Here’s an attempt. Beyond a doubt, it is a wordy and flawed attempt, so bear with me if you want and skip the post if you don’t want.
Background
First off, let me lay down some background.
Every game has an element of “smart play” to it. “Smart play” is whatever leads you to success within the game system itself. For instance, if the game system allows xp for both gold and monsters, and gold gives more xp, then smart play focuses on the acquiring of treasure over the slaying of monsters. In games where gold translates to better equipment (and thus more in-game power), smart play still focuses on acquiring treasure, although not necessarily more than whatever grants xp. In games that don’t include story awards, there is no game incentive to rescuing the princess for free, and adventure design usually includes a cash reward for the same.
“Satisfying play” is not necessarily the same thing as smart play. For example, the oft-quoted “greyhawking” of dungeons is an artefact of smart play, but it is often described as unsatisfying. This is one of the primary reasons for including wandering monsters in early editions – the monsters create an offset for greyhawking that allows players to be “smart” while achieving a satisfying experience.
To use a non-rpg example, in the game of Monopoly, smart play at the beginning of the game usually means trying to buy up as much property as you can, and landing in jail is something that is undesirable. Later on, you might have to avoid landing on other people’s hotels, and landing in jail is a good thing. Smart play is entirely dependent upon the rules. It has nothing to do with whether you believe landing in jail is fun or not fun.
It is my belief that, when designing a game, it is desirable to not have smart play interfere with satisfying play.
Sandbox Play
The purpose of sandbox play is not to make each individual game session (or group of sessions) into a single “story” with a beginning, middle, and end. In fact, instead of being a story, within a sandbox a series of “stories” are going on, all at the same time. The DM lays hooks and background material, generally giving a context so that the players can make meaningful decisions. The DM is allowed to make any rules about the world, but he is emphatically not allowed to tell the players what they will do, where they will go, or when they will do it.
The best sandboxes do not just have a single PC group involved in the setting; instead, there are multiple PC groups, all of whom interact with (and through their interaction, change) the setting. The interactions of all of these characters make the setting richer, and that is often the primary draw of a sandbox setting.
It cannot be emphasized too much: The biggest draw of a sandbox setting – for players and DM alike – is the ability to immerse in the setting. This is a function both of the richness of the setting itself, and how the rules interact with the players.
I could (and almost did) write an essay on the job of the DM in a sandbox setting. Instead, let me point out two things:
(1) In a sandbox, there is always more to do than can ever be done. The world is larger than the characters. Players can only be offered meaningful choices between X and Y if, in some sense, choosing X excludes Y. In a sandbox, the most precious commodity is time.
(2) The DM of a sandbox game never tells the players what to do. He merely supplies the context for their choices. It is okay for the DM to say that IF the princess is not rescued THEN Country A goes to war with Country B. It is okay for the DM to say IF you spend too much time greyhawking, THEN you will meet more wandering monsters. It is okay for the DM to say IF you go to the Mountain of Death, THEN you might meet the undead giants. It is most emphatically NOT okay for the DM to say that the PCs must rescue the princess, must not greyhawk the dungeon, or must not go to the Mountain of Death.
One more thing about verisimilitude in a fantasy world. It is my opinion (and YMMV) that richness in a campaign setting – the primary draw of the sandbox – cannot be achieved unless the mundane world is reasonably understandable to the players’ experience. Creatures like wolves and bears are important to encounter. A gloss on a normal ecosystem is important so that players can gain clues from what they see around them. Players must understand how gravity normally works, must be able to assume that there are various normal small animals in the Vesve Forest, etc.
Fantasy elements, of course, exist in this world, but the fantasy elements are largely understandable by (a) how they contrast to the mundane elements, and (b) how they resonate with the players.
Schrödinger’s Wounding, Healing, etc.
The best solution to the wound/hit point problem in 4e is to completely disjoin game-world damage from hit points. Some go so far as to claim that this is what the rules imply, although I do not believe that this is true. Hit points in 4e are still, for example, described as representing damage to some degree, and “healing surges” does not imply shifting damage from one type of representation to another type of representation. “Healing surges” as described strongly imply that they are performing actual healing.
Disjoining in-game wounds from game mechanics leaves the players in a situation where they have no mechanical reason to role-play their characters as wounded. In previous editions, low hit points would cause players to behave more cautiously the more they became wounded. In this way, smart play followed satisfying play, assuming that one believes it is satisfying to have greater (rather than lesser) verisimilitude.
It has been pointed out that healing in earlier editions can be extremely fast, but there is a difference between magical healing and mundane healing. Different people have different points at which verisimilitude is stretched past the breaking point. 3e was very, very close to that point for me, and 4e is well over it. YMMV.
Be that as it may, in episodic play, there is a space between adventures which is narrated by the DM. It is simplicity in itself, in this form of play, for the DM to say “Two months pass, and you’ve healed all of your wounds.” The narrative covers the discrepancy between mechanics and what the mechanics (IMHO) should be describing. Schrödinger’s Wounding exists, but isn’t experienced as a problem, if it is even noticed at all.
4e Enters the Sandbox
The sandbox wants verisimilitude. It wants richness in setting. It wants a ruleset that allows the players to interact with the setting in a way that is both smart (in terms of game rules) and satisfying (in terms of setting).
Satisfying sandbox play requires immersion. Indeed, if you aren’t looking for immersion and decision making, one wonders what value you would get from a sandbox. Immersion requires narration and events that do not strain credibility. You neither want to avoid narrating damage, nor to only narrate damage that is, effectively, no worse than a hangnail.
Smart play includes pressing on, because the most precious commodity in the sandbox is time. The sandbox also, perforce, includes reasons to make the decision to slow down, rest, heal, etc. IOW, in order to allow for satisfying play that is also smart play, both game and DM must include elements that allow the players to be “smart” for making decisions that lead to more satisfying play. Although this sounds odd, increasing the complications of decision making within the sandbox also increases the satisfaction of making those decisions.
(Ed Greenwood brought this up in an excellent article in The Dragon, long ago, and before TSR bought the Forgotten Realms.)
Lacyon has it essentially right when he says
Lacyon said:
It essentially boils down to this*: PCs in a sandbox are reasonably likely to, at some point, be narratively described as pretty physically beaten up, at a point where there's nothing mechanical in the sandbox** to suggest that they should stop and nurse these narrated injuries (or seek magical healing, or whatever). So they don't - they'll push forward to the next tomb, or the next level of the megadungeon, or whatever, and keep going. Which means the next time they're at a reasonable stopping point, they'll have even more narrative physical injuries, with still no mechanical effect. This can continue ad absurdum; even if you like this kind of thing in moderation, it can grate when done to excess.
If one were incredibly dismissive, one might say
If it grates for the sandbox players, they can choose stop and rest - it's their sandbox!
If it grates for the sandbox DM, he can narrate less-severe injuries next time.
But this doesn’t really help. The players in a sandbox don’t want to choose between smart play and satisfying play. Tell them that they must, and they begin to wonder why they are in the sandbox at all. If the DM only narrates less-severe injuries, he (a) removes expected potential consequences from player decisions, thus effectively devaluing them, and (b) removes expected description from the narrative, thus damaging immersion.
Nor is "Then the players need thicker skins Or, of course, they need to watch Die Hard again, and
then get thicker skins" anything more than dismissive of the playstyle, and the problems trying to use 4e causes with that playstyle.
I remember one pre-release blog entry where a designer suggested that 4e was going to be a good edition for sandboxes. It mentioned the possibility of a revamped Forbidden City from
Dwellers of the Forbidden City that would essentially be a big sandbox. I was looking forward to that. Even if I didn't switch editions, I could convert the module!
4e isn't the first edition where problems arise in sandbox play. 3e had two glaring problems: (1) power curve made it harder for players to determine what are appropriate areas in sandbox play, and (2) too much "homework" in designing areas makes it hard for a DM to prep a sandbox.
All IMHO, of course.
And, probably pretty flawed overall......but there you have it.
RC
Or, of course, they need to watch Die Hard again, and then get thicker skins.
