Contextual Specification – Examples.

Through unintentional irony, L.B Jeffries has helped me prove my assertion that:  “Abstract concept can be powerful but are difficult to appreciate without specific examples.” I’d like to claim that I had intended for my previous post to be overly abstract so as to prove a point, but unfortunately it was merely the result of poor editing. Therefore I’m going to continue with some specific examples of embedded and emergent boundaries and their effects on the experience of play.

The overarching context of Far Cry 2 is that of being a mercenary in a war torn African nation. The embedded boundaries of the game present this fictional setting, imposing limits on both logical and physical exploration; where the player is free to go and the actions they are free to take. A part of these embedded boundaries are the mechanics handling the implementation of the buddy system. This includes not only the logical rules explaining buddy behaviour and interaction but also the textures, models, animations and audio lines related to each buddy. All of these are elements embedded in the game, crafted by the designers and artists; immutable. Though different players can meet and interact with different buddy characters the rules governing those interactions and the assets used to present them are selected from a predefined range of possibilities.

The road less traveled…

When I play Far Cry 2 I am bringing, often unintentionally, a set of subjective emergent boundaries with me. My interactions with the character of Nasreen Davar might have relied on the predefined rules and assets that are embedded elements of the game, however my reaction to her was influenced heavily by my own perceptions and beliefs. My personal play experience was still within the confines of that defined by the embedded boundaries (as is the experience of everybody who plays the game), but the specifics of that experience were further shaped by the emergent boundaries I had erected. The motivations I assigned to Nasreen and other characters was not something hardwired into the mechanics of the game, it was an emergent conceit born of my interpretation of the provided fictional context.

Nowhere in the game rules is there anything that explicitly defines a scene of implied rape, however during my time that particular portion of the game it was something I was very conscious of. Nasreen had been contextualised as a female mercenary and through interactions with me had been deemed by the game, and myself, to be my buddy. When she was taken away the embedded boundaries restricted the actions that were available to rescuing her or escaping on my own, but my  own emergent boundaries restricted those two options even further to the singular activity of ensuring she was safe.

The emergent boundaries served to reinforce the embedded boundaries.

A further example of this form of reinforcement can be seen in the way that Far Cry 2 handled physical movement. Though a lot of the country is reachable by road there are various checkpoints on these roads that when approached will cause you to get be fired upon by the mercenaries guarding them. In the abstract: “certain locations on the map have clusters of respawning objects based around them.” As with my previous examples this abstract concept could apply to many situations, however in this case the locations are contextualised as checkpoints and the respawning objects as hostile mercenaries.

Far Cry 2‘s world is one at war, knowing this I came to the game with certain assumptions, certain emergent boundaries. One of which was that as the primary means of travel roads would be guarded and therefore dangerous. Entering the world I naturally tested this assumption, and finding it to be accurate I made it a point to avoid roads as much as possible. Doing so I found I was not attacked as frequently and I would often find diamonds, tapes, and alternate routes to important locations that I would have missed had I not gone off the beaten track.

The context of the game caused me to make certain assumptions, which were reinforced by the mechanics of the game, leading to a change in my play experience.

Emergent behaviours are ones influenced more by players own emergent boundaries than the embedded boundaries provided by the game.

That’ll be expensive… Or not…

This process of reinforcement is not always the case however, consider Grand Theft Auto IV. The context of the game is that of a Serbian immigrant, Niko Bellic, arriving in America ostensibly to meet his successful cousin Roman. It’s clear from the moment you meet him that Roman is not exactly living the “American Dream” and money is a problem. Throughout the game the need for money is brought up as a motivation on numerous occasions often being the driving reason for Niko’s willing participation in criminal activities. For each of these activities Niko is rewarded with a resource which is contextualised as money, specifically American dollars. These are all embedded elements of the game.

Understanding the concept of money, and having a fair idea of how much a dollar is worth on average I was willing to accept being paid certain amounts for certain tasks. However within what felt like an unnaturally short period of time I had earned enough to buy a much better apartment than my current safe house, and live a significantly more affluent lifestyle. Despite this the game still made a point of encouraging me to take on missions to earn money.  The embedded boundaries of the game were in conflict with the emergent boundaries formed from my understanding of the concept of money and the relative worth of the dollar.

My emergent boundaries were undermining the embedded boundaries, there was a conflict, a dissonance between what the game was explicitly telling me was important and what it was implying was important.

I stopped caring about money as a motivation, and subsequently stopped caring about Niko Bellic as a character because his stated motivations were transparent falsehoods, he clearly didn’t need the money.

It all went horribly wrong…

Having just completed a mission for Nasreen I was heading to a distant safe house when my dune buggy was rammed. Jumping out I threw a Molotov at the pursuing vehicle. The Molotov hit the driver setting him on fire and killing him almost instantly. Ducking behind my dune buggy I drew my silenced MP-5 and after a brief game of cat and mouse around some nearby trees I was able to to finish off the second mercenary with a burst to the chest. While I’d been otherwise occupied the fire from my Molotov had ignited their vehicle and as I watched it started to spread toward mine. I sprinted back in an attempt to reach it and drive away before it too could catch fire. I was forced to turn away at the last moment as, already damaged from the initial crash, it exploded, taking a significant portion of my health with it and leaving me standing in the middle of nowhere.

Automatically my finger reached out towards the F9 key, time for a Quickload…

Wait that’s not what happened…

… I pulled out my map, orientated myself with the nearest safe house and started walking. Everything that happened over the next ten minutes, a checkpoint skirmish, a run in with a Zebra, and locating one of The Jackal’s tapes, occurred because I’d managed to blow up my own buggy.

Grand Theft Africa..?

In one sense I’d failed, and in a myriad other games I would simply have reloaded and tried again. In other games there often develops a compulsion to do things ‘correctly’. A need to isolate the optimum route through an encounter so as to maximise efficiency and minimize use of resources: ammunition, medical kits, time. Though I’ve fallen prey to that mentality myself I do wonder why it’s so easy to fall into that mind set. A byproduct of the arcade era when failure meant death and the inevitable need to feed the cabinet more loose change?

Thinking back over games I’ve enjoyed I’ve found the strongest memories are not of moments where my carefully laid plans succeeded, but moments where everything went horribly wrong. Mistiming a blackjack attempt on a Mechanist in Thief II: The Metal Age and having to leap off a balcony to get away; the flight from the police in Grand Theft Auto IV that lead to a head on collision and Niko Bellic’s body flying through the windscreen into the water.

There is a pleasure in succeeding, in forming a plan and executing it flawlessly, but there’s also pleasure, of a different kind, in failure. Consider the moments in games you remember clearest, I’ll go out on a limb and say that likely three quarters of those moments are not when things went according to plan, they are instead those times when everything went horribly, anarchically and brilliantly wrong.

There’s something distinctly personal about failure. If you succeed you are following the the path of dozens of people before you, but each failure, and your reaction to it, is uniquely and specifically yours. Nobody else has failed in quite the same way or under quite the same circumstances as you.

It’s time to embrace failure, accept it and keep playing, adapt and improvise. Some of the most interesting experiences you’ll have with a game will stem from those times when you are required to think on your feet, to react to the unexpected and the calamitous.


I don’t play Grand Theft Auto IV anymore. I have neither reached the conclusion of Niko Bellic’s story, nor achieved anything approaching one hundred percent completion. I have ceased playing because I’ve found that for all this sand-box nature it is inherently devoid of freedom.

I can murder a dozen people or more in the space of a minute and, putting aside the dubious moral nature of such an act, I feel nothing. I was arguably free to murder those people or not, but it is a freedom that doesn’t feel liberating in the slightest. The more I am given free reign to murder, steal and cause chaos the less appealing each of those options becomes. When I have the option to murder one person that is a choice of massive significance and consequence, when I have the choice to murder dozens the action itself means very little.

Liberty City is a manifestation of Immanuel Kant’s Categorical Imperative, when I am free to steal at will the entire notion of property breaks down and without property theft means nothing; when I am free to murder anybody life itself is devalued.

This sensation of devalued choices is compounded by the narrative of Grand Theft Auto IV itself. In the course of the story and character arc of Niko Bellic, murder goes from being of major significance to something he agrees to with little regard simply to earn money he does not need.

As a work of fiction I can accept acts of violence within the context of the world and so do not feel a specific moral outrage at being asked to commit murder at certain points. However each time I am explicitly asked to murder there is a dichotomy between that moment and the rest of the game. Either murder is significant and therefore I should only do it selectively or murder is meaningless in which case having the choice of whether to let a particular character live or not is an empty choice.

Of course murder is hardly unique to Grand Theft Auto IV, it is also the primary means of interacting with the world in real time strategy games or first person shooters. However, at the expense of sounding callous, in those titles the act of murdering a virtual character comes as the final action in a sequence of choices. You decide which weapon to use, which enemy to target; where, when and how. Murder in these types of games is abstracted to the consequence of a sequence of tactical and strategic decisions. In such cases murder has a meaning because, even though it is performed repeatedly, each time the world changes based on your actions.

Murder in Grand Theft Auto IV is made meaningless because often there are no consequences, and it is so easy than you can even do it accidentally; take a corner badly and you’ve probably committed a hit and run without realising.

I am dealing specifically with virtual murder because it’s the most obvious example available, but the core problem remains regardless of the choices on offer. If I have freedom to go anywhere, nowhere is special and therefore the value of choosing where to go is diminished; if I can buy anything, nothing has any value.

The greater your freedom the less that freedom means, the more choices available the less value each choice has in and of itself.

A matter of character.

A cast from all walks of life: young and old; rich and poor; the law abiding, the lawless, and those somewhere in between; prostitutes, thieves, murderers and lawyers. Characters that are exaggerations yet still human and all the more memorable for it. Crime, violence and a think vein of social satire, not to mention racial controversy.

I’m not actually talking about Grand Theft Auto IV, rather the works of Charles Dickens. Even a century after his death Dickens philosophy and style is alive throughout contemporary western media, either via straight adaptations of his work, or through his clear influence on the extensive ensemble cast and interconnecting stories and motivations of shows like The Wire.

Amongst his numerous skills as an author Dickens is probably most remember for his vast wealth of imaginative, memorable and ultimately believable characters. The majority of his stories hang on the actions and motivations of the characters within them, a philosophy game writers would do well to consider.

Most game stories are still about things, objects, the magical plot “MacGuffins”. How often is your primary motivation in a game to locate an object, usually in several pieces each one in a different “corner of the world”?  What memorable stories in other medium have depended as heavily on an object as most game stories do? There are some but the most memorable stories are those about characters. Objects exist in all stories often as symbols or tokens. But those symbols have no power of their own, they mean something to, or are a representation of an aspect of, a particular character.

Steven Gaynor recently wrote an essay where he concluded that: “The greatest aspiration of a game designer is merely to set the stage.” That’s a very solid definition of the goal game designers should be striving for, to create a context in which player actions have meaning. But in a play the stage is a composite entity. There is the physical stage itself, there are the props, and there are the characters.

Games are about action and interaction, and the current technology for interacting with characters is still far from the fluid natural responsivness such a character focus really calls for. Often interactions between players and characters are limited to a set number of choices picked from a dialogue tree or similar. Maybe this is a case of thinking about things the wrong way round?

What I say is important, but what I do is really what people will remember. I can say I’m a friend as much as I like but unless I show that to be true it’s just words. Maybe a player’s actions should take a higher priority in terms of character interaction than a player’s dialogue choices?

Consider a game world full of interesting characters with differing allegiances and motivations. The Liberty City of Grand Theft Auto IV, or the Coketown of Hard Times. Into this world you, the player, are thrust. You have agency within the contextual confines of the world; you can act as you feel is appropriate. Your relationship with the characters in the world is then defined by your actions. If you act in opposition to the goals of one character then they and those aligned with them will grow to resent you. Work in support of the goals of another and they will grow to like you, possibly even help you. The choices you make aren’t limited to those predefined by the story, but by the verbs most appropriate to the situation you are in, your actions are contextualised by the other characters in the world and their perception of you; we are defined by our enemies.

The characters themselves would make decisions on which actions to take next based on the state of the world and their own motivations, potentially using a form of Goal Orientated Action Planning to determine their future plans. Trust, friendship, and betrayal, might potentially be emergent behaviours from such a system. Working in alignment with one character for a long time and then doing something in opposition might make sense to you but be treated as a betrayal by them.

In such a situation your actions would be limited by the boundaries of the simulated world and not by the means of direct character interaction available to you. It might even be better to create such a system without dialogue, a world where actions really do speak louder than words. Imagine being a photo journalist in a war zone where you don’t speak the language, your only interactions with those around you would be through the pictures your take and their reaction to, and interpretation of, them.

It might even be more interesting if the players role in the world was not that of protagonist, or antagonist, and in fact they were simply a supporting character. Then a narrative puppet master could adjust the thematic elements of the world to fit the protagonist’s impressions of the character. Becoming more light and warm if you are friendly towards them, or darker and more oppressive, if you are in conflict.

Characters are what make good stories, not objects. Dickens understood this, as did Agatha Christie, as does Francis Ford Coppola. Believable characters in games are hard, actions are easier, so why not define character through actions?