Monday, 21 November 2016

Conceptual density (or 'What are RPG books *for*, anyway?')

I recently picked up Shadows Over Vathak for free in an online sale (summary: a perfectly good two-page setting document mercilessly stretched out into an interminably lengthy setting sourcebook), and it got me thinking about, well, ideas...

Imagine this situation: you're running a game of D&D, and your players arrive in a new place - a tavern, a market, whatever - and you happen to mention that one of the people hanging around the place is a dwarf. This dwarf has no special significance: you just came up with him on the spur of the moment, as the kind of person who might be in a place like this in D&D-land. But one of the PCs latches onto him. She asks questions about him. She goes over and talks to him. She wants to know what a dwarf like him is doing in a place like this.

At this point, you might have any number of brilliant ideas about how to work this dwarf into your campaign. But even if you don't, there are three options which are always available to you.

Option one is the path of least resistance. 'His name is, um, Dain Steelbeard. He's a blacksmith. He's short and muscular and he has a massive beard. He's in town to buy, um, mining equipment for his clan, who operate an iron mine up in the hills.' Anyone with even the most basic knowledge of traditional fantasy can come up with this stuff ad nauseam. It's boring and cliched, but it'll do until something more interesting comes along.

Image result for dwarf
If he isn't called Dain Steelbeard, I bet someone else in his clan is...

Option two is the path of free-association. 'His name is, um, Ronald Cakeman. He's a cat trainer who travels the land in search of designer shoes. His beard is dyed bright green and he's wearing a bowler hat. He's embroiled in a sordid bisexual love triangle with a centaur and a harpy and his sorrows have driven him to drink. And he rides around town on a yak.' Coming up with stuff like this is easy because it is meaningless random bullshit. It's more fun and colourful than yet another beer-loving, axe-swinging dwarf with a bad Scottish accent, but it's also more disruptive and less likely to cohere with the rest of the setting. Too much reliance on this sort of free association will quickly turn your game into surrealist nonsense.

Option three is the path of filler. 'His name is Dain the, um, Corpulent. He's really fat. He has bright green eyes and he works as a carpenter, repairing the local mill. And he hates mice.' This is the kind of stuff you can get off any one of a thousand 'random NPC traits' tables. Random detail about occupation. Random detail about appearance. Random detail about personality. Just enough to distinguish him from the next dwarf, but nothing that's actually particularly meaningful or interesting. After all, why should your players care that he has bright green eyes?

Now, there are all kinds of other, better things you might do with him as well, but these three options are always available as a kind of baseline minimum. You can always fall back on cliches or invent random nonsense or string together plausible but irrelevant details until they accumulate into something resembling a (rather boring) person. They are the most basic implements in the GM's toolbox.

I think it thus follows that, if an RPG supplement is worth reading (let alone buying), it has to give you more than that. The contents need to be something better than you could come up with, unaided, simply by following cliches and/or random madlibbing and/or coming up with irrelevant filler. Otherwise, how have they improved your game?

I say this because quite a lot of them don't. Like, they'll detail an orc tribe, in enormous detail... and it will all be exactly what you would expect from the words 'orc tribe'. They'll be fierce and warlike and brutal and love violence and live in a cave and have a thuggish chief who likes throwing his weight around and a creepy shaman who's always demanding more sacrifices and you could have thought of all that yourself. It boils down to 'this orc tribe is a tribe of orcs who look like orcs and think like orcs and act like orcs and fight like orcs'. Sometimes entire campaign settings will be like this, spending hundreds and hundreds of pages telling you that, yes, the not-Vikings sail around in longships and raid coastal settlements and the not-Egyptians build giant monuments to their dead Pharaohs and all the other stuff that you could have guessed just from being given the most basic overview of the world. There's no added value.

Image result for orcs
Orcs who are angry violent savages? And fight with axes? My mind! It is blown!

Or else they'll take the filler route, and just fill their word-count with irrelevant detail. The headman of this village is tall and old and cheerful. The headman of that village has a long black beard and drinks too much. Maybe this isn't exactly what you would have come up with if you were inventing these NPCs out of whole cloth on the spur of the moment... but the chances are that if you had just made something up on the spot, then whatever you did come up with, possibly with the help of a random table or two, would have been just as good. What's the point of writing things down in full detail when a single 'random NPC quirks' table could have done the same job a hundred times over?

You've hopefully all read Zak's 'One More Idea' method for adventure writing, where you start off with something totally basic ('a room with three weak monsters, then a room with one big monster') and just keep going back over it, adding one more idea to each concept each time, until it's developed into something interesting enough to run: his example shows how adding just three ideas to each adventure element is enough to get you from a bare skeleton to a compellingly weird, densely interconnected scenario. But that only works if each idea is actually adding something worth having. If your first concept is 'dwarf'' and your next five ideas are 'loves gold', 'hates orcs', 'massive beard', 'expert blacksmith', and 'fights with an axe', then frankly you might as well have stopped with idea number one.

So this is where RPG books come in. They give you better material to work with. They are full of examples where the author added one more idea, and that idea was really good. Slumbering Ursine Dunes, for example, gave me neanderthal Cave Dwarves with flint hammers and stone-and-metal Robo-Dwarves with robotronic eyes and evil space elves who talk and dress like Ming the Merciless, and since reading it I have had that many more options available to me when thinking about what the random elves or dwarves hanging around my games might be like - options which aren't just the same old cliches, and which are likely to be much more powerfully imagined and conceptually coherent than whatever randomness I might ad-lib on the spur of the moment. Most RPG supplements, even pretty dull ones, will have at least one or two such ideas buried somewhere within them. But the really good ones will have stuff like this on almost every page.

This is what I think of as 'conceptual density', and it is, bar nothing, the single thing that I look for most in RPG materials. A weak book will take one moderately interesting idea - 'What if a bunch of kobolds started worshipping Cthulhu?' - and eke it out over ten, or twenty, or thirty pages, filling in all the gaps with either cliche material (kobolds doing either generic kobold stuff, generic Cthulhu cultist stuff, or a mixture of the two) or with the kind of random details which are no better or worse than the stuff you could have come up with on the spot. A good book will use that one idea as a kind of basket into which it proceeds to stuff as many other new ideas as possible. Think of the sheer number of ideas packed into, say, Qelong: war-torn fantasy Cambodia and animated stupas and a river which is a Naga and it's stirring in its sleep and its dreams are warping people into monsters and a magical device is leaking arcane radiation into the water and it collects in people's hands and the villagers cut people's hands off to be safe and those hands then animate and crawl around the countryside strangling people and and and and... 

Image result for qelong
and and and it's all going to fucking kill you, but at least getting eaten by a miles-long snake-river-goddess with four heads will be original...

I guess what I'm getting at is that, quite a lot of the time, I feel that a lot of the content in RPG supplements - especially adventure modules - is superfluous. 'Trolls raid nearby settlements, capturing villagers to sacrifice in an ancient stone circle in the hope of awakening their ancient monster-god' is a decent enough basis for an adventure. But if everything that follows is just pretty much exactly what your reader would have expected based on that description - trolls acting exactly like trolls normally do, villagers being cliched fantasy villagers, the monster-god being a gribbly tentacle beast straight out of Cthulhu Central Casting, and so on - then you might as well have left it as a one-line adventure prompt rather than writing a whole module about it. The extra page-count should be justified by its content: it should be full of stuff that is, on average, better than most people are going to be able to come up with on the spot. 

My favourite RPG supplements are the ones in which the page count corresponds to the number of actual ideas, and thus maintain a very high level of conceptual density. One or two ideas in one or two pages is fine. Dozens of ideas in dozens of pages is fine. But a handful of ideas stretched out over dozens or hundreds of pages, Shadows Over Vathak style, is just wasting everyone's time.

Probably coming next: behemoth Pathfinder mega-campaign Rise of the Runelords distilled down into a single blog post!


  1. This post encapsulates my thoughts so perfectly and succintly, it now makes me want to read Qelong again.

  2. Great post. I feel it's a summary of something I was trying to articulate, but wasn't able to do it so clearly and so precisely. And also I think that, at least for me, it's going to be a nice framework to evaluate the work of others of my own. Very nice indeed, sir.

  3. Take your concept and brainstorm ten ideas around it. Then take that sheet and throw it out. Think of ten more ideas. If those are as cliche as the first ten then throw those out too. Now the work gets harder to come up with ideas but the ideas get better.

  4. I agree with your point regarding published products but I think you might be underestimating the opportunity provided by players' capacity to latch onto inconsequential details when you say "why should your players care that he has bright green eyes?" I once played in a game where a party spent a good 10 minutes of game time "examining" a turnip in the middle of a road. When I had my character pick up & take the turnip, the reaction was along the line of "Oh my God! What has he done?" We continued to encounter turnips every thirty feet or so along the road & the rest of the players reacted the same way every time we encountered one & I picked it up. When we finally caught up with the peasant with a cart full of turnips on his way to market, I said, "Hey, you dropped these" and, as gratitude, he gave us a ride to town.

    1. Oh, the ability of PCs to obsess over random details is legendary. But usually it's the result of assuming that the importance of everything is equal to the prominence it's given in the GM's narration - so the turnip *must* be meaningful, or else why would the GM have mentioned it?

      So if you're not in the habit of describing the physical appearance of NPCs, and you then mention that a dwarf has bright green eyes, then your players are quite likely to think that must mean something. ('Maybe he's a green dragon in disguise!') After all, if it didn't, you wouldn't have mentioned it, right? But if you're in the habit of giving a little bit of description to most of the NPCs they come across then the chances of this sort of behaviour will be reduced, although never entirely eliminated. You know what PCs are like.

  5. Worth reading of only for the captions on the images!

    1. I'm glad to see that my fine caption-work is finally getting the recognition it deserves...

  6. Well written, and some great comments too. I have used your method but I don't think I ever conceptualized it like this before. Thanks!

  7. How would you (or would you) differentiate between "good" ideas and "novel" ones?

    1. Well, it's good if it leads to better play. Novel ideas will *tend* to be better than cliches, because they make everyone sit up and take notice, rather than autopiloting their way through yet another cave system full of goblins with an ogre at the end. But novelty isn't a value in and of itself: that's the point I was trying to make with the Path of Free-Association. Ronald Cakeman the Dwarf is novel, but he's not likely to lead to very good play.

      What makes an idea good? Novelty, yes, but also clarity and coherence and what Patrick Stuart calls 'imaginative energy'. Good ideas are the ones that immediately make you sit up and pay attention when you read them, the ones that you can instantly and vividly imagine using to great effect in actual play. Originality *helps*, but a powerfully-imagined take on a cliche is better than a pile of random gibberish strung together in the name of novelty any day.

  8. I don't think novelty is vital. The category of answer that is most likely to interest me as a player, I think, is "His name is, um, Dain Steelbeard, etc. Also, his brother was killed by the Orc warchief from chapter 2, he has a crush on the alchemist you bought potions from yesterday, he's working on his blacksmith's master's exam and would like you to help teach him and/or procure ingredients for him, and he's selling iron spike traps that I'm sure you'll be able to come up with an interesting use for.

    You can add that kind of humanising (dwarfising?) detail to a stone and metal robot dwarf if you want to, but I think it will be easier and possibly more effective on a more vanilla dwarf, essentially because it has less of b) and c) in it.

    You wrote a post here where you set out the advantages of a shared standard image palette. I think this is another reason why Dain Steelbeard is an easier canvas to paint a person on than Robot Dain Steelbeard is.

    1. I agree that if you can actually integrate your random NPC into the ongoing narrative on the fly, complete with story hooks, then that's a better option. I think it might be orthogonal to the other issues involved, though.

      As for shared image palettes, I think that a lot of good RPG stuff is just about combining stuff that everyone's already familiar with in effective ways. Like, I read a book recently which featured a goblin tribe that lived in a network of tunnels under an old stone circle, and kept lobotomised human abductees down in the tunnels with them as slaves and emergency food supplies. That's just a combination of three 'stock images' (goblins in tunnels, creepy stone circle, mindless lobotomites), but putting them together gives you a scene that your players are unlikely to forget in a hurry!

      Genuinely unfamiliar stuff is much harder to work with, but it's not really what I'm talking about here. Anyone who knows what dwarves and robots look like can imagine a robot dwarf...

  9. Good Stuff

    d100s i do meant to simplify and present simple cliches and some absurd ideas. You can paint a portrait of adventure area and provide game short cuts easily.

    I loved 40k universe but I found the RPG focus on Calaxis sector so painfully detailed I have no room to move, a canon that none shall pass. For all the minute detail it was logical extension of basic setting, not really imaginative and limited the universe.

    Dense minimalism beats bloat any day

    Personally oldschool gaming for me is about DIY more than product. If big companies vanished tommorrow i wouldnt care.

    1. Honestly, your blog has had a big influence on the way I read RPG supplements. As I skim through them, I quite often find myself thinking: 'This is OK, but it's no better than what you'd get by making a few rolls on one of the random tables from Elfmaids and Octopi. And Chris gives those away for free!'

      'Dense minimalism' is a good phrase. And, to stick with 40K, compare the minimalist way that setting was presented back in the 1980s to the way it gets presented today!

  10. Heroditus wrote a damn good RPG book
    History and classics are better value than glossy coffee table books

  11. This is why Open Page Dungeons are a godsend. Maps and ideas without metaplot (usually) and other bloat.

  12. Great post ! As a good example of density, I reviewed some time ago an adventure for "Beasts & Barbarians". The adventure costs less than... 3$ and is 11 pages long (not counting stats), but the author takes the best of the rule system (SW) and an inventive GM can add some more stuff from his own or other supplements so that there is stuff for several evenings :