Welcome to the Sandbox

In the realm of video games, and, more specifically, in the world of that strange beast that is the Massively Multiplayer Online Roleplaying Game (MMORPG) there are few more sought after and yet largely unfound beasts than the cryptozooic sandbox. As defined on Wikipedia:

[sandbox] is a type of video game level design where a player can roam freely through a virtual world and is given considerable freedom in choosing how or when to approach objectives.

At the dawn of the MMORPG era, there were many aspirants to sandbox-style games, including Ultima Online, EVE, Star Wars GalaxiesShadowbane, Tale in the Desert and more recently, albeit in a non-massively format, the incredible success of Minecraft – but more on that later.

A central point for all of these games – whether ultimately successful or not – was the emphasis on player driven gameplay.  This gameplay evinced itself in several different areas – economic, crafting and territorial control.

While several of these games still exist today, and in some cases continue to thrive, the cold hard fact is that vanishingly few new MMORPG games are built inside an open world, sandbox philosophy.  As a whole, the industry has instead opted for very carefully controlled and contained player experiences, frequently emphasizing story and cinematic experiences and a generally linear path that is difficult if not impossible to deviate from in any significant way.

What happened?  Why did the industry shift gears?

There are segments of the industry who continue to be passionate advocates for the model (I consider myself one of these), but by and large the development community has turned against the model.

There are a lot variations on the arguments against the sandbox model (and at times, this can get a unnecessarily pejorative – “Well, when you have more experience as a developer, you’ll understand…”) but the core argument has more to do, I think, with the trend in the market towards larger, more inclusive games that rely on a broader market. Appealing to a broader market means you need to worry about accessibility, intuitive gameplay, and not screwing the player too badly for making poor choices.

All of these are very hard things to deliver in an uncontrolled, open world environment.  Not impossible despite what some have claimed, but in fairness, non-trivial as well.  Moreover, there remains layered on top of this quandary the deep divisions in design philosophy between the controlled experience and the emergent experience.

The design philosophy of the controlled experience is the opposite end of the spectrum from the sandbox approach, and is essentially what I think of as “cinematic design”.  Designing a game like this (and, note, I don’t think there’s anything wrong with this intrinsically) means you want to control the player experience so you can optimize it, either for accessibility or simply for effect.  Single player horror games like Amnesia, for example, rely heavily on this, and this is hardly unknown in the MMORPG market as well – consider Star Wars: The Old Republic and the astonishing amount of resources the development team sunk into voice acting and cinematics.

The cinematic design philosophy does have a lot to speak for it.  It is more accessible to a broader audience which means more new players making it past that critical first two hours of gameplay.  It also makes it much easier to tell any sort of story.

There is a fundamental problem with the cinematic design philosophy, however, and that is maintaining the pace of content development.  Players will tear through carefully constructed content at a rate that far outpaces the ability of any development team to keep up.  One possible solution to this – though not yet proven out – is to integrate player originated content.  Cryptic’s Neverwinter, for example is aiming to incorporate a more advanced version of the player generated content tools first deployed in Star Trek Online,  but it remains a valid question as to how well this strategy will work in practice.

The sandbox design philosophy, on the other hand, relies on systems and procedural content.  Essentially, the game provides the Lego pieces and the players either in competition or cooperation with each other generate the “stories”.  In Shadowbane this took the form of the rise and fall of guilds and sieges on player-owned cities.  In EVE this took the form of economic and military conglomerates and the competition over resources.

The two reasons I argue are in favor of sandbox design are, first, the benefit in long term retention by players who feel like their choices matter, and second, that the development bottleneck of content is removed.  The downsides, of course, is it is very challenging to make this kind of design balanced for experienced players, and be accessible to new players – both critical components of MMORPGs today.

So where does this leave us?

In short, the current emphasis on cinematic design is understandable in light of the other sea shifts that have overwhelmed the industry over the last ten years.  That being said, I think there is absolutely room for systemic, player driven gameplay, and I think the industry is doing a disservice both to the community and to itself by excessively focusing on cinematic design, especially since cinematic design has its own crippling content deployment challenges that it continues to fail to satisfactorily address.

There are definite problems and challenges to a sandbox systems approach to MMORPG design, but I feel the real reasons for the general abandonment of sandbox design has relatively little to do with the difficulties of designing this kind of game, and much more to do with the market shifts that have emphasized and rewarded games that attract much broader bases of market support, even if this has meant sacrificing long-term retention.

So, much of this can, I think, in fact be argued as a too-fervent favoring of initial eyeballs over player retention.  The current market focus is all about getting new players, and has strayed from retaining existing players.  Where the sandbox model can excel is in the realm of retention, particularly where it can address the balance challenges of its model and the sandbox model’s tendency to excessively punish failure to the point of encouraging players to quit rather than rebuild.

2 thoughts on “Welcome to the Sandbox

  1. I won’t (and suppose can’t either, as I’m a player not a developer) get into the business side of things, but it has always been my firm belief that the “best” way to design an MMO would be a hybrid. A sandbox world — that is to say a meaningful virtual world rather than the themepark “go here for this level bracket and never return” flawed approach — but with pockets of directed content. I’ll use Skyrim as the best example that comes to mind this morning. I don’t have to have a “goal” per se, I can be perfectly content just wandering around killing random monsters, collecting random resources, maybe head to some POIs that I haven’t checked off yet. Often I’m perfectly content to reflect that I may have started off heading in one direction thinking I would do [this] then suddenly I change my mind and go do [that] instead in a totally different direction. But then sometimes I want to have a goal, and that’s when I can activate one of the quests in my journal that are waiting for me and go do that for a bit of directed adventure, possibly some story or lore contained therein that I don’t get from simply wandering around slaughtering wildlife.

    Also I believe while theme park MMOs can also be fine and good in their own right as “mere” games rather than meaningful virtual worlds, that adding pockets of sandbox elements into them can also reinvigorate and revitalize the game and its player base.

    Here’s one for you: your two most recent and most famous additions to Star Trek Online were the Duty Officer system and Fleet Starbase system. Would you describe those as merely systems in and themselves, as directed systems or as sandbox-y systems? Or something else entirely?

    • I agree, some degree of hybridization is both almost inevitable and a good thing, whichever extreme your game is – even Minecraft has its dungeons and monster spawners,

      Skyrim, in fact, encapsulates my own preference for this balance. The game is very systemic at a largescale level, but there are individual Lego pieces of content that are themselves very highly refined, and then combined in many different ways.

      Granted, I am betting most of it is statically placed, but that doesn’t, I think, change the argument; there’s systemic design in the world, and also systemic design in the development implementation, which is more the direction Skyrim did. For example, Skyrim reuses the same mobs, similar AI procedures, similar mission structures extensively, just combining them in different ways.

      I think the duty officer and fleet systems in STO fall into this model as well, actually. Both are highly systemized but with discrete pieces of repeatable, refined content. Assignments are random, which is an example of systemic design in the world, but the modular nature of the system also makes it very fast to build new assignments, rewards and even mechanics by re-using existing pieces in new ways.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s