also imagine a further variant that includes a neutral condition for rolling without either advantage or disadvantage when neither side has the upper hand.
In sum, this is a neat mechanic for modeling evolving situations that are very sensitive to player action (and inaction). By eschewing alternate keys or meta-game action points it is cleaner than Goblin's Henchmen's original, although it loses some probabilistic nuances. And you can still get some of the same inertial energy produced in the original by the probability tilt of the bell curve by setting player inaction at rolling with “disadvantage”. Inaction flows towards defeat; if they snooze they (probably) lose.
Entanglements
Entanglements by contrast uses a tanglegraph to represent the relationship between a small number of NPCs, accompanied by a sheet with notes on what each NPC knows. It’s presented as a tool that is usable especially with small mysteries that have a fair number of tightly interrelated NPCs, where what each knows is crucial. The tanglegraphs are attractive visual representations of the relationships of NPCs. Indeed, as drawn by Earl they have the aesthetic of dynamic representation of the atom, suggesting relations between particles in motion bound together by forces. But there is also something occult, perhaps kabbalistic, about their trailing parabolas connecting circular nodes.
Where there are four or fewer NPCs, each NPC has an arrow going to each other NPC. Thus the relation between any two NPCs is modeled by two arrows one coming from each of the NPCs to the other. Each arrow is labelled with one or two words representing how the source NPC relates to the NPC at the end of the arrow. For NPCs of five or greater number Earl uses only a single arrow for each representation. The arrow might say, for example, “estranged brothers”, or “blackmail”, or whatever. This makes the visual representation easier to take in, and also accommodates the fact that it’s hard for players to keep track of so much information—so it pushes the DM to think a little more holistically about the relationship. The tanglegraphs are accompanied by a key that names the NPCs at each node and provides space to write what the NPC knows.
In sum, this PDF, although little, packs a punch. If you’re in the market for this kind of tool, the aesthetically pleasing worksheets are worth looking at.
A Tangled Web
The relationship maps, like the one above, are essentially tanglegraphs. But instead of Earl’s occult atomic energy, these illustrated maps are presented in a folksy way. They cater to people who prefer visual representations, with a dual key, using a different icon for each NPC, as well as for each type of relationship. By breaking free from the rigid format of Earl’s numinous templates, they allow for more flexible and simpler representations. For example, rather than choosing whether to use a single arrow or two one-way arrows to represent a relationship, McCown and Stryffler use both. If the relationship is mutual they put an arrow at each end. If the relationship is one way (i.e. unrequited love) they put the arrow at just one end. So, you could have two npcs with no relationship at all, two with one-way relations, or two with one mutual relation, or even a mutual relation and a one-way relation.
Personally, I find the dual key, although charming, a little unwieldy, since it requires multiple simultaneous decoding to read what is an otherwise quite simple relationship map. If I were making these for my home game, I would use names rather than icons to depict the NPCs.
The NPC key provides the name of the NPC, a 1-3 sentence capsule description, and one or two similarly terse pieces of information. Instead of focusing solely on knowledge as Earl does, these pieces of information are labelled Secrets, Flavor, Flaws, or Desires. This information is geared towards the very human mysteries that each of the six scenarios presents, where a combination of secrets, foibles, and desires create an unfolding mystery. Flavor presents welcomes notes on how to characterize or play a memorable NPC. The structure here is good and potentially relevant to different genres of play. You've got the relationship map, with a small key of different kinds of relations. You have a brief description of each NPC. Pick half a dozen tags that correspond to the features most relevant to the style of play for your game. Organize your notes that way.
The randomized tables that follow the relationship map allow one to roll to flesh out central elements of the small-scale mystery by providing further detail on motivations, schemes, or unfolding crises of two or three NPCs. I like the idea of further customizable options in a product presenting a series of small scenarios, but I have mixed feelings about the use of random tables to present them. I’m a fan of rolling on tables to represent unfolding events that are sensitive to chance (and player action as in Earl’s Flowers). I’m also a fan of rolling on tables in the moment to produce unexpected results during play, as with random encounters. During prep I’m a fan of rolling on tables for inspiration that get the imaginative juices flowing, as in Matt Finch’s glorious Tome of Adventure Design. But I’m less a fan of rolling to produce swappable customized elements of an otherwise fixed scenario during prep. Why would one do that? It doesn’t produce unexpected results in the moment, or get the creative juices flowing. So, what role does randomization play here? Would anything be lost if we presented them as a menu of options from which the DM may pick what seems most interesting? Of course, one can always treat a table as a menu of options if one likes, so the complaint is small.
The mystery scenarios presented range from quirky and low-key to sinister and high stakes. They are systemless, but written with the dressings of fantasy play, i.e. vaguely medieval village mysteries with demi-human characters, suggesting use for games like D&D. At one end of the spectrum, scenarios like “Critter Bones” and “Blackmailing the Blacksmith” read like fantasy versions of an episode of Northern Exposure. Take the first of these. A pair of town busybodies suspect a loner dwarf of being up to no good. As it turns out, he is struggling to learn necromancy, but he likely has harmless motives (e.g. reviving his pet or starting a zombie circus). Meanwhile, two waifs (a child urchin and a pickpocket) are about to get themselves in trouble with low-stakes criminal activity! Personally, I have trouble imagining spinning fun sessions out of this kind of scenario at least using a ruleset like early editions of D&D. They read like scenarios for Dogs in the Vineyard, but with D&D tropes substituting for the fire and brimstone.
At the other end of the spectrum are scenarios that seem like quite strong material for adventure. For example, in “Fish Tales” a bossy fishmonger has found an artifact in the gullet of a fish that will bring doom to the village (and perhaps the whole region) if the players don’t crack the mystery and intervene. My favorite scenario is “The Little Druid” in which a young druid is being taught shapeshifting by a talking wolf, while the pack (unbeknownst to the druid) attacks villagers. Meanwhile and unrelatedly, an assassin who can barely control their magical powers has slain the wrong victim with tragic consequences. There’s a lot going on in this scenario, and it has an air of fairy-tale like mystery to it that I can imagine shining at the table for old school D&D, especially for 1st or 2nd level characters.
The NPC key provides the name of the NPC, a 1-3 sentence capsule description, and one or two similarly terse pieces of information. Instead of focusing solely on knowledge as Earl does, these pieces of information are labelled Secrets, Flavor, Flaws, or Desires. This information is geared towards the very human mysteries that each of the six scenarios presents, where a combination of secrets, foibles, and desires create an unfolding mystery. Flavor presents welcomes notes on how to characterize or play a memorable NPC. The structure here is good and potentially relevant to different genres of play. You've got the relationship map, with a small key of different kinds of relations. You have a brief description of each NPC. Pick half a dozen tags that correspond to the features most relevant to the style of play for your game. Organize your notes that way.
The randomized tables that follow the relationship map allow one to roll to flesh out central elements of the small-scale mystery by providing further detail on motivations, schemes, or unfolding crises of two or three NPCs. I like the idea of further customizable options in a product presenting a series of small scenarios, but I have mixed feelings about the use of random tables to present them. I’m a fan of rolling on tables to represent unfolding events that are sensitive to chance (and player action as in Earl’s Flowers). I’m also a fan of rolling on tables in the moment to produce unexpected results during play, as with random encounters. During prep I’m a fan of rolling on tables for inspiration that get the imaginative juices flowing, as in Matt Finch’s glorious Tome of Adventure Design. But I’m less a fan of rolling to produce swappable customized elements of an otherwise fixed scenario during prep. Why would one do that? It doesn’t produce unexpected results in the moment, or get the creative juices flowing. So, what role does randomization play here? Would anything be lost if we presented them as a menu of options from which the DM may pick what seems most interesting? Of course, one can always treat a table as a menu of options if one likes, so the complaint is small.
The mystery scenarios presented range from quirky and low-key to sinister and high stakes. They are systemless, but written with the dressings of fantasy play, i.e. vaguely medieval village mysteries with demi-human characters, suggesting use for games like D&D. At one end of the spectrum, scenarios like “Critter Bones” and “Blackmailing the Blacksmith” read like fantasy versions of an episode of Northern Exposure. Take the first of these. A pair of town busybodies suspect a loner dwarf of being up to no good. As it turns out, he is struggling to learn necromancy, but he likely has harmless motives (e.g. reviving his pet or starting a zombie circus). Meanwhile, two waifs (a child urchin and a pickpocket) are about to get themselves in trouble with low-stakes criminal activity! Personally, I have trouble imagining spinning fun sessions out of this kind of scenario at least using a ruleset like early editions of D&D. They read like scenarios for Dogs in the Vineyard, but with D&D tropes substituting for the fire and brimstone.
At the other end of the spectrum are scenarios that seem like quite strong material for adventure. For example, in “Fish Tales” a bossy fishmonger has found an artifact in the gullet of a fish that will bring doom to the village (and perhaps the whole region) if the players don’t crack the mystery and intervene. My favorite scenario is “The Little Druid” in which a young druid is being taught shapeshifting by a talking wolf, while the pack (unbeknownst to the druid) attacks villagers. Meanwhile and unrelatedly, an assassin who can barely control their magical powers has slain the wrong victim with tragic consequences. There’s a lot going on in this scenario, and it has an air of fairy-tale like mystery to it that I can imagine shining at the table for old school D&D, especially for 1st or 2nd level characters.
The random tables included after the six scenarios at the end of the zine are extremely terse. I puzzled over how they were to be used in play. For example, there are three separate tables to roll on that each give a one-word description of the voice of an NPC with results like, "Indian accent", "exasperated", "haughty". Another table of weather has "clear full moon" and "sunny" as possible entries. What is the weather situation, such that it's an open question whether it is daytime or nighttime? The zine also ends with what I assume, perhaps, to be a die drop random event table. It includes things like, "A dog shows up out of nowhere. What could it want?" and also "The next roll in your game is a critical failure." I struggled to see what productive function this table might play.
Setting aside this brief foray into random tables at its end, and focusing on its original contributions, what A Tangled Web shows us is that it is one thing to have a way to represent and store information, for example, a tanglegraph of the relationships of networks of NPCs. This is an abstract tool, like Goblin's Henchmen's hexflowers that can be put to any number of purposes. What is at least as important is seeing how this tool can be put to a definite use--in this case, prepping small-scale mystery scenarios. Once one has seen six deployments of the tool for this purpose, one is well-equipped to carry on with one's own mysteries. I recommended a tangled web on these grounds, especially if small-scale mysteries with a big heart is your jam.
Where to Go Next
In my mind, the question is where to go next with devices of representation like this. In particular, how could one tailor them to questions of larger scale and greater complexity than homespun small town mysteries built from broken hearts, land speculation, and brassy fishmongers? What forms of representation would be required for adventures focusing on political intrigues among rival factions in an urban environment? How can we depict the jostling fractious politics of wizards in shifting consortiums, with relations to the strange entities of the Overworld or fey courts? Part of the key, surely, will be settle on a set of tags to organize their notable features different than either of these zines introduce for organizing the information about NPCs relevant for small-scale mysteries. For example, for the wizards we might ask what domain their occult researches are in. For general factions, we might represent what resources they possess. And so on.
I think it would also be interesting to experiment with nested tanglegraphs, with say each major faction in a locale (a city, say) getting its own tanglegraph, and separate tanglegraph representing the relationship of factions to one another. It would be fun to somehow figure out how to keep a big whiteboard that had the nested tanglegraphs available electronically in a single representation. This would even allow you to draw lines connecting elements of one tanglegraph to elements of another, where one player in a faction had a secret alliance with a player in another faction, although this would quickly come to look like those giant boards in movies that disgraced police detectives erect in their bedrooms to track some hopelessly complicated case they were never able to crack.
