DESIGN SINGULARITY

In my free time I’m working on a mod with little robots as main NPCs (you’ll end to see a lot of them). While thinking about the narrative component, I was worried about what makes me care about a character in a videogame if it isn’t playable.

Personally, I use to care at least a minimum to my digital companions, but most of the times it’s because I try to be part of that world, in despite of the poor job of the developers.

So, it’s not about my personal attachment, but the kind of effort that as a developer I can in put in my creation, because in general the good examples are very sparse.

Speaking of robots, two good source of inspiration could be WALL-E from the homonymous Pixar movie and the companion cube from Portal.
They are both robots but they pertain to different media, so they have to adopt different approaches.

You know, movies can use real people so a big part of the problem is automatically resolved, but in games, with the actual technology, it’s the opposite: you take for granted that the NPCs is unreal.
Here I think that the uncanny valley is responsible for this lack of credibility, in fact it’s difficult to take something seriously when it’s so damn pixelated!

Actually the ongoing graphical improvements will permit us to create stunning works of art – some high profile games such as Crysis are a good insight; nevertheless, the overall components of reality doesn’t scale relatively. Characters might seems “real” but they don’t act properly. I argue it’s a problem of interactivity.

Characters who moves in our same environment inside the game are limited to a finite set of actions, designed once forever by the developer. This is a strong remainder of the fiction you are playing in.

However, sometimes a limited interaction could enhance the realism and consequentially the immersion.
I think about Cleverbot, an AI to dialog with by writing sentences on a command line. It’s enough reliable until you understand how to trick it (one minute), but it’s fair.

Returning to the examples made above, WALL-E is far to resemble an human, however it can express a varied range of feelings, and stimulate others in the viewer too.
In this case, it’s mostly useful as a reference on the design of the robotic character: I really like how it’s eyes can change expression simply rotating. The same is valid for all the rest of the cast, where none have big anthropomorphic round eyes stamped upon.

The Companion Cube of Portal is said to create attachment in the player, even if in effect it’s only a clever gimmick that plays with a player’s common behaviour.
Surprising, what I find remarkable isn’t the time that you spend with it, but the part when (spoiler alert) you have to leave it behind, incinerating it too. Actually, it’s an undoable action.

In modern games you can quicksave, regenerate your health using medikits, take advantage of checkpoints, and so on. It’s ok to respect the player by not forcing him to spent more time than he wants, but in our daily reality there are no do-overs. Maybe it’s why we consider NPCs not so seriously.

I pretend to think that narrative in games have not to be linear, thanks to the interactive nature of the medium, to fully take advantage of it. Having a character to merely instructing us about its life is a passive method typical of movies, that works on building empathy with the viewer.

Games more commonly motivate players to care about NPCs by having them to give at your character items, quests or experience points. This seems to me like promoting a relationship with a drug-dealer. You are probably going to like him, or dislike him if he doesn’t feed you. Pretty dull indeed!

What I’ve understood is that we should not treat NPCs like peaces of a puzzle. They should not be part of a problem that the player has to solve or something superimposed.
The dog of Fable II now comes in mind. I thought from the previews that it could have been a great chance to create a credible relationship, but it ends up being immortal – the player can’t dispatch or leave it behind.
It’s even worst if the NPC has different skill than you have, obviously assigned to be exploited to solve puzzles and pass locks. The character appears merely as an extension – an appendices.

The NPC should be a “source of choices”.
From a designer point of view, problems serve to guide the player, instead of choices that serve the purpose to diversify/enrich the experience.
When you start thinking of a NPC on the base of its stats or its role in the game’s economy, no deep connection could appear. You are implicitly not treating him like person, but a cog in a machine that you want to work as efficiently as possible.

As an alternative, if the character is part of a choice of yours and, for example, something bad happens to him by your fault, then a stronger feeling could arouse. Actually you are losing an useful companion that you have chosen to be such, and as you can see this imply a certain degree of not linearity.

This lead to me to think that the NPCs could depend on the player, instead of the contrary; moreover, they could choose to not join us or maintain a certain distance, based on our attitude towards them.

In my mod, having to do with little robots, this seems to work fine. They could be innocent, indefensive, maybe a bit childish, and so dependant on the player. I could also build a mutual relationship – they serve to me as I serve to them.

Any good example not already cited here?

I use to divide the design cognitive process in layers, where each is an abstraction or a specialization of the other one, depending if you start from the up or from the bottom. In the former, you start from an idea, that you contextualize and elaborate to find the core features, that are translated in basic mechanics and actions available to be performed by the player.

During the experience, new actions are introduced, developed and finally exploited. These actions or sets of actions form a pattern, that serves the purpose to keep the player’s interest.

Usually in games you have nested patterns: when the player masters a pattern, a new one is introduced and so on.

In the best games however, all these patterns are critical and interact with a reactive gamespace, i.e. that changes during the play, to create a clean and lean design.

More than difficulty curve and learning curve, this is about interest curve. Any given pattern or loop has only a limited space of maneuver, before the player becomes bored of it. For example, in an FPS there’s only a bunch of situations funny with the same weapon and same enemies, in fact the loop is soon expanded to add in variation, commonly providing different enemies and equipment.

Gamespace is pertinent in defining which part of the loop is accessible to the player: at the beginning the protagonist has a single gun and only then he increases his weapons count.

This has to do with the learning curve, because managing so many guns from the very start may be confusing, but is also about the difficulty curve, because for the same reason may be overwhelming.

In this discussion however, restricting access is the functional approach to handle the interest curve.

A play loop can have a series of variations, but only few will be fun to be played. For example, arming the player with a chaingun and putting him in a level with only snipers *may* not be a real source of fun, even if having to do with chainguns and snipers separately is a real joy.

The restrictions force the player to focus to a subset of the whole possibilities – those interesting permutations that can be showed before having to change loop.

This leads to fun games but that are capable to hold interest only for a limited time before running out of loops for the gamespace. It does not creates a clean and lean design as mentioned at the beginning. It simulates it and works great, but only within tight restrictions.

In the best games, loops are tightly bounded together, eventually feeding back to themselves in meaningful ways. Think about Tetris for example: it has the single loop (falling blocks) that feeds back into itself smoothly.

Katamari is another great example: it has two loops (maneuvering and collecting) linked together, because you have to move to collect, becoming bigger and continuing to move; moreover, Katamari has a dynamic gamespace that changes every time a variations is presented. When a loop is mastered, it simply turns back to an altered original play loop, instead of presenting a new one. There’s no need to slice a subset of the loop to maintain the interest, it’s a reactive gamespace, something that entangle without resorting to pile up secondary loops on secondary loops.

Clean and lean design.

Yeah, it’s not easy :D

+=
New side project, due to be released *soon*. Guess what?

I used to enjoy a lot Diablo 2 at the time, but now a find myself incapable of the same joy.
Well, I had this same feeling also at the time, but it was somehow mitigated by the amount of time at my disposition.

Actually I feel very punished by RPGs that let you choose your upgrade path in a skill tree.
Sooner or later you understand that the path chosen makes your character suck or it’s not as efficient as it could be.

This becomes a painful issue when you can upgrade each level, but you can’t take back and correct your previous decisions. Usually I end stacking up points after points until I decide a meaningful path. This is utterly bad, because most games scale encounters and challenges to your level, so I mostly play under-powered.

The natural attitude to maximize is trapped in those skill tree that let the player choose a “bad” ability. The apparent variety of option lets you image that you could create an equally variety of builds, but it’s more like a series of right and wrongs choices.

RPGs works in such a way: you try a path and then you understand what’s could be bettered the next time, so you restart the game from scratch. This isn’t replayability, it’s annoying trial and error, because I have to lose my progress to enjoy the character I would.

With Mass Effect I finished the game with 11 points to arrange yet. It’s a shame, even if there wouldn’t been sensible changes in the gameplay.

Now, World of Warcraft and others give you the possibility to rearrange all your points, and it’s good, but it’s more an issue with games that in general does not let you work to gain points.

If the new level simply unlocked a new set of skills and the player had to do something (kill stuff, etc.) to buy them, levelling would be only a matter of time.
This soften the concept of error, because the player does not take back from a misstep, but he has to spend more time to develop another path.

This appears to me a rather fairer solution, and also states that now I really prefer to experience the character than perfecting its build.

L4D lacks of the usual static scenes of scripted narrative, but has a bunch of fragmented dialog bits that get recombined algorithmically. Characters' lines are context-sensitive and no more activated by triggers, but the big difference with the standard FPS does not lie in that.


Multiplayer emotion is based on shared resources.

MMOs and LARPs are pretty confident on this practice. L4D is more similar to the former, where primitive emotions are generated by sharing survival resources. In LAPs usually storylines are shared between players, producing a wider set of emotions.

Emotions pivot around constraints, such as how difficult resources are to obtain; in fact, an independent character is likely to express less emotions out of a situation.
In L4D you can heal by yourself, but you absolutely need your teammates to escape from the tongue of a Smoker. It’s not their job to help you, they’re not medics, and it’s generally your fault if you need another aid-kit to continue. You’re going to be thankful, at least.

In MMOs difficult situations lend to two digits groups of people, but relationships are inevitably watered down, because you can’t precisely say who helped you.
L4D provides a more tight experience, because actions and errors are more visible in a squad of four elements. Moreover, the game constantly highlights who helped who, the most generous player, the most active one on killing, etc.
Psychologically speaking, this transparence lend players to be polite and not acting like a jerk.

L4D offers multiple levels of shared resources to lock players together and to cause the emotions desired. For example, the same teammates are resources, because they can advice about an incoming threat (Bill: “I heard a Boomer around here”).

These resources need to be in a divenendo and difficult to preserve, otherwise they lose their interest. From the example above, single missions start with four aid-kits, one per player, but the infected zombies are likely to force you to consume them all, so you strive to reach the next outpost, helping the others to help you to stay alive.

Multiplayer emotion is dynamic storytelling.

Again, the dynamic dialog system has a tangential importance, most because it’s a bit unused in the game.

L4D makes a little step back and a great step forward: Half Life 2 elevated ingame narrative integrating cut-scenes with the gameplay, but eventually the whole structure and pace were pretty static. This heritage is forwarded by the secure rooms, filled with writings on the walls and spare discussions between the survivors.

In L4D, storytelling is more on the vein of Doom and Quake: the plot is not a big deal, usually it’s merely outlined, while the real story is lived directly on the field.
The series of events that happen to the characters forms the story; informations are no more inoculated to a passive audience, but conveyed directly through the gameplay.

This system is obviously more centered on the characters, because the story resolves essentially on what happens to them, but there’s also space for background stories, though very marginal. Take the example of a corpse lying one the ground, covered by a cloth: it says that there were other survivors before, that they had losses and that they had the consideration to cover their friend, but the need to keep going forward.

The big step consists in making these events dynamic.
The usual structure of the story is distilled to obtain single moments, capable of delivering the desired emotions. Those moments are then instanced at will by the AI Director, that analyzes a wide set of parameters to calculate the best moment to cast an event.

Players end to experience different stories each time, made of the same basic events rearranged. For example, a sudden threat on a team-mate (the attack of a boss-infected) causes the short-term goal to save him – what may be labeled as “rescue-event”.

For a good reason, L4D singleplayer campaign is less incisive: without real mates, there’s no social play that sustain cooperative storytelling and interplay emotions.

I’ve resumed this old draft for its pertinence with the previous post.

When I first played The Sims, I was upset by its time conception.
You play this simulation with little characters, hurry to carry out mundane activities in quarters that equals to real one hours.

The time compression is particularly common in games with small characters. The idea is to make the experience denser and bigger than the real life. You may put the whole playtime in a slider, where interesting tasks make the things bumpy: more the thinks per hour more the fun.

Implicitly, it’s said that those activities are too long to be enjoyable in a game. Cooking pasta takes 15 minutes? Let’s do 5 seconds. Running on the beach? Only 2 minutes, otherwise you starve. Travel from city to city? Just a click in the map.

Is always the destination more important than the voyage?

In The Sims I had the impression that time was my enemy: I couldn’t listen to music o play sports freely, because the game forced me to go to the bathroom! Obviously I wasn’t playing in the “right way”, as the game wanted me to do.

The Sims seemed to offer the possibility to play another me, but with that time pressure I wasn’t able to immerse myself in the role. I guess that this perfectly fit our modern life-style, full of tasks and dealings (…thanks to Numa Pompilio’s calendar), but it hinders the focus on the character.

The The Sims example well served me to criticize the general tendency to make recreative events efficient. It’s not by accident that you lose the knowledge of time during pleasant activities.

Lost in Blue for DS followed the same design practice. You are in a beautiful island wondering for a place to fish, but a constantly ticking clock reminds you to move forward. For me, this pressure is unsuitable and ditches the impetus to walk around and talk to people.

On surface Bully by Rockstar seems like a GTA-clone but its open world is regulated by a clock in the upper right of the screen. I’ve appreciated the whole system, unfortunately you have no ways to judge how much a task will take, nevertheless, if you’re yet around for 2 am, inevitably the character falls down exhausted (andthe active mission is aborted).

There’s a good reason why there’s no clock in casinos! Would you remind someone who’s giving you lots of money that it’s late to play?

What makes the difference between a vivid virtual world and a mere agglomeration of polygons? Is enough to reproduce every detail and hope that the life springs out magically? In other words, what else do you need other than the sum of the all parts?
It does not to be efficient!

I’m going to help me with examples from the Star Trek universe.
The TNG series used to build a story around every element, from the ignition button of Data to the flute of Captain Picard. This made the experience pretty fake for me.

In the real world, a stone is only a stone. We don’t bother about the story of that particular stone, except if it’s set in circle at Stonehenge, and it's right this that make them so relevant: there’s no other equal stone structures*.
Uniqueness is such thanks for an amount of other non-unique stuff. In the same vibe, we can define what is alive by defining what is not (do you remember Eraclito of Efeso?).

The DS9 series of ST is different: it’s a live world full of insignificant details, maybe thanks to the fact that is not filmed in the same four stages of the Enterprise starship.

Being efficient is part of the development philosophy, because making a game is very expensive, but the urge to fill more content in the same space is counter-productive, if you’re going to build a seamless world.

In Oblivion (or Fallout, GTA, etc.) you have a primary quest and a bunch of minor ones. Why could we not have only one massive quest?
Well, not only it’s not time-wise player’s friendly, but it’ll end seeming like a castle built in air.

That mass of secondary quests appears as the background for your main adventure, it let you remember that there’s something alive outside the boundaries of the “dungeon where you’re battling”.
For the same reason, I usually don’t play a lot of side quests, because if I finish them all, the world instantly lost its realness.

Mass Effect turns mainly around its big quest, substituting side stories with secondary activities. You can land with your MAKO tank on a planet surface and gather resources, and that's all.
It take no time to understand that those sections are pretty disjointed and, in the end, fake. What happens there does not affect your main voyage in any manner. Fortunately, the main quest is good enough, so you don’t bother much about the lifeless void left outside.

Far from being the only reason, inefficiency seems paramount for a real virtual world… or have you named all the stones in your courtyard?

*it's the same for other megalithic sites in Armenia, Scotland, Germany, etc.

Personally, I’ve spent a good amount of time thinking about the implications of rendering space, structure and relationships in games.

During the eighteenth and nineteenth century we had systems which altered or augmented perception. Take for example the diorama and the camera lucida - both examples of representation machines. From the same point of view, we could read the spatial qualities of games as a testing ground for exploring perception, physics and geometry.

Far from being enough, the results of that exploration should lead to impact the gameplay, because games are about navigating space, with only few exceptions. From the open levels of a first person shooter to the restrictive dialog tree of a RPG, it’s such a deeply integral element that it’s natural to tweak and explore its possibilities.

The very act of moving around the game world and looking around is mostly the first thing to have in place, at the beginning of the development.

It’s meant to test the design, but actually it remains the same whereas other gameplay loops are exploited (shooting, building, collecting experience, etc.).

Many games treat space too generically, without taking advantage of rules and tradeoffs inherent in it.

In those games, space is treated as very generic, because the only difference between here and there is that here is closer to that arbitrary point, chosen by the designer.

I’d like to cite few examples to further explain the point.

Chess has a space as a set of hard and chunky coordinates, meant to be “explored” on a turn-base. Then Pong combined those chunky coordinates with a real-time speed. And became so popular! …Do you know Pong right?
Now, its addictiveness is almost faded away because it’s no more a brand-new way of navigating space that few people had conceived of.

Gauntlet is another milestone in its take about space. You can see much further than you can walk (compare it with Super Mario Bros.), so you have a big reaction time but, being in a maze, you really don’t know which way to go.

Sonic the Hedgehog is almost the opposite. It’s fast and presents complex levels.
Usually it’s compared with Mario, but the pace is pretty different and much of the content could only be explored racing through at top speed.

Soul Reaver had a fascinating way to represent space, having two different versions of the game world, connected to overcome obstacles and enemies, because moving through one moved you through the other.

I’m not going to cite only popular titles to sustain my point. The truth is that these kind of twists about space are really fun to play with. In fact, there’s a bunch of more recent games that revolve their focus on navigating space, and springing the fun out of it.

Prince of Persia, and its spiritual successor Assassin’s Creed, allows you to do fantastic acrobatics in same kind of space, where your run of the mill FPS allowed you only to walk forward and jump.

Psychonauts twists and breaks the norms of navigation in a lot of its levels. Moreover it manages to have a kind of variety and novelty in themes that we can find only backtracking to Super Mario 3.

Echochrome for PlayStation allows to direct a mannequin through levels that resembles an impossible construction of Oscar Reutersvärd. Each level is an isometric projection and is navigated through perceived geometry rather than actual geometry. Optical illusion is the essence of the navigation.

The indie-game Fez is another title which explores the idea of a reconfigurable vantage point. It’s a side-scroller in which the camera can be rotated that extends the representation shifting meme, much like Paper Mario and Crush made.

Portal revolves around the use of wormholes to shoot into surfaces, allowing you to pass and cut through space in a non-linear manner.

All these games, and many others, change the ways you look at and move through levels. It’s their tweaking of the everyday norm that sets them apart from other games.
I can’t advice any reason for not continuing to tweak this way, a lot!

Indie-games have an easier life experimenting about changing how you travel, how you think of distance, even the nature of space itself.

For example, Determinance and Parabox try with different extents to do that, one letting the player fly instead of walk, the other allowing the player to rotate the world to fall and move forward.

A logical solution, to make the way a player perceives and travels through space intrinsically linked to some part of gameplay, is mapping space to something besides space.

Both my solution and recent trope is linking one or more parameters to space. The idea is that changing the position changes the game dynamics.

If you think about it, falling is basically that – a simply modifier on z-axis: you understand that moving here (ex. upon a pitfall) sends you there (ex. down to death). Absolutely nothing nearly complex.

Another similar example, taken from the real world: as you walk higher a mountain, oxygen gets fewer and it’s harder to breath. Translating it in a game context, there is this relationship that affects your gameplay in an interesting way (challenge for climbers).

You’ll find out that the trope where you stand affects how you play is capable of so many dynamics.

Imagine a game where space is mapped to daylight circle. Walking right you can see the night, walking left you illuminate the environment. If this was the case of a side-scroller, it could reveal enemies that usually hide in shadow. Then you might have to dispatch them, who are invisible when you are near enough to shoot.

How about a game where moving right means changing your evolution state, from larva to homo-sapiens?
You can really map space to anything, from the gravity to the health-points of your opponents.
I can’t bother about increasing in complexity if amazing-ness continues to seems so proportional.

Then, we haven't even touched different ways of viewing space. I’m trying just that in Parabox, representing the forth dimension of a 3d game-space, in 2d.

Mirror’s Edge is in touch with that too. The protagonist Faith possesses the ability for reading the geometry of her surroundings. In the playfield, objects and surfaces that comprise the environment are colour coded according to accessibility. Considered how architectural elements are demarcated, the city becomes an huge text to read to navigate. The gameplay revolves around wayfinding and spatial problem solving rather than button pressing bravura.

The example above of Super Paper Mario is also one of this kind. Space is portrayed in 2d and suddenly you learn to carve a third dimension in the scene that opens to you other possibilities.

Then, think about the vision mechanic of Metroid Prime. You can extrapolate different information from the environment, based on which lens you equip.

Obviously there’s a lot of space for improvements.

I remind Collage Rebus II from 1970. In his drawing Daniel Libeskind delineates an interiority that is more about arranging linked rooms than rendering plausible spaces. Curvilinear walls irrationally subdivide the assemblage of rooms, ladder lead to nowhere and edges ends in white space.
This hypothetical design describes unbuildable architecture. It’s an interesting example of spatial reasoning devoid of interactivity. Collage Rebus is frozen in motion without any navigable configuration.

Then, the field of Augmented Reality offers good suggestions about how to change your view of the reality, maintaining interactivity in real-time.

Well, I have a lot of thoughts on this matter. More to come in part 2.