Thursday, 26 January 2023

The Motivated Sandbox Search or 'Haystack' Campaign

There has long been a trope of fantasy fiction in which the main character or characters have to search a wide geographical area for certain items - often fragments of what is itself a larger item - in order to assemble them into a great artefact of some kind (often to forge a weapon so as to strike against some Dark Lord or other). It is almost certainly inherited as a storytelling device from ancient myths and sagas - in which heroes often have to perform a set of certain 'labours', slay a series of monsters, etc., in order to realise some greater ambition.

I can't remember the first time I ever personally came across this trope (it was almost certainly through reading some Fighting Fantasy gamebook or other) but I instantly latched onto it; when I was about 10 or 11 a friend and I both sat down together to launch what we were sure were bound to be stellar careers as novelists, and my book - The Spear of Eternity - concerned a quest ranging across the continent of Snith in search of the four parts of the eponymous spear, so that it could be reassembled and thrust into the heart of the 'Death Lord', Keshin.

There is something addictive about this structure. It provides a basic formula or pattern which possesses the power to satisfy precisely because of its repetitive nature - we can see what's coming, but, like in an episode of Columbo, the enjoyment derives from the manner in which the resolution comes rather than the ending, which we already basically know. This method of storytelling also builds tension by creating a series of mini-climaxes and denouements before the final moment of victory. And it can, let's face it, give the author a bit of license to really string things out. It is no accident therefore that writers still latch onto it - as JK Rowling did, of course, with Voldemort and his horcruxes. 

(Indeed, my 5 year old daughter is a big fan of the Rainbow Magic books, which consist of seemingly dozens of series, each of which involves the two main characters, Rachel Walker and Kirsty Tate, invariably having to journey across Fairy Land to find seven items which seven fairies have dropped so that they can foil the arch-villain Jack Frost in the final volume. Even to an adult, despite the crushing repetitiveness of the formula, there is a kind of pleasure to be derived from discovering how the authors have managed to write yet another variation on the same, rigid framework.)

Interestingly, the RPG equivalent of this storytelling device exists in that strangest of places - the Lagrange point between a sandbox and a railroad. On the one hand, there is an overarching plot. But on the other, there is no preordained way in which it has to play out: there's a rod with seven parts (or whatever) scattered around the world, and the PCs have to find them. Where they look and, how they go about looking, is up to them. It is perfectly possible to play this kind of campaign in, essentially, a 'sandbox' fashion: 'You have to find the 13 Rings of Yamtether; they're somewhere out there in the continent of Flax; here is a big list of rumours about places they may or may not be found - go!' With a big hexmap and lots of places to explore, you've got yourself quite a game going.

The motivated sandbox search, as it will be henceforth known (but I also rather like the term 'haystack campaign' - feel free to use it) also manages to achieve the difficult task of marrying player agency and freedom with the idea that the PCs are the 'heroes'. Giving them free rein to go wherever they like provided they are searching for the Six Ladles of the Holy Chef in order to slay the Bouillabaise Demon of Doom creates space for a unity of 2nd edition-era 'good guys' D&D with the roguish sensibilities of old school play, and may be worth fiddling with for that reason alone.

(And one could, of course, just as easily run such a campaign entirely within one massive megadungeon - of, say 9 levels, each containing one part of an item which needs all the others in order to be put together into a greater whole.) 


[I am currently running a Kickstarter for the 2nd edition of Yoon-Suin, the renowned campaign toolbox for fantasy games. You can back it here.]


  1. The Teeth of Dahlver-Nar: 32 little relics scattered about, and the PCs will of course want to get them all. They are the tiniest of all of the artifacts and relics, and as such they are very easy to hide.

  2. You fool! There are *seven* Ladles of the Holy Chef! By the wrath of Great Michelin, the stars will darken and become void, and the bitter laments of the diners shall sound across the land.

  3. My own preferred "motivated sandbox" is to have a chief villain, ruling the land or whatnot, who is rather beyond the PCs' ability to defeat by themselves. The players are to depose him or her, typically by uniting or recruiting various factions, finding items of power, gathering intelligence and identifying weaknesses. It makes for a similar game to what you describe but the players' options are more open-ended since they can scheme however they want and the more they do the better their chances of success, rather than hunting down specific MacGuffins.

    1. Yeah, I can see that working - although it requires a lot of planning and moving parts.

    2. One of my best campaigns ever, my Wilderlands Ghinarian Hills Campaign, worked out like that, with the PCs rising in power, uniting factions, ultimately fighting and defeating the Nazi-esque Black Sun. There was almost no planning from me though, and certainly no scripting, just a lot of creation of motivated NPCs and factions, and setting things loose to develop organically in play.

    3. Things like the Sine Nomine faction mechanics can use there too.

  4. "On the one hand, there is an overarching plot. But on the other, there is no preordained way in which it has to play out:.... Where they look and, how they go about looking, is up to them."
    Actually, it's not. The levels structure of existing RPGs precludes this - unless the PCs are of such a high level that they aren't expected to raise in levels throughout the whole thing. And most systems break at such levels anyway. %))) Otherwise, they have to go through locations from the lowest-level one to the highest-level one, in order. Of course, the DM can cheat placing the low-level encounters to wherever they decide to go first, but that's really the same thing. %((

    1. This isn't even true in 5e. They can go to somewhere too dangerous, and learn to run away. They can go to somewhere "too easy", and speed through opposition anticlimactically. Either way they can learn to gather intelligence and investigate things.

      (Support: this is happening in my current 5e plotted-sandbox.)

      It may even be easier in 5e than in some other games, since bounded accuracy, but :shrug:?

    2. I agree with Tom. I do see your point, Mike, but there is a bit more flexibility involved - and of course the campaign wouldn't just be about finding the seven parts of the rod or whatever. The sandbox area would be full of places to explore, investigate, and so on.

    3. I find myself more agreeing than disagreeing with this concern; to a certain extent it is a false freedom. You *can* go after the part in the high-level are but you probably aren't going to successfully acquire the thing.

      It is an interesting idea, although I think JC makes a good point about making it optional rather than the planned central focus of the campaign.

  5. In a megadungeon in particular, there's a kind of halfway house between your 'motivated sandbox' and the pure version - where parts of some greater whole can be chanced upon with their significance only become apparent later: "Among the gnawed bones and tawdry trinkets, you find a strange bit of carved red stone. It looks like part of a statuette - a left arm, perhaps, broken off just above the elbow, with a six-fingered hand pointing or saluting with three digits. The stonework is very fine, and the whole thing is just four or five inches long ..."

    The item might be ignored or discarded or stuffed away in a backpack, but when the PCs find a leg or part of a torso, excitement will automatically ensue - and especially if the two fragments magically meld together.

    Someone, somewhere, will know what the statuette is. And someone else may know what it *does*. The party might even remember things they've seen before to which they attached no significance: "Remember that plinth with the broken foot on it in the room where we fought the skeletons? Was that the right size? Was it RED?". And of course, if the PCs ask around, other interested parties (in two senses!) may prick up their ears.

    This kind of thing is different from the 'motivated sandbox', in that there's no driving *need* for the PCs to find the other bits. But they might create that need for themselves as the campaign goes on (even if it's just to thwart rivals or make a killing from selling the item back to whatever peculiar cult it belongs to), and you get a naturally arising strand of 'story' every time they hear of or discover one of the fragments.

    Beyond a megadungeon, these kind of things are probably too small as needles and the haystack too big; there's no point having the statuette's head in a different dungeon unless there's a trail of clues to it. But in a closed environment, you've got the potential for motivation if the players decided to act on it - and an increased sense of depth even if they don't ("I wonder if those red arms we gave away to the kobolds and newt-men were once part of the same statue. Oh well - I suppose we'll never know ...").

    And of course you can have lots of this kind of thing going on at the same time - and not all need be 'completion' tasks; they could just be connected by commonalities: "The Temple of Knowledge will pay well for anything written in Auld Draconic".