Mar 16

No Man's Sky Versus Exploration

As the June 2016 launch date approaches for No Man's Sky (NMS) -- the game from Hello Games that will offer quintillions of procedurally generated worlds -- questions are coming up about how Hello Games are implementing the exploration they say they want players to enjoy in NMS.

Comments from the developers (quoted below) suggest that, despite building such a vast world, players who try to savor the qualities of a particular place instead of racing around to quickly see large quantities of places will be penalized through various means. That's raised the question: is a game still "about exploration" if features of that game deliberately penalize exploring it deeply? Does exploration only mean shuttling between locations as quickly as possible?

Some have defended Hello Games and their design of NMS by pointing out that, to make a game, developers must impose rules on players -- the rules are, in fact, what define a game. Even open-world games have rules that limit what players can do, just the same as in NMS.

I disagree. My feeling is precisely that it's not the same in No Man's Sky, that this game is different from open-world and other player-centric games in an important way that has an enormous impact on how exploration-friendly a game feels. The best way I can come up with to describe the difference is "systemic consequences versus direct (developer-imposed) consequences."


It's completely correct to say that the developers of a game create the rules of that game. They do indeed dictate what players can do, and what they can't do. But having agreed with that, I hope we can also agree that this still allows developers a lot of leeway in just how much control they impose over player actions. If games are a collaboration between developer and player, specifically implementing features to push players away from some actions and toward others results in a low-collaboration game.

Games built to enable systemic consequences, on the other hand, move the Amount of Collaboration needle in the direction of players. So what do I mean by "systemic consequences," and how does that maximize collaboration between players of a game and the developers of that game?

First, some proposed definitions. Direct consequences are in-game effects imposed directly on player characters through features specifically implemented by the developers to prevent player behaviors they don't want. And systemic consequences are unscripted effects that arise when complex systems bump into each other, sometimes but not necessarily due to player actions.

Open-world games, and games that try to let players choose their own kind of fun -- such as Ultima Underworld, System Shock, and Thief by Looking Glass as well as Deus Ex by Ion Storm -- tend to implement systemic consequences for player actions. Bethesda's open-world Elder Scrolls and Fallout games also favor systemic effects for player actions. You want to kill an NPC? Go ahead. Within certain culturally-imposed limits (e.g., children being invulnerable), you can do that. The developer won't do anything to stop you; you're free to make that choice, and the developer will not have implemented any kind of direct penalty or punishment or reward to try to push you toward or away from that action.

But there can be systemic consequences -- things that happen, or don't happen -- because that NPC is no longer in the world. Maybe they had a unique quest. Maybe they were the only character who knew some interesting information. Maybe they would have helped your character, or opposed your character, if a particular sequence of events was triggered. Maybe they had (were given by the developer) the power to create new things in the game world, and now those things will never exist. Maybe they were abusing another character, and now that they're gone that other character is free to help you.

The point in describing these scenarios is to show that even though the developers may not have deliberately designed some specific direct consequence for killing an NPC (or taking some other action in the world), it's still possible for there to be consequences for actions. The beauty of designing a game to support systemic consequences is that it's not the developer saying, "I won't let you do that" -- it's the player saying, "Hmm, maybe I don't want to do that... or maybe I do!"

In short, systemic consequences respect player choice.

OK, if that's the kind of experience enabled by designing a game to offer systemic consequences, what are "direct consequences" and what's the effect of making a game that strongly favors them?


Direct consequences are effects implemented by a developer for the specific purpose of exerting control over player actions. End the wrong NPC and the game literally ends in a fail state. Try to explore a side path and get your face bitten off by an invulnerable monster, or get blown up by land mines, or smack into an invisible wall, or get killed by an undetectable sniper. FPS games are like this -- their main goal is generating exciting sensations, so exploration that would slow down the game gets discouraged by any means necessary.

And in fact, Valve games are like this, too. They have small areas allowing some tactical improvisation or puzzle solving, but as soon as that's done you're back on the tightly-controlled linear path. Valve are masters of the scripted sequence, and excel at world-building, so that your on-rails trip is fun (and I really enjoy Valve games)... but Valve games take a long time to produce because they spend a lot of time testing and then taking away possibilities from players. The fact that speedrunners manage to find a few loopholes to exploit is more a tribute to the ingenuity and perseverance of the speedrunners than to any intention on Valve's part to let players enjoy their games in the way they (the players) prefer.

In games like these, consequences are there because the developer explicitly implemented them to check player freedom. Here's how Sony's Ryan Clements describes a design decision by Hello Games:

But be wary of what you take. Steal too much from a planet and the omnipresent sentinels — stewards from a bygone era — remove you from the world. Forcefully. Defeat a few of the smaller machines and larger walkers come to finish the job.

That is a feature specifically implemented by the developer to directly alter the in-game behavior of players. That is the developer deciding for all players what kinds of fun are permitted. It's not, "OK, you can do that, but it might have side effects." It's "If you do that, eventually we will punish you for it by killing your character."

A game design preference for direct consequences is a preference for controlling players because the developer doesn't trust them to know how to entertain themselves.

Another example: PC Gamer's Christopher Livingston asked Sean Murray, "How do you know it all works?" Murray's answer:

"I have had to make my peace with the fact that there are..." Murray trails off for a moment. "And I know this sounds bad, but there are bound to be balance-breaking planets out there, that one player will find a way to get rich quick, right? And I think that’s sort of cool."

"Sort of" cool. And he had to "make [his] peace" with some players playing what is basically a single-player game in a way that they find entertaining, despite the systems that Hello Games has implemented to try to prevent those expressions of personalized fun.

Still don't believe me that Murray and Hello Games aren't interested in supporting players being able to have their own kinds of fun? Again from the PC Gamer interview:

Murray tells me he’s already seen players in No Man’s Sky who want to settle in for a while rather than make a hyper-speed dash for the horizon.

"We do find people in playtests just staying on a planet for far longer than I would have expected. And that's cool, I like that. Even in doing demos and stuff like that, which are quite short, you find people finding a hazardous planet, and they just want to see how long they can survive on it. Or you find people being on a hazardous planet, and then they go to a really lush planet, and they feel like 'This is nice! I just want to hang out here, I want to chill,' and that's good as well. They have their own way of playing games and their own particular play style and then they find a planet that suits that better and gives them the experience that they want, and they'll settle on that for a little bit. And they might stay a few hours there and they might decide to move on."

No Man's Sky, however, is built specifically to push players to find new places rather than encouraging them to get to know any one place in particular.

"In terms of features," he says, "in terms of design, decision-making, and stuff, we always say 'Does this encourage exploration or does it not?', basically. And we use that to cut out features. So things like base-building: we don't have base-building. And the reason is, it would just make people want to stay where they are and not explore. And we've built this whole huge universe, and that would be a shame! We want them to go out and explore. Or, for instance, each planet could contain loads of different biomes within it, we could have polar ice caps, and all that kind of thing, but then it wouldn't make you want to go and visit other planets. So we don't have that, and that's really purposeful, and that's our kind of vision for the game."

In other words, Hello Games have decided that "exploration" means "flying around a lot except when forced to collect resources required for flying around except that collecting 'too many' resources will be punished." As PC Gamer's Livingston put it, Hello Games want to "push" players. You want to chill someplace interesting? Well, you can't -- not for long, anyway, even if that means deliberately excluding features from the game. What individual players actually enjoy doing will not be allowed to conflict with the developer's vision for the game.

Now where did I hear something like that "vision thing" before...?

I hate to disturb you when you’re playing SimCity, but I'd like to offer some straight answers on the topic: Always-Connected and why SimCity is not an offline experience.

Always-Connected is a big change from SimCities of the past. It didn't come down as an order from corporate and it isn’t a clandestine strategy to control players. It's fundamental to the vision we had for this SimCity. From the ground up, we designed this game with multiplayer in mind -- using new technology to realize a vision of players connected in regions to create a SimCity that captured the dynamism of the world we live in; a global, ever-changing, social world.


[C]ould we have built a subset offline mode? Yes. But we rejected that idea because it didn't fit with our vision.

Remember how that worked out?


I know this may be coming across as sounding harsh toward Hello Games and NMS, so I want to state clearly that there are aspects to announced NMS features that sound great, starting with the fact that Hello Games are at least trying to make a game that strongly endorses exploration as primary gameplay. That's a Good Thing. The problem is that the descriptions of how they're implementing exploration play are wildly at odds actually letting people explore! And imposing direct consequences for trying to explore, rather than trusting to systemic consequences and to players knowing what they enjoy, is the primary example of that disconnect.

Games designed to favor systemic consequences over direct consequences are a statement that the developer considers entertaining different kinds of players more important than imposing their own "vision" of the correct way to play the game on all players. A game meant primarily as an artistic statement can, and should, unapologetically deliver the developer's personal vision... but a commercial game, which must try to entertain paying customers, is better when it's designed to be more collaborative with players and respectful of their interests.

Building a wonderfully massive, procedurally generated world like that of No Man's Sky, and then implementing direct punishments for players who try to enjoy that world in their own preferred ways, is like building a Ferrari and fitting it with a governor on the engine limiting it to 40 miles an hour.

That, and not people just chilling in the awesome world of No Man's Sky if that's what they enjoy, is what would be a shame.

Sep 15

Environmental Storytelling

Weapon at the ready, you slowly open the door.

No movement ahead. No sounds behind you. You relax slightly and enter the room.

Along a far wall, beneath pale strips of peeling wallpaper, is a bed: a sagging mattress atop broken boards, covered with the decayed brown remnants of a bedspread.

On the bed is a small skeleton. One of its arms reaches toward the room's lone window, and one short leg rests inside a moldy leather brace.

The window is boarded up, but a pale golden light sneaks through the gaps in the planks. On the floor below the window lie a baseball glove and a weathered wooden bat. On the rotted windowsill, glowing in the sunset, sits a baseball bearing the faded signature of someone named "Dizzy Dean."

Around the window are drawn simple images: stick figures playing with a ball on a sunny day; a house with something warm being taken from the oven; a brilliant flash of light; multiple angry stick figures wrestling with two horizontal stick figures drawn with their eyes closed.

You step back and close the door, leaving the room as you found it.

As you read this description of something that might happen in a game, did you create in your mind a story to explain these observations? Do you also do this when actually playing a game?

If so, you've been enjoying "environmental storytelling." It's one of the great pleasures of exploring game worlds.

Environmental storytelling is the art of arranging a careful selection of the objects available in a game world so that they suggest a story to the player who sees them. (I'm using "story" in a fairly loose sense here, so let's not get hung up on what is or isn't a story -- I just mean a sequence of events that has emotional meaning.)

Many games featuring characters in an imagined world tell their stories overtly. Developers may animate a non-player character who talks to your character; they may dramatize events using cutscenes; they may place notebooks or audio logs; they may simply display text as exposition. In all these cases, a story is imparted directly to you, the player, with little or no interpretation necessary.

Environmental storytelling is less direct. Instead of explicitly describing events, environmental storytelling shows the final outcome of a sequence of events, then it invites players to make up their own stories about what happened to cause that outcome. This mode of telling stories is more collaborative than performative: a game's content developer will select objects for a location, arrange those objects in some way that feels meaningful, and then leave the final interpretation of that tableau to the player.

In combination with more direct storytelling, this is a wonderful way to deepen the player's immersion in a virtual world. Players can be told directly what happened and why, and then developers can give that dry information more emotional punch by creating vignettes that allow players, through their characters, to perceive the consequences of the events described elsewhere. As players invent stories to explain these vignettes, they become more immersed in the world of the game. Do it often enough, and well enough, and many players will even begin to discover meaning in things that the developer never thought of.

A game whose developers have taken the time and care to arrange its objects in ways that living people might have done is a game that is a joy to explore, even if the story that some environment tells is not a happy one. Each new scene helps to give the game world more meaning, and makes it a place to which one wants to return, because who knows how many such scenes remain to be found, or were missed the first time through?

It should be said that environmental storytelling probably has the most value in game worlds that, for whatever reason, don't contain many non-hostile NPCs. Beyond just being an additional way to communicate the nature of the game world, the nature of environmental storytelling -- arranging objects to show the outcome of a sequence of events -- is a particularly valuable content delivery tool for game worlds that are post-apocalyptic or rest on the bones of long-lost cultures. Games about things ending benefit from showing stories about the endings of individual people.

It's probably not a coincidence that the two major franchises of Bethesda Softworks -- The Elder Scrolls and Fallout -- are both deeply wrapped around what could be called "the melancholy of lost civilizations." There aren't many people around (to deliver story) because they're mostly dead. So the stories to be told, by those whose lives ended long ago, are rendered through the local environments in which they lived and died.

A skeleton reclining in a bathtub filled with empty gin bottles.

A decayed cake, in the center of a circle of toys, with a party hat placed on a single chair.

A dessicated corpse with one arm draped over a trapped chest.

A skeleton outside the entrance to a bunker, slumped over a chair next to a pistol with one round missing.

A tattered clipbook containing faded pictures of two girls laughing.

A room filled with plungers attached to every conceivable flat surface.

Two skeletons on a bed, their hands touching.

A secret room whose shelves are filled with cheeses of every description.

You get the idea. Behind the door to any room in the game world might lie the last moments, frozen in time, of someone who once lived in that world, and whose ending (as revealed by the last objects that mattered to them) tells you a little something more about that world.

I can't think of many game worlds that wouldn't benefit from some thoughtful environmental storytelling.

Apr 15

Dynamic Ecology: Local and Global

I commented recently on the announcement by OtherSide Entertainment that their in-development Computer Role-Playing Game (CRPG), Underworld Ascendant, would include some form of "dynamic ecology" simulation.

Naturally I did have some questions about that. 🙂

Here's something else Underworld Ascendant (UA) Senior Producer Chris Siegel said about the dynamic ecology idea (from UA's Kickstarter Update #39):

Whether the player [alters the ecology] intentionally or by accident, he will be able to see the effect of his actions in his environment almost immediately.

What I'm curious about at the moment is the "almost immediately" part of that comment.

That will probably make sense in a lot of cases. Players will often have a short-term goal in mind, and (as with the rest of the Innovation Engine) they'll expect to be able to perform some action and then watch as it immediately helps them achieve that goal.

What about longer-term goals? In Chris's example, damming a stream caused the downstream herbivores to move away, frustrating the local predators... but is it plausible that this would happen immediately, right before the player's eyes?

Would it be OK if there are some (not all, just some) goals that can't be attained instantly through any player action, but that instead require time -- and maybe some repeated "encouragement" -- to achieve? Would it be OK if manipulating the local ecology was one of those kinds of effects?

For example, this is basically what agriculture is -- a longish-term consequence of a specific intentional alteration of the local ecology. Domestication of animals is an even longer-term example. I'm not saying these should be part of Underworld Ascendant, just that there can be gameplay value in having effects of changes to an ecology that take time to manifest.

Even better from a game design perspective: in addition to somewhat predictable short-term and long-term effects of fiddling with a local ecology, what about unanticipated effects?

Whether deliberate (by the player or an NPC), accidental, or as the result of some natural physical process, it feels reasonable to me that alterations to an ecology would also have long-term effects that are harder to predict and control. The removal of minerals from soil, for example, is a side-effect of agriculture. The long-term unintentional ecological effect of nitrogen extraction is to cause a place to become unfarmable -- or even a "dust bowl" -- without adopting additional practices such as fertilization and crop rotation.

Again, I'm not suggesting that the nitrogen cycle needs to be part of the dynamic ecology of an exploratory/simulationist game like Underworld Ascendant. There's already a Farming Simulator 2015. 😉 The point I'm making here is that in addition to desired long-term changes due to deliberate modifications of a local ecology, it's more interesting if there may also be side effects that accumulate over time, causing new kinds of gameplay challenges to emerge. That's a much more active world.

Suppose an NPC accidentally drops a naturally-powered magical light source. It rolls down a ravine, bounces off a ledge, floats down a subterranean stream, and finally comes to rest on the sandy shore of a formerly pitch-black grotto. A place that had been dark for years, eons, is now brightly lit. As days pass, weeks, months, is it possible that plants and animals might begin to adapt to the availability of light in that location? I don't mean "blind fish suddenly sprout eyes" (although maybe magic could speed up evolution). I mean different plants might grow; new herbivores could migrate to the grotto to feed on the new plants; new predators might arrive to take advantage of the sudden abundance of prey... and all because of a completely accidental interaction between an ecological input (a light source) and the systemic behaviors designed into the ecology engine. Of course it would be nice to integrate this sort of thing into active gameplay in some way, but for the moment I'm more interested in whether it could happen at all.

Finally, what about non-local ecological changes over the long term? To continue the series of agricultural examples, in the 1600s the practice of fencing animals out of crop areas was replaced with fencing-in. This was a much more efficient practice for maintaining a healthy balance between nitrogen usage and fixing by combining crop and livestock farming. As this process became adopted widely, it ushered in an agricultural revolution that fed many more humans. What had been a local ecological modification spread to change all of Western civilization and beyond.

And to stress this once more, I'm not suggesting that a simulationist action-adventure CRPG needs to care about whether NPC livestock are fenced in or not! This is just another simple example showing how some kinds of local changes to an ecology might spread. Over time -- maybe during the course of a game -- local changes might become non-local changes, altering the very appearance and function of the game world. These changes might be positive, such as growing more food... but they might also be extremely negative, such as spreading a diseased plant throughout the gameworld or eventually attracting the attention of a terrible apex predator.

These particular examples from agriculture are for the most part pretty mundane, so they're not well-suited to an exciting game. But I think they suggest some possibilities for ways in which a dynamic ecological simulation could generate new kinds of challenges -- both immediate/local and long-term/global -- for NPCs to face and for the player to enjoy exploring.

Apr 15

Worldiness and Underworld Ascendant

One of the games currently in development that I'm following from the perspective of an Explorer gamer is Underworld Ascendant. It's being built by OtherSide Entertainment, the reconstitution of the much-missed Looking Glass Studios.

I'm particularly interested in this game, and in this developer, because it -- and they -- are all about designing games so that challenges can be solved in multiple ways. Not only does that mean "exploration" becomes a deliberately-supported way to play a game, I think this player-centric design style is itself a rather Explorer-like mindset.

I'll have more to say about Underworld Ascendant as its development contines. And I will definitely have more to say about Looking Glass Studios and its games, given their centrality to Explorer-oriented gaming! (Although this recent story from Polygon on the demise of Looking Glass and its rebirth as OtherSide is worth reading in the meantime.)

But for now, I'd like to start by talking about the recently-announced "dynamic ecology" feature for Underworld Ascendant ("UA" for short).

The latest update (#39) by Senior Producer Chris Siegel on the UA Kickstarter page describes how the OtherSide development team is thinking of how to include a form of ecological simulation as part of the game world.

The Underworld will have a working ecology. ... Part of the idea here is this just goes on without the player interference. There is a matrix of eat or be eaten that has been going on in the Underworld for, well, forever. And, without any outside interference it could go on indefinitely.

But, enter the player. Talk about a monkey wrench! Players will be able to affect, change and mold the ecology in ways not seen before in games.

This makes me happy. 🙂

For a lot of people, when you say "game" to them, their first thought is of mechanics: the stuff you (through your character) actively, directly do in the world of the game. And that's fine. Mechanics matter.

But I've always been fascinated with the world itself as it's implemented in a game. Doing things is fun... but doing things to what, or with what? "Game" to me means not just what I do, but the feedback loop of the world reacting to what I've done and creating more options to explore.

Part of that comes from the look of a game, and from its characters and stories. For me, though, it's always been mostly about the dynamics -- the active and reactive behaviors of the game itself.

What are the underlying dynamic systems of the world of a game? How do they respond to the player's inputs? How do they interact with each other to generate emergent effects, some (but not necessarily all) of which may be situations with which critters and NPCs and the player can interact, which in turn kick off more consequences leading to still more situations?

A game that only responds in a few carefully-constrained and predictable ways to player input can be fun. But a game that carefully uses simulationist principles to enable the emergence of a living, breathing, dynamic world... that is a special kind of game. Not only is it a unique species of exploratory fun, it is important for game design (and, I think, for the whole game industry) because it is a kind of fun that only computers can offer. Computer gaming needs simulated worlds.

All of which is a long intro to: yay for dynamic ecology. 😀

A (relatively) simple but functional dynamic ecology is a valuable component for building a gameworld that feels like a plausible secondary reality. Simulating the "circle of life" to some level that's appropriate for a game -- meaning it's complex enough to be interesting but simple enough to generate known gameplay utility -- immediately kicks up the worldiness of Underworld Ascendant several notches.

That's not just cool for the sake of being cool. A system like this one, with visible and interactive emergent effects, isn't merely about having another back-of-the-box bullet point ("dynamic ecology!"); it will actually produce additional content for every player to enjoy. Reactive systems are not a waste of time; done well, they are content-generation features that can pay for themselves. In fact, I'm assuming UA's ecology simulation will be integrated with the Innovation Engine so that in addition to being the source of content to be consumed, its systemic behavior means it can be treated as a tool with which to provide useful inputs back into the game.

In Kickstarter Update #39 for UA, Chris Siegel continued his comments on dynamic ecology by suggesting some knock-on effects of tinkering with the local ecology: damming a stream drives off downstream herbivores, causing the local predators to begin raiding a nearby elf village. In the example, that seems like a Bad Thing... but what if you don't like the elves? What if driving them away leaves their village open for occupation by your friends, the dwarves? A dynamic, manipulatable ecology becomes another way to innovate solutions to gameplay challenges and opportunities.

Being able to explore the reactive behaviors of the local ecology, to perceive its patterns, and to manipulate its second-order effects to one's advantage gives players more ways to solve problems beyond "hit it with a stick." I consider that a great example of The Looking Glass Way of Design.

A dynamic ecology means a better world, and a better world means a better game. I am extremely eager to see that world in action.

Mar 15

Super Star Trek (1985)

In 1983, I discovered a file called WHTHSE in the Chemical Engineering files on the campus MVS mainframe.

Curious, I opened the file. It was obviously a computer program, but in a language I didn't recognize, and it seemed to contain references to Klingons... and the Enterprise.

What I'd found was the PL/I source code to a souped-up version of the venerable text game Star Trek, probably based on the original version popularized by Dave Ahl.

Naturally I couldn't resist tweaking it myself.

After teaching myself PL/I, I started modifying this game with a few new twists of my own. A couple of years later, I left the university, but not without dumping a printout of the game's source code.

In 2014, I had just finished a game prototype in DHTML, and it occurred to me it might be fun to see if I could get Super Star Trek running in that more modern language so that today's gamers could enjoy it. In a real sense, this game predated Rogue, so to call it a roguelike is a little inaccurate... and yet it's very accurate, because Super Star Trek includes many of the features we associate with roguelikes today, including permadeath and random generation of worlds. It's a very simple (early) version of these things, and so it's a remarkably harsh and unforgiving game -- but it's not really meant to be won. (Although it can be won.) Mostly it's about seeing the different ways of not winning, because there are many. 🙂

I present Super Star Trek to you in as close to the same way it was originally enjoyed as I could make it. There's no internal help; you were expected to play it a few times and figure out for yourself how it worked. But I can tell you that you can simply hit the Enter key to bring up a list of game commands.

Also, while I've worked to make the game playable in IE 8+, Firefox, Chrome, Opera, and both desktop and mobile (iOS) Safari, there may be some quirks. If you do notice anything seriously wrong, please feel free to let me know.


Mar 15

Let the exploration of games begin!


This is the opening post for the new Explorer Gamer web site.

In the days to come, this site will begin to feature in-depth posts and discussions on gaming from the perspective of those of who enjoy exploring worlds and systems.

We'll think about what it means to be an Explorer gamer. We'll look at different games to see how well they satisfy the Explorer's interest in deep systems and interesting dynamics. And we'll consider ideas for game features that Explorers are likely to find especially satisfying, and think about ways to achieve those ideas.

If your idea of fun when playing a game is to map its possibilities, to combine elements and enjoy being surprised at the results that emerge, to discover new features and new vistas, this site is for you.

And about time, too. 🙂