Ask Angry: Playing in the Sandbox

Banner Patreon 800 x 100

Do you have a question for The Angry GM! Well, because people can’t get it right, I had to set up a separate page with instructions. About how to send a f$&%ing e-mail. Ready How to Ask Angry to learn how to submit a question.

DM Solomani asks:

You seem to not like sandbox type games. Being an old DM I have used them off and on for 30 years. But usually as a one off adventure, rather than a campaign. I am curious – do you feel it’s possible to even run a sandbox adventure well, let alone a sandbox campaign?

People always do this. They take something I say repeatedly and loudly and assume that somehow means something about my preferences. Just because I say things like “sandbox games piss me off” or “Isle of Dread is everything that is wrong with sandbox games” or “I f$&%ing hate sandbox games,” that doesn’t mean I hate sandbox games. I say a lot of things.

The problem – as always – is that sandbox is a nebulous term and everyone THINKS they know what it means but everyone actually has a different f$&%ing definition. And that makes it hard to discuss. The problem is that sandbox speaks about the structure of a game, and game structure is something that’s on a spectrum.

For example, the true, pure definition of a sandbox game is one in which the players are given no objectives or goals but are merely placed in a game world and told to make their own fun. The game world simply provides tools for the making of said. The best example is a game like Minecraft, though even Minecraft does present a few goals (like kill the Enderdragon). In point of fact, there probably isn’t a pure video game example of a sandbox because the programmers have to choose a finite number of activities and interactions to put in the game and, in the end, those end up limiting the player.

A pure D&D sandbox is what I call a “f$&% around game.” The GM creates an environment. The players do whatever antisocial, pain-in-the-a$&, destructive things they want, and the GM responds to those choices. Those games tend to die out very quickly due to boredom, GM frustration, or the idiot players trying to kill one another until the game just self-destructs. There’s no narrative thread in such a game because as soon as there’s a narrative thread, it’s not really a sandbox anymore.

And that’s actually the deep, dark secret of sandbox games in RPGs. GMs and players THINK they are running sandbox games, but they usually stop being a sandbox very quickly once the players choose one or more goals for themselves.

Take a look at the Minecraft example again. I was playing Minecraft not too long ago. Every so often, I boot it up and start playing. I don’t know why. I just do.

If you’re not familiar, Minecraft is a game in which the world is basically made out of building blocks. You are dropped into the middle of the world somewhere and you start breaking things apart and combining them however you want. You break trees down to make a house, you mine ore to make tools, you build structures, farms, domesticate animals, whatever you want. There are some monsters that roam around the world too. But mostly, you’re just in a big ole world made of building blocks and you can use the blocks to build what you want.

Well, I decided I wanted to build an underground train that took me from my island (where I lived) to the neighboring Isla Del Pollo (which I named because it had chickens on it). If you came in and watched me play at that point, you’d see me hunting for iron, chopping wood, and digging for gold and mystical redstone, all of which were necessary to build a powered rail system. Or you’d see me digging tunnels, widening tunnels, redirecting lava, or bridging chasms. Or you’d see me laying rails. All of those things were parts of the process of building my little train system. But if all you saw was that part of me playing the game, you’d think the game was an elaborate train construction simulation with very oddly blocky graphics.

So, at that point, is the game really sandbox? You might say yes because I chose the goal. But if you compared it to a train construction simulator game, the answer is no.

Likewise, at various times you’d call my game of Minecraft a lighthouse construction simulator, a multi-colored sheep breeding simulator, or a skeleton and zombie hunting game.

And the same thing happens in most sandbox RPGs. The game is only a sandbox until the players choose a goal. After that, the game is a goal-driven, structured RPG adventure like any other. The only difference – and honestly, this is a really minor difference that people make a LOT of dumb noise about – is that the players chose the goal rather than the GM assigning it.

Most sandbox RPGs go even further and offer the players a choice of several goals. This is where properly designed hexcrawls tend to sit. You have a map filled with things to do and the players pick a thing and go do it. And while they are doing it, the game is generally indistinguishable from any other goal-oriented RPG out there.

So, do those things count as sandboxes? Does it even matter?

So, given all of that, why do I make so much noise about how terrible sandboxes are? Well, there’s a couple of reasons.

First of all, I’m sick and tired of the elitist morons out there screaming that a sandbox game is somehow more “pure” because the players choose their goals as the game goes on instead everyone building a game around a goal. That’s a load of bulls$&%. As long as the players and the GM agree on the goals of the game, it doesn’t matter where they came from or when. And once a goal is chosen, a sandbox game is indistinguishable from any other game.

Second of all, the confusion around the term sandbox has lead some GMs (and players) to believe that sandbox games must be devoid of ALL structure. And that’s terrible. In order to be satisfying, narratively, a game should conform to a good narrative structure. It should be well paced. And you lose that in the “f$&% around game.” I hated Arkham City as compared to Arkham Asylum because City had no pacing. Too much was going on and you were free to run around in whatever direction you wanted and could ignore or put off missions whenever you wanted. I know some people get off on that. But I’m too interested in a well-told story to like that s$&%. And I’m definitely tired of hearing how every game should work that way because “it’s more free.”

Third of all, that same confusion often leads to the “find the fun” game. A “find the fun” game is one in which all of the fun things to do are hidden somewhere on the map and the players can’t engage in them until they locate them. Some exploration is okay, but when the game gets bogged down in a “find the fun” slog, it’s gone too far.

So, I don’t HATE sandbox games. And I certainly think they can be done well. My games tend to be a mix of sandbox and structured gameplay precisely because new goals appear as the players discover them and make choices, usually in an organic way. But I hate what sandbox games do to people. I hate the fights around sandbox games. I hate how smug and superior they are. And I hate what happens when they are done wrong. And it’s easy to do them wrong.

Banner Patreon 800 x 100

12 thoughts on “Ask Angry: Playing in the Sandbox

  1. Did you find the main part of Baldur’s Gate 2 equally frustrating? You’re dropped into a large city area and told “Go raise 20,000$”, and that’s about it, but every step you are hit with a different quest you can pursue, or not (It’s plenty easy to raise the funds ignoring 90% of the quests). Personally I loved it, it gave a great feeling of exploration and investigation.

    • There isn’t much of a main story in that part of the game, but all the quests you can take are stories themselves. And since there are only few maps, which are each relatively small, and NPCs are clearly visible, it’s really not hard at all to find pretty much all of these quests pretty quickly.
      Getting the money is only the hook to get you started at looking for quests. Once you’ve started a few you quickly end up picking up new ones faster than you finish them. Instead of one big story, you have lots of small stories, which are mostly each worth playing in their own right.

      The same could be said about Skyrim, but my problem with games like that is all the walking and searching for paths and going through very long dungeons in which nothing happens to get each tiny bit of story. In Baldur’s Gate II the pacing is much faster.

  2. It’s interesting to note that sandbox games aren’t even the best way to satisfy the “exploration” aspect of fun.

    When you played Super Metroid, you pointed out how the exploration worked because of how carefully, linearly designed everything was. And how important that is to avoid leaving the player lost and frustrated. Sandbox games claim they’re great for exploration, but have none of that careful design which makes the exploration fun.

    That said, my preferred way of building adventures is to have a main storyline, but craft enough detail into the setting that it can “go sandbox” anytime the players choose. Because players never want to do what they’re supposed to. It’s like they can smell preparation, and have an instinctual desire to undermine as many GM-hours of prep they can.

  3. Thanks Angry. Nice piece. I had hex crawl type games in mind when asking the question. But now I think back even those games had an overarching goal and the sandbox part was more like a mystery as players tried to find the McGuffin-of-the-Day.

    I am currently running Valley of the Brain Collector which fits that description and is working well. The other one I remember running in a similar sandbox way was Test of the Smoking Eye. Both took a lot of prep work with the characters only seeing a fraction of the locations. Both were fun but not sure it’s worth the prep time.

    DM Solomani

  4. Pingback: The Toolkit: Walking the Earth | The Mad Adventurers Society

  5. The thing that always bothers me about sandbox elitists is the crowing about how players can stumble into situations far above their capacity and must learn to read their environment and be prepared to run. I don’t think that’s a bad thing at all – but most elitists go about it poorly. They sprinkle enormous threats randomly among the hexes without establishing the narrative history of the ecology. The monsters are just there, in their lairs, waiting for the poor adventurers to wander into the wrong cave. There might be some modicum of interaction with the world, such as a dragon who holds a nearby village in her thrall, but for the most part they’re static entities.

    The world could respond naturally to these threats – the ogre in 25B might contract with the orc tribe in 24A so that they can drive out the beholder in Q46; the villagers in thrall to the dragon might secretly establish a deal with a wandering cultist, turning over their allegiance to a dark god in exchange for protection against the dragon – but imposing any sort of narrative at all, even the natural narrative of the evolving world, is often too much for sandbox elitists to bear. And don’t even think of asking for motivations to be applied to these world bosses! They don’t need any motivation other than “TPK or bust”.

  6. Thank you. I get very, very tired of always hearing about the inherent superiority of sandbox games myself – especially since every time I’ve actually played in one, I’ve ended up frustrated and confused. What’s wrong with a good old-fashioned plot, I say? You can give the players the chance to make meaningful decisions and affect the way the story pans out without expecting them to come up with the story for you.

  7. Hey Angry. Recently discovered your blog and loving it.
    Though in this article, I am sorry but I think you’re so wrong I had to comment. I think you started with a false assumption: that sandboxes are somewhat opposed to the idea of having a goal.
    “There’s no narrative thread in such a game because as soon as there’s a narrative thread, it’s not really a sandbox anymore.”
    That’s just not true. Actually, a sandbox doesn’t even mean the DM shouldn’t help/guide the players to work towards a goal, it just means the DM has to make it so that the players can choose another goal, do another thing, or even help the bad guy instead of completing the usual quest by killing it/him. The keyword is freedom, not lack of goals or motivations. Minecraft is a sandbox game (bad one imo, but still) because despite the fact that there’s some kind of goal (to survive, and build up some defenses to be able to continue surviving), you are absolutely free to do whatever you want and choose whichever goal fits you. There is sandbox in almost every TTRPG I played, because even in a “scripted” dunjeon you have a lot of options. You can use different strategies, different tactics, and if things gets intense, most of the time you can just run and get the fuck out of the scripted hell to do something else. That’s “game freedom”, which u may also call options. Sandbox = infinite number of options, or so many that you can consider it infinite.
    Now, I had this argument with a friend who then replied “yeah, but in that case, every game is sandbox, cause there’s no way a game wouldn’t provide at least 2 options (in TTRPG he meant). And he’s kinda right, but not absolutely. I’ve played game where social interactions with important NPCs just couldn’t go wrong (meaning: not the way the DM wanted to), cause the DM made the character forgiving, open-minded, and etc… The players could insult him, slap him, whatever, he ultimately would tell us about how the big bad guy is evil and someone has tu put him down. Even if we killed the guy, there would be another NPC that would do it. If we kill the whole town, I’m pretty sure the big bad guy would come to us and say “Hey guys, I’m evil, come fight me in my tower”. Of course I’m exagerating, but not that much.
    To take a better example (cause you would just say “crap DM” or whatever), let’s take a look at The legend of Zelda. Despite most of the games from the serie being brilliant, using clever game design mechanics and bla bla bla, this is the worst linear game ever. There is NOTHING “sandboxy” in it. I talked about options earlier on, well, this game provides you with one and one only option every single encounter. You find a bow in the dunjeon? Well, you can be sure the only way to beat the boss is using the bow to shoot in a plant that crush the boss or whatever. And if somehow you manage to hit the plant with your sword, that won’t count, you had to use a freaking bow.
    On the opposite, a game from the The Elder Scrolls series will let you a SHITLOAD of options. Found the boss of the “dunjeon”? You can use a bow, a sword, an electric spell, trapping the door and making the boss go through the traps, etc…

    That translates into a TTRPG game in the way the DM has written an encounter. If he wants to narrate a story, he will have some encounter that can only end one, maybe two ways. And that’s cool, you can still make it feel like a sandbox if you let the players get to that way with all the options they could come up with. Anyway, ultimately, if you wanted the big bad guy to win, you just had to say that the bad guy lost his mind and doesn’t recognize allies and ennemies, and there, even if the players want to ally with that bad guy, they can’t, and are forced to fight him. That doesn’t break the sandbox part of it, as long as they can TRY to be ally with him, and they can TRY tons of way to kill that guy. But having the goal (kill that bad guy), having clues about what to do (“go kill that bad guy”), and/or having to do it for logical reasons (“he killed friends of us, he’s a threat for our town, we like to kill bad guys, whatever”), doesn’t mean the game ain’t a sandbox.

    So, I’ve been rambling a lot but my main point is that what you call a “goal-oriented RPG” is actually most of the time a sandbox game. There’s just different degrees of freedom, thus “sandboxy” games. You can have a game that’s a sandbox but has “linear” parts, because you choose story-telling over freedom of choices in specific moments.

    I’dd like to end by saying that anyway, players don’t want an “absolute sandbox”, they just want to feel like they’re in an absolute sandbox, meaning the want to feel like they have the freedom to do whatever they want, even though they can’t really. But if they can’t, it’s not because the DM wouldn’t let it, it’s just because the game universe itself has limits: there will always be bad asses in the game that will kick the ass of the players if they start killing everyone on sight. Players know that, but that’s part of the sandbox: choices have consequences.

  8. I’ve been running the same sandbox world for 26 years and I’m no elitist. My players are not bored. They are not psychotic murderhobos. They are people who want a very specific type of game play. They act. I react. They chase whatever interests them. The world does its own thing around them.

    A sandbox is not goal-oriented or non-goal-oriented. A sandbox is alive. A world that gives zero fucks for the players, but is worrying about itself. The world came first, not the players. Nothing is scaled. Nothing sits in its cavern chamber waiting for the party to happen along. The ecologies exist because it makes for a dynamic world. The global/regional/local events that escalate/de-escalate/vanish happen because the world is alive.

    My sandbox world, anyway, isn’t without structure. It has 40,000 years of history. It has a shit ton of moving parts. The characters in that world are free to interact with whatever they like, and they trust me enough not to pull some Quantum Ogre bullshit on them. Whether or not they chase some narrative goal isn’t what defines a sandbox. What defines it is if the world exists on a scale that doesn’t give a damn about the characters being special snowflakes. They can become that, if they want to, but they need to earn it. To work for it. Most of all, they need to go out and find adventure without having their hands held.

    • See, for how beautiful and romantic the idea of that kind of sandbox is, it doesn’t really make for the best actual play experience unless that awesome play experience was already there. People like it when things make sense, and they like don’t like feeling like they’re being led around by the nose. You don’t need to write 40,000 years of history and make sure to give noncommittal shrugs when your players ask about narrative structure in order to accomplish that.

      The feeling of “Oh my goodness, puppy-huffing gracious! The world is alive!” tends to die pretty quickly, after which all you’re left with is your raw interest hooks and encounter design. Quantum Ogres aren’t a problem unless they’re fucking obvious.

  9. This is largely why I disliked that Kingmaker AP, and why I can’t see how it gets so much praise. I’ll full admit I didn’t play much Kingmaker though.

    What happened was that I joined a Kingmaker game in progress because I wanted to have a single game where I wasn’t the GM that didn’t anger me with how wrong everything was. And our primary goal was to “fill out the map”. I shit you not, it wasn’t even a “find the fun” game, it was a goddamn checklist of zone descriptions with a single battle every 3 in-game days, which we all know ends up being pure tripe in attrition-based games like D&D.

    I luckily came in near the end of filling out that map though, so I figured I’d sit through it. Cue literally 8 hours of “kingdom building” over 2 sessions, where we would go month by month rolling for stats for our kingdom that I didn’t really bother learning because I would rather play the actual game than deal with trash like loot and associated number-wanking.

    I stuck with it though! I stuck with it, because it was obviously the end of a chapter and I was hoping for the next one to be good. “GOOD NEWS EVERYONE!” I felt like the GM said, as we were shunted into a new, empty map full to the fucking brim of unexplored hexes, with little to no explanation for why we would ever want something like that.

    I quit the group almost immediately.

  10. I’m currently running a sandbox campaign.

    The premise was simple. The party is exiled from a city, with no clear objectives. In order to create urgency, I told them that they needed to collect resources such as food and water in order to survive. Given that my game is set in Dark Sun, this is an immediate challenge. From there, I simply let them wander around, and used the lore in the Dark Sun Campaign Setting to give them things to do, quest lines and such.

    Now, as the campaign progresses, the Sandbox feel has given way to a more linear quest-line format. But the sandbox is still there; at any point, my party could reject a given quest line and simply do whatever they please. I prepare session to session, and spend most of my time thinking through every possible choice that the party could make, and obsessively creating contingencies for those choices.

    I don’t think this is just a good way to run D&D. I now believe that this is the *only* way to successfully engage your players. As soon as you railroad them into a questline, you’ve removed the most fun aspect of being a player in D&D, which is the limitlessness of choices available to their characters. It’s unique to D&D; in a video game RPG, the architects *have* to railroad the characters in order to save on REM and production-time. But with D&D, the only limitation is the players’ imagination and the GM’s improvisation.

    This isn’t to say I don’t railroad my players with questlines. Of course I do. I just motivate them with XP bonuses, instead of punishing them for rejecting the questline. Carrot > stick.

Leave a Reply