Story-Driven vs. Sandbox RPG Design

(I don't know why I'm posting so much about RPGs lately; I haven't written an RPG since the days of MUDs, and my current project is totally different... but it's fun to think and talk about anyway!)

So we have several other threads going on now about the proper role of quests and depth of NPCs in RPGs. It occurred to me today that some of the differences of opinion we're seeing may boil down to a distinction we haven't clearly made between different types of RPG design: story-driven vs. sandbox.

A story-driven RPG has a clear, well-developed plot written by the game designers, and to "complete" the game, the player is expected to advance through this plot, participating in heavily scripted developments that unfold as the player reaches certain trigger points (completes certain quests, kills certain monsters, etc.).

A sandbox RPG is one with little or no plot (though it may have plenty of backstory). The player is expected to advance their own character, in whatever direction and by whatever means they choose. Such a game may have no actual ending; you can keep playing as long as you want, though advancing your character may get harder and harder. Or it may end when you reach a certain level (or achieve a singular objective that only a high-level player could possibly do, like getting the Amulet of Yendor.)

I think these lead to a quite different emphasis on various design points. The story-driven RPG can get by with thin NPCs, but needs lots of scripted quests. The sandbox RPG would benefit from deeper NPCs, and can get by with a lot fewer quests (in the limit, no quests at all). And the quests it does have, might best be generated procedurally.

What do you think? What other design implications are there for story-driven vs. sandbox RPGs?

 

I'm gonna say you've got your wires crossed on that one. For something more story driven, I have far higher expectations of good characters being present and of there being far fewer sidequests to spend time one. A sandbox RPG should be 90% sidequests though as that is how you dig into understanding the world, while characters don't really matter because you'll leave the town soon enough anyway.

A story driven game is about the characters, while a sandbox game is about the world.

 

That's why I said designing a game around a dynamic, learning, thinking AI =/= designing a sandbox game.

The sandbox game is... Well golly, it's based on the notion of an actual sandbox. Build castles, dig a moat...

I think I know of a game with really dynamic characters, where the story changes moment by moment and you never really know what's going to happen. It's called... oh wait, I'm thinking of real life. Whoa! I'll be here all week.

 

@RockoDyne: Well isn't that interesting! I see it completely differently. Thus we have the grounds for an interesting discussion.

In a story-driven RPG, the characters have as much depth as the author gave them. So as long as you stick to the script, they'll react appropriately and feel real. But if you do something unexpected, it quickly becomes obvious how thin such story-scripted NPCs are. Tony's "bucket on the head" example is a good one; it's unlikely the author anticipated me sticking a bucket on anyone's head, so the characters will not have any scripted dialog to deal with it, and will probably continue to prattle away as if everything's normal. But this is OK, because it's a story RPG; you're expected to be following the plot, not sticking buckets on heads. That's why I think a story RPG can get away with characters that have no real (simulated) life to them; their lives are like the lives of characters in a book.

But in a sandbox RPG, I will not leave the town soon, and you can't make me. You might leave the town soon, but somebody else might well decide not to leave at all, instead buying a stall and setting up a blacksmith shop, or renovating that abandoned shack by the docks, or shacking up with the mayor's daughter. That's the whole point — the player can do whatever they want. So, if the characters are thin, then this becomes really obvious (much faster than in a story-driven game), and much less interesting to spend time with, to the detriment of gameplay.

As for quests, a story-driven RPG is pretty much moved along entirely by quests, as far as I can tell. Even if they're not handed out and tracked in a quest log, you're told you need to slay such-and-so, and you know the story isn't going to advance until you do.

A sandbox RPG, on the other hand, can do fine without quests at all. Or it can have them (what Teila calls jobs or tasks) just to give you another way to earn money/XP/skill.

 

I'm reposting because it makes my point about the 'bucket on a head' issue. There is no way to account for everything a player might do and that just has to be accepted. Additionally, it's upon the player to suspend disbelief and just accept the game as it is. Think about it, what stops a player in a historical-fantasy environment from creating gunpowder or a steam engine? No game has ever left the player to do EVERYTHING they wanted because there will always be constraints.

The thing about story driven games is that they can have a handful of well written characters, while a sandbox is bound to have a bushel of barely written character.

 

@JoeStrout, great topic and summary of story-driven vs. sandbox. I think you sparked a run on torches and pitchforks with your opinions of NPC depth, though. Perhaps you could say that story-driven NPCs are narratively deep but computationally shallow, while sandbox NPCs are the opposite. Sandbox NPCs can adapt better, but (at least so far) their stories are less nuanced. The thing about writing (or any art) is that you're sharing something with the author, not necessarily with the characters. When a character gets a broken heart, it's like you and the author telepathically share the thought, "yes, I know what it feels like to lose the love of your life. I hear you, I feel where you're coming from." Communion with the author is the strength of story-driven RPGs.

There's something to this. But you're not really building a castle. You're manipulating an (imperfect) simulation of stone blocks. In the same way, you can model an (imperfect) simulation of living things. Your simulation of a human of course isn't going to be the same as a real human, but I think it can be made complex enough to be fun.

I imagine you could have a "sandbox" with just a couple computationally-driven (versus story-driven) NPCs and nothing else. You could interact with each, pit them against each other, set them up with each other, etc.

@JoeStrout - in a sandbox, once you get past the idea that you can do whatever you want, what do you do? From a design standpoint, what compels the player to do something?

 

Now, now... don't bring all your fancy logic in here. Sure you can talk about facts and reason... And you can try to focus on game design... but in the end, this is supposed to be about imagining the possibilities. Don't constrain your thinking based on the limitations of the medium, open up your soul and imagine a romance with an NPC. Oh, what if it was an extramarital NPC romance? Dirty laundry...!

I don't think spending time "hanging out" with NPC's is even the intent of NPC's in the first place.

 

Yeah, what you said!

Right. Just as there are many different levels of simulating a castle — from a static, unchanging, artistically created model of a castle to a bunch of blocks you can arrange any way you want — there are many different levels or approaches to simulating people. A sandbox game benefits from more complex people simulations that you can interact with in more interesting ways.

I imagine there could be lots of different motivations. Exploration is one; the sheer joy of discovery is enough in some cases (although, as @RockoDyne points out, this one leads the player to not spend too much time with the NPCs).

For me, a better motivator is development. Developing the character is OK; developing a piece of the world is better. I'm the guy in our D&D group that keeps saying stuff like, "What, we can cast a permanent light spell on any old object? Why don't we abandon the quest and just open up a shop selling lights? These guys live in squalor and darkness, we'll make a fortune!" (Of course the rest of the group never goes for this.)

So for me, acquiring a bit of property, building or improving a house, running a business, improving the simulated lives of the NPCs around me, dealing with the occasional bandits that come by to terrorize town, etc. would all keep me sandboxing. If the game is designed well, these self-assigned objectives won't necessarily be quick and easy — I may have to go on my own quests to find needed materials, acquire skills, or find people who can help with what I'm trying to do. Or, just buff up my character to the point where I don't have to hide when the bandits come to town.

I think it's the same as with any other sandbox game: provide lots of opportunities for the player to create their own stories, and motivation will mostly take care of itself.

 

And then the DM/GM decides that all permanent light spells dissipate a few hundred yards away from the caster if the spell isn't attached to a lamp made with platinum and twenty masterfully cut diamonds, while an army of angry customers is marching on your shop because you undervalued the gift of magic as something that comes cheap.

The problem is we are talking about videogames. Even if you could invent new technology, the likelihood is that all the NPC's could appraise is the speed of a horse. A car would be such a foreign object that the NPC's wouldn't be capable of interpreting it as something useful. God forbid I extrapolate the argument to space travel. What then, is the game going to start generating completely new celestial bodies? WILL IT EVEN KNOW ABOUT ORBITAL DYNAMICS, NOT JUST SOME NEWTONIAN BULLSHIT?

If you really want to do anything, you're in the wrong medium. Do you want me to start bringing in Peter Molyneux quotes for comparisons of crazy talk?

 

You get them Rocko this stuff is getting out of control every single thread its just the same stuff.

 

You guys need to check it because this is a game design forum not the speculative science fiction forum

 

...that's not a game. That's just using your imagination. That's called playing pretend. That's un-structured play, games are structured play. Why even bother with games? Sounds like what you really want is to transport into a fantasy world where all the characters are really real people. Unfortunately, that's not possible and games are not the gateway to such a world. You should write fantasy novels, that allows you to go much deeper into the lives of your characters and you can share it with the world.

 

I'll assume you've honestly missed the point (rather than just being purposely difficult). The point is, a sandbox RPG lets players do whatever they want as much as possible. There are many MMORPGs where players can buy a house, settle down, and make a peaceful living fishing, crafting, blacksmithing, etc.

In single-player games, that's less common, but not for any good reason as far as I can see — I think it's an element of fun most game designers have simply overlooked.

And since it seems to be necessary, I will add again that not all players are alike. Maybe you don't like sandbox RPGs (or actual role playing, for that matter). That's fine, I'm not going to twist your arm to make you play them. But please don't try to deny the existence of people who DO like them, or try to stop discussion of designing games for those people.

 

You know what, have at it. You obviously are never going to stop creating these threads, so it's up to me to stop reacting to them.

 

So, apart from NPCs and quests, what other design implications are there for story-driven vs. sandbox RPGs?

I'm thinking about crafting, for example... for someone trying to advance a story, too much crafting might feel like a diversion. But for a sandbox game it might be just the thing; it provides ample opportunities for the player to want things (ingredients), and so have to go looking for them or fighting monsters to get them or whatever. Indeed, Minecraft is based almost entirely around this (though Minecraft is quite weak as an RPG, but really strong as a sandbox).

What other impacts does this story- vs. sandbox decision have on your RPG design?

P.S. Of course I realize many games do attempt to combine these. And maybe that even works to some extent. But I don't think it works very well — when I'm in a heavily story-driven RPG and wander off to do some sandboxy things, then I always feel like I'm letting the main plotline characters down, or imagine them sitting around, tapping their foot, wondering why I'm out collecting flowers and crafting baskets when I'm supposed to be fetching the Amulet of Plot Advancement from the Big Bad's slowly cooling corpse.

 

If typical story driven games give the player a few things to do at any given moment, sandbox games give the player a large, mostly uncountable, number of things to do. There is nothing wrong with either, and I quite enjoy both. So I don't have any issue with giving the player a lot that they can do.

The problem I have is taking the "large number of things to do" of a sandbox and thinking "a large number" can in any way approach infinity, much less thinking that the game is going to respond "realistically" to every random player behavior. For a hypothetical example, let's say a player goes into a busy city street, puts on a leotard and a tutu, and starts jumping and twirling around. Is any NPC going to tip that player? Of course not. Most systems wouldn't even be able to tell that the player was even doing anything, with the rest just noticing a spurious pathing issue. Yet every player passing by will realize it's a street performing ballerina.

Yeah... jRPG's have been doing this for over a decade. Look up the Atelier series.

 

Agreed. But the fact that we can't be perfect doesn't mean we shouldn't try to be better.

Neat, I hadn't seen those before. Not exactly what I was imagining, since it seems here that alchemy is a central, required mechanic in the game — failure to accomplish certain tasks within a certain time can cause the game to abruptly end. But it does illustrate the basic idea that crafting in itself can be a fun activity for players, and I think it fits nicely as one of the many activities a sandbox RPG could offer.

 

Ha!! Morrowind had mods where you could romance an NPC and even marry them! They were quite popular. One of my favorites was rather romantic, lots of fun. I have to admit, I prefer hanging out with other player characters than with NPCs though, but I know lots of people who play MMO's but hanging out with people. I always wondered why they didn't play single player games instead!

Could be that if Joe's ideas come to be, then those loner types that enjoy the MMO openness but don't like other people might be a rather large niche group who enjoy his games!

Oh, and I have had fans beg us to let them marry NPCs. Mostly because of the heir issues and they don't want to "marry" another player in game, probably for personal reasons. LOL Of course, there are other ways to get heirs but some players are just strange I guess.

 

I don't want to not respond since you quoted, but at the same time I'm sick and tired of these compulsive AI depth/personal interaction threads and can't handle any more.

 

Can't say I ever had a time issue with them, so it's either really easy or from the very early games or the more recent ones (assuming it's not just talking about the day/night cycle).
The crafting was mostly just a pacing mechanic, so you ended up doing things in the home/HQ and not just running immediately back out into the field to speed through.

 

Then don't read them! Easy fix.

 

Yes, people often get attached to their companion NPCs, despite them being quite shallow — see this for example. (Seriously, read it through — it's a great article.)

I wonder how much deeper (or more common) this sort of attachment would be if NPCs weren't cardboard cut-outs?