The promise of sandbox play is that players can choose to do whatever they’re going to enjoy doing in a wide-open environment. In practice, though, it often doesn’t work this way.
Some of the problems come at the beginning of the campaign, when a lack of information prevents players from translating “what is there for our characters to do” into “which things will be fun for me as a player.” This is a pretty well-discussed problem, with excellent suggestions from the classic sources including the West Marches and Rob Conley’s Bat in the Attic.
A problem I haven’t seen discussed as much develops in a sandbox campaign that’s well underway. The players have made a choice about what their characters want to pursue, and they’ve really gotten invested in it. The problem comes when that investment turns the sandbox into a tunnel of the player’s own making.
In the White Sandbox campaign, we saw that happen between the second and third level of Caverns of Thracia. The players had identified “killing the Beast Lord” as the thing that was going to be fun for them. But the intense opposition they faced as they drew near his domain was pushing them towards a style of play they really didn’t enjoy. Hiring a big force of mercenaries and pushing these disposable troops in front of them seemed like the only solution available to them. They wanted to kill the Beast Lord with the same madcap brio they’d dealt with previous encounters, but the way the dungeon was set up made this difficult to impossible. (Ray Weiss told me that this is an emergent property of dungeons stocked using the OD&D procedures; perhaps, as Oban was saying, “saturday night specials” are assumed to become important in this zone, so that the generated treasures no longer have to carry the load of character advancement.)
As the referee, it was really clear to me that the sandbox was full of other dungeons that would support that style of play – many of them also designed by Paul Jaquays. As I watched the players becoming frustrated with Caverns of Thracia, I suggested in increasingly overt terms that they might want to try going on some side treks which I knew they would both enjoy more for themselves, and would also yield the gold and magic items that would allow them to become powerful enough to deal with the Beast Lord’s forces in their accustomed style. But there was a remarkably strong commitment to continuing to bang their heads against the same wall.
This seems to me to potentially give the lie to the sandbox promise: all the opportunities to choose what you’ll enjoy are for naught if you can’t unchoose a previous decision that is proving not to be enjoyable.
In the Adventurer Conqueror King demo we ran earlier, I refereed for characters who were about the same level as the Grey Company of old. The key difference was that the players had also previously played the characters who were the mentors and lieges of these “adventurer”-level characters. In this role, they chose which mission their low-level characters would be assigned to.
The intentional design feature was to highlight the ways that the different spheres of activity in ACKS come together. Over a long-term campaign this can become clear, but in a demo where each player might only participate for a few hours, we needed to highlight right away all the experiences that ACKS supports. Having players switch their viewpoint between three characters at different levels proved to be very successful in this regard. (I initially resisted the idea because I don’t love the troupe style of play in Ars Magica nearly as much as I love its noun-verb spellcasting. I think the difference is that in Ars Magica, a grog will never become a companion will never become a magus, so switching viewpoints feels like playing different games. In ACKS, the organic progression from adventurer to conqueror to king makes switching as natural as reading Conan stories outside the character’s internal chronology).
The unexpected design benefit of this is that it offers a way for players to switch out of the mindset that leads to frustration. In the Abandoned Monastery, the low-level party ran into bugbears tough enough that the characters had to retreat and rest after a single fight. Normally this is where the dogged “never surrender even if it becomes a bitter grind” approach sets in.
But because I thought it might be a good switching point, I said “OK everybody, your lieges don’t want to see you get killed and they do want you to come back with information. Do you want to try to return through the wilderness and report – in which case we’ll play out what the kings do with this new data at the domain level, and you’ll get a pat on the back? Or do you want to go back in the dungeon and get something for yourselves, whether that’s treasure or revenge?”
Framing it this way turned around their initial beat-head-against-wall tendency. I think it’s because it offered a choice where both results would be fun. A choice between admitting defeat and going back for another beating is never fun. So introducing the option of switching to play the characters who had a different range of things to do, related to but possibly independent of the situation with the bughears, restored the wide-open possibility of doing lots of enjoyable things that is, to me, the essence of sandbox play.
One of the design posts of Adventures Great and Glorious mentions that players will play factions instead of characters, which I suspect is going to afford the same kind of anti-frustration switching of perspective as we’ve evolved through the ACKS demo and the playtesting thereof.
Charlatan
Eric Minton
maldoor
Greengoat
David Wellington
James Nostack
tavisallison
The Mule Abides 
I try to be loyal to the plausible pseudo-physics of the fictional world. But when players are miserable for more than one session, I change things up.
If, for example, the players don’t like the grind of battling through to the Beast King – the Beast King might decide to take 95% of his forces and teleport away. The players would quickly crush the remaining “skeleton crew” and capture prisoners who would give them backstory. The Beast King teleported away, and it was mysterious, etc. etc.
On a separate angle of the same issue, I think Grand Theft Auto – San Andreas can teach us a lesson here. GTA can be a very unfun game for me, but I stuck with it because I loved navigating the environment. I loved seeing the detailed landscapes. I loved listening to the radio and blowing off the missions. (It was strangely like the fun of ignoring one’s unpleasant homework to read a pleasant book.)
I think D&D had so much success because its early players were shocked and delighted to navigate an imaginary landscape that resembled fantasy fiction. They were starved for fiction, and they were willing to tolerate un-fun combat in order to get that fiction. Of course, it’s hard for the DM to make a landscape or dungeonscape or cityscape interesting, and our modern palates are more jaded than they were in 1981.
I’ve experienced that same frustration while playing some sandbox-style video games. Your character isn’t powerful enough to tackle the next event in the “main” storyline, so you’re stuck on side-quests and random exploration. Although those elements should make sandboxes fun, they feel like busywork and grinding if all you want to do is continue the main story.
“the sandbox was full of other dungeons that would support that style of play” – Perhaps you could redefine the unused side-quests to convince the players they are part of the larger Beast Lord-plot? For instance, one of Beast Lord’s lieutenants relocated to a nearby dungeon with a small detachment of the horde and plans to assault an undefended village. It’s up to the characters to stop him.
I rarely argue for an Evil GM approach, but if the players are still stubborn and refuse to acknowledge the hint-hint-nudge-nudge warnings that they aren’t ready for the challenge, then perhaps it’s best to let the characters lose. A TPK is harsh, but a year of slavery to the Beast Lord before their eventual escape might teach ’em a lesson.
I prefer the “closed matrix” approach to sand-box, where there are choices, but very specific things happening to drive them to one of a few choices. Then, when things go on for a while, you can get into true sandbox because they’ve established their characters, choices, and have context to choose.
I love the “multiple levels of characters” idea so they can switch when they want to.
I agree with A3. Natural Selection should apply to players.
Tavis, another emergent property of the OD&D treasure rules is the relative value of coin as characters level. For a first level character, 1,000 copper pieces is a nice hoard, even if it weighs as much as two people can reasonably carry. 20 exp is nothing to leave behind. As characters level copper quickly becomes less a treasure than a distraction, then a trick, then a trap. Silver follows. Do two fourth level characters want to burden themselves with 500 sp each, when it means 50 exp each in exchange for a 6″ movement rate on the way out of the dungeon? We have seen transport of gold become a huge mid-level challenge in your white-box game. Both Tobias and Xeno have spent precious spell-research effort to researching solutions to this.
OD&D indicates one early solution to the problem by carefully defining how much each summoned monster can carry. Summoned Djinn and elementals can carry thousands of pounds out of a dungeon, an ability worth its weight in… well you know.
I am wondering how ACKS handles the intersection of relative exp. value and encumbrance. Not only for dungeon adventures, but for the day-to-day of domain management. Do you worry about tax collectors, treasure caravans, local change-houses and similar logistics? Perhaps the King who spends too much time conquering foreign lands gains a robin hood problem at home: a problem that her adventurers can be dispatched to handle (or paid “protection” money to prevent).
To extend the idea of higher level characters leaving behind sections of dungeon full of silver, copper, and recently slain beasts: to what degree is this an opportunity for lower-level parties? Just thinking out loud. Adventurer and scavenger are but a semantic distinction.
In the demo, there were definitely times when the higher-level characters were finished with a dungeon and the players considered sending their lower-level team of PCs in to pick up stuff the original explorers had left behind. In campaign play I expect this would happen often, and be super fast and fun because a simple errand + hand-me-down map + player-calculable risk of terrifying wandering monsters = awesome mini-game.
And to actually address the topic of this post: the original conception of the dungeon was a vast and malleable map where sub-levels were constantly being discovered and passages could appear or disappear overnight. If characters were stymied by some part of the dungeons under Greyhawk, they would only have to wander to a different part. It is only when you see the dungeon as a set-piece that the static, linear nature of it can become a logistic problem. Unfortunately almost every published product I can think of, even brilliant vast constructions like the Caverns of Thracia are necessarily static. (Or seen as static – most of the good ones do point out that the referee needs to change and adopt each product to suit their players).
Once you get into an underworld, “built by generation of mad wizards and insane geniuses” you are talking about an entirely different sort of product. Closer to Vornheim or some of the old Judges Guild products providing random generation tools. The dungeon becomes the sandbox, or at least adopts a lot of the characteristics of one.
So, three possible ways that a dungeon of low-level treasure has come to be:
1. Lair of low level entities
2. Place where earlier low-level adventurers have died
3. Former lair of higher level creatures, now strewn with only the cast-offs of their hoard
… each of which suggests something a bit different about the dungeon itself, and the creatures encountered within.
> player-calculable risk of terrifying wandering monsters = awesome mini-game.
Yes, awesome! Also, if the higher-level characters have actually done the clever job of sneaking around and stealing treasure without actually risking combat, it becomes even more interesting for the lower-level party. I wish I could have been to Gen-con and seen the demo…