Collision: Crafting Memories

“Some memories are realities, and are better than anything that can ever happen to one again.” ~ Willa Sibert Cather

“I’m in my tree talking to the Dixie Chicks and they’re making me feel better.” ~  Dawn Campbell (I Heart Huckabees)

A little while ago I embarked on a project that was borne from a number of things; genuine inspiration from the work of others; my desire to make a mark with my rhetoric about game design; and to get noticed or perhaps just respected by a few. Now the same project has, like all good things become intensely personal, and an attempt, finally, to make sense of the world a bit in a way I hope will speak to others. This piece is an attempt to force my hand into working faster by “letting it out” about the project, and more likely is, like all my games writing, to encourage a conversation so we can all feel smart for a while. With that in mind I’ll cut to the chase, but not before you go through the steps I did:

The inspiration:

It may strike you as odd that I’d make my own foray into modding sound utterly derivative by taking about what inspired me to make one, with specific examples. I will only go on to talk about them because they taught me a number of lessons, and it seems a lot more dishonest if I just drop the names later as though they had nothing to do with it. No, I learned a few things and these clashing with my ideas brought other new ideas, and I now confidently feel I can discuss what unifies and separates our respective philosophies on game design in the first person. The mods: Dear Esther, Korsakovia, and Radiator.

The clue was “mods” wasn’t it? The mod is cheap way of developing something that looks like a proper game but isn’t. I’m not in this for the money, and I can’t fund a full game so yes, my project is an experimental mod that – like a pilot fish – will mooch the algal particles of assets off the back off Half Life 2 (and probably Episode 2 as well). Note to self: thank them in the credits.

So I decided for better or worse about two years ago that I was one of those “games are art” people. I wasn’t going to make a big thing about it, and to be frank it was off the back of Braid. In a lot of ways I have a lot to thank Jonathan Blow for. He got me through the rabbit hole, really, but like Alice I’ve discovered the White Rabbit is sometimes a pretentious ass and that he was not the only one down there (I’m not sure if Alice really did think that about the White Rabbit, but we’ll never know now). I had never bought into being spoken down to as a “gamer dude”, and that culture of achievements, rivalries, high-scores, and “clocking” games never appealed to me. I suppose I was always that “arty” person but it felt like it needed declaring if this game thing was going to go any further. I needed to put that stamp on my identity in case I ever got to GDC. Fast forward a year and a half, and I’m writing a blog (like everyone else) and more importantly I’m trying to recreate my workplace (a pub) in Hammer (for Left 4 Dead) so I can learn how to use it.

This was the first step in actually making something, so when I came to play Dear Esther and co. it shouldn’t be surprising that the mod as a medium for my ideas would come to the fore as a potential outlet that would actually get played. Enough background for now… I won’t explain what the mods are, I’ll provide links to each one, and here is a link to a previous piece I wrote on them.

Dear Esther by Dan Pinchbeck and thechineseroom (music and sound design by the marvelous Jessica Curry) was a “Braid moment” for modders. It divided people – some heralding it as amazing and a landmark in game design, and others questioning its right to exist at all, calling it a waste of time, and “not a game” (as though that’s an insult). To me the truth lies somewhere between the two, and not just because I quite like it and I’m taking a moderate view. Partially I am, maybe, but more importantly it struck me quite clearly that no, Dear Esther is not a game. And it is a landmark. Sorry if that makes me a pretentious blowhard, but then maybe I am pretentious and I blow hard.

Before I played Dear Esther I thought I had a nice-sounding, valid, researched, theory on what made a good gaming experience. It was elegant, I even disagreed with Richard Rouse III with it (on my own – listening to a podcast). I’ll summarise: a really good game is a combination of two core things; ludology (obviously! Ha!) and narrative. These things weren’t created as though one compensates for the other and you make concessions, no, no (as though this is easy), you weave them around each other like some sort of Persian rug-maker. The idea being that a game’s ludology represents the core themes of the narrative, and vice-versa – so that changing the style of play in said metaphorical game – would then make it feel like a stilted unfinished experience, and changing the narrative would make it feel thematically jarring. I’m not sure how you extrapolate that to the rug. Something about loose strands, probably.

You could argue that Dear Esther has one mechanic: walking. Substitute “walking” with “exploring” and then think of its narrative. There is certainly a theme of exploring – there is an explorer mentioned, a car journey which is never explained, and more importantly the player explores their morality when ruminating on the issues of the narrative. Does that sound like a stretch? It might be. It’s actually evidence of trying too hard to fit a work into our own restrictive theories on game design I feel. I often think we try as critics to cushion our interpretation of new game experiences in our accumulated wisdom a little too quickly – it’s not all our fault, though, any game has connotations to something else via the medium’s existence (you can’t watch a movie without thinking “I am watching a movie.”) so it becomes difficult to apply fresh eyes to everything. Dear Esther’s achievement is that, in quite a simple way, it encouraged us to do that.

The method was alarmingly simple to me (and unfortunately unrepeatable by me). It was a Half Life 2 mod that bore absolutely no similarity (besides a few unfortunate hangovers – sound effects for one) to the original game. Not just the setting, or the voice acting, the length, or the music, but there is no shooting, no gravity gun, no physics puzzles, and no head-crabs. “What is this?” is a question that comes from the second the voiceover kicks in, and everything you see and hear just repeats it. It seduces you like that into caring about the voice work, the symbols, the red light, and the apparitions. If it were a movie it would be a bit schlocky, but because you have no idea if where you are is meant to be real or meant to be a dream/hallucination/mixture of both, you are constantly asking yourself “What is this?” because your instincts shrug unknowingly. You’ve seen nothing like it.

As I hinted there are two things I got from Dear Esther: 1. If you give people fragments of narrative they will try to fill in the blanks themselves, and give coherence to an incoherent narrative – A practice which people find engaging, and from the ambiguity find powerful. And 2. That the most important thing you can do with a game space is exploit its lack of reality to aid narrative, and create such ambiguity.

Korsakovia, the follow up (though certainly not a sequel) taught me a whole bunch of things. Things I probably won’t be using in this project since my aim isn’t to scare the living crap out of people, nor is it to intentionally frustrate them. That said, my project does aim to confound the expectations people have when they approach the mechanics of games, and it does intend to exploit the thing Korsakovia really got right: Unreality.

Korsakovia simulated the experience of one man (Christoper) in a state of extremely destructive psychosis suffering from Korsakoff’s syndrome. The narrative plays with unreality by putting us into his world in his mind, and by offering us audio of conversations with his doctor which are triggered at specific moments in the action, which – combined – blur the lines between reality and fantasy (at one point it becomes clear that the doctor may also have once been real, but at an unknown point have been merged into the fantasy).

The videogame is the perfect tool to provide this experience as videogames are synthetic from the ground up. Because games are made in engines with underlying systems, it’s as though they exist independently in a nondescript universe of intangible natural law. This means that Christopher’s delusional world is, in terms of a game, just as real as any other in-game environment – which neatly carries over the message that Christopher’s world is just as real to him as any other world. This is the key thing to be learned from Korsakovia. Anything is real (or not) when you put it on the screen.

I actually didn’t play this mod a great deal, as it was very effective in making the player uncomfortable. The combination of aggressively chaotic sound design (it’s thematically relevant, and disturbing in both its nature and how it is mixed), intentionally vague sign-posting and level design, and frustrating “combat” made Christopher’s mind, aptly, a place I didn’t want to spend much time in. No doubt I have learned a few things, then, about creating an unreal environment, and about a mind I wouldn’t mind being in. Its only real failing was that it used too many recognizable Half Life assets, including the iconic crowbar of Gordon Freeman – which is evidence that supports the points made about Dear Esther – people didn’t care as much about Korsakovia, because it seemed cheaper for the reuse of assets (despite costing £10,000 to make), and that recycling gave people something to latch onto. In turn they didn’t question it as much. It became a horror mod that unfortunately lost some respect for not re-skinning a silly old crowbar. This is a pity, because in some ways this is the most inspiring one of the set.

In a close second, at least in approaching the project, was Radiator by Robert Yang. Astute readers will have noticed I headed this piece with a quote about memory, and memories feature in Radiator as spaces the player can explore briefly. More astute readers will note that memories feature in all of these mods, and arguably the other two mods are set entirely in memory. While the flashbacks that can be visited briefly in Radiator Ep. 2: Handle With Care are explorable there is no narrative or ludological consequence to anything you do in them, you just observe – and they are practically motionless, like dioramas – not really like memories at all.  In fact, while visually striking, the flashbacks are perhaps the only moments where it really lets itself down. No, my inspiration from Radiator was in the sense that it had characters that seemed a bit more like real people, just nuanced enough to make you fill the gaps, and the things I really took from it were a general tone I liked and, in Episode 1: Polaris, multiple endings that were handled in a specific way. I’ll let you make what you will of that, as it seems counter-productive to go on that line of thought. I will say that Radiator is actually my favourite of the three to play, but memories were a device in the plot of Radiator and none of them mattered in particular, they were just beans to be counted, but in my own work this is not the case.


So after I played these I was left with a desire to make a mod, and specifically to make one that was not a traditional game in any sense. I didn’t want to make a mod where you walked around setting off audio triggers, and nor did I want to make a mod which put you into a situation like Handle With Care did – with crazy box puzzles (they make sense in the interior logic of the mod, but I felt as though it didn’t encourage much thought once you’d gained its significance, making the episode something of a one-trick-pony). There had to be a way to utilize the unreality of a game environment to suggest a theme without battering you over the head with it. There must be a way to utilize this unreality to create an experience that doesn’t suppose it knows what “success” and “failure” is (not that I’m claiming the above mods did, just a contention I often have with games), and there must be a way to make you play without it being a “game” but more natural than that. I wanted to move people with a narrative that would make them think, that was ambiguous and left open to interpretation, but not paralyzingly vague.

How do you make something feel grounded in human emotions, ambiguity, but that doesn’t seem contrived or far fetched? You put it in the emotional, vague, ambiguous parts of our brain – our memories and dreams. I liked that the experimental mods I’d played also toyed with metaphor, too, but I felt like I could stray too far from my goals if I made a metaphorical situation the core of the ludology. Having said this, the main event of the mod and what everything spirals from is based on a metaphor – a car crash that signifies a turning point in the course of the relationship between the protagonist and his partner.

Though “games” are treated somewhat patronizingly as pure entertainment , I believe there is a lot more possible. The first thing I felt I had to do was tear away the notions of success and failure. They do not exist in art, and they are illusions in life – there is no great “success” that everyone may bask in. It is entirely subjective. Society is built on rules and laws – like a game, but  it is easy to look at life through a lens and see it’s a game where no one wins or loses, because our defintions of winning and losing are not concrete.  Life is game you can’t win or lose. It is a a game in which you can choose to ignore the success and fail states of other players (religion, politics, and philosophy). With this in mind, why on earth would I make a game about life I could win?

I want to make the rules that govern our behaviour in game as divisive and vague as they are in real life, and for those interactions to be governed by rules that are transitory that reflect ever evolving situations. The idea is to make people think about how and what they are doing in a videogame more than they usually have to, and not by threatening them with failure or by funnelling them into it, but by instilling in the player the coherence of the game universe that is not as hugely removed from the real world like we are used to. Of course this is practically impossible to do accurately. My intention is to focus on social interaction. It’s remarkably simple, really: you will be defined by what you have done, not what you meant to do. In games traditionally, for instance, going to speak to a character will often let you get away with a great deal before rewarding you for finally pressing the “talk” button. I want to make a game that recognises you tried to jump on a table and drop kick the person you’re about to talk to and, accordingly, thinks you’re a twat.

The interest in memory sprang, however, because it would be wrong to assume the opinions of others were the sole manner in which we judged ourselves, and our life experiences. Our lingering, important, memories and our interpretation of them over time go a long way to defining who we are at any one time. Memories and how they work  form determine our success an failure states in life – we push forward, and to do that we look back. The rub is these success states are ephemeral uncertain places that are sometimes conflicted, with our definitions ever-changing and contextually driven by our psyche in relation to our circumstance. I talk of the difference between wistful regret of what might have been one moment, and quiet gratitude for the same life experience the next. From deciding that a memory is painful, to then deciding it was necessary, valuable, and in some ways exciting and vital to making us who we are. Game narratives and mechanics combined have shown us the complexities of making decisions by facing us with the possible consequences, but they rarely give us an opportunity to ruminate on why our opinion on our memories changes over time. It would seem like an impossible task, to try and experience a memory as though the memory itself was old, changing, and degrading into a blissfully bright old fable – but in the unreal world of gaming I can see no reason why not to try.

The idea that fail states are interpreted qualitatively by the player instead of quantitatively by the game’s scorekeeper is the crux of the mod. The multiple endings I spoke of will therefore focus on the avatar and player’s perception of events, and perhaps not their real-life nature.  I wanted to tell a story that remained open to players to play the way they wanted to, and offer an appropriate response to that – with that sort of thing you really only have one “hit” to get it right first time, as any subsequent playthrough will be informed by that first one. However, I still wanted to do this because A) I feel a player can still experience a lot, and gain insight from repeated playthroughs, an B) because I wanted whatever happens to be the player’s responsibility, and not to be cushioned by a strict linear (and unobservant) narrative. What you do counts, to you. Hopefully.


Collision (working subtitle: A relationship car-crash simulator) is a first-person mod in the Source engine designed for the player to navigate the inevitable breakdown of a relationship, and gauge how they feel they can interact in the game-space, and what they would do if it was made clear that they had agency over the situation as a whole (most players won’t be aware of the paragraph above). It should be made clear as they are playing, and moreso at the beginning that their interactions are being monitored in a way unusual to a game, for instance the act of looking at something will mean more than the customary precursor to shooting something at it.


The plot is fairly loose as your agency in the various scenes determines certain elements. Basically the whole game takes place in a theoretical never-place with two distinct time-lines. The first is the “present”, in which the evolving scenario of the car crash is taking place. The other is the “past” which is presented in a non-linear series of flashbacks which chart the relationship at key points flitting between recent memories and further in the past. These also vary greatly in their presentation and ludic mechanics. As is the nature of memories their significance may not be apparent at first, and sometimes it may feel as if not much of anything is happening at all. The locations will generally be in the UK, as the couple are English.

The story revolves around a man and woman whose relationship is falling apart. When we meet them they are in a car on a winter’s night on the A1. A song is playing but they aren’t listening because a bristly conversation about the journey is taking place that wells up into an argument. Due to it snowing and it being dark there is poor visibility on the road and the car hits a pothole which puts it into a spin. It is moments into the spin that our perception of time comes to a halt, and we enter the memories of the man whose relationship/car is in the spin.

And so:

Now this was where I got with the idea a couple of months ago. I have spent time pottering on, coming up the manners in which I’ll depict the memories as they shift and decay, thinking carefully on how not to unnerve the player too much, or make them feel at risk (unless I’m trying to), and I read a couple books, watched a few movies to get research on how others approached these issues, mainly so I could try to avoid ripping anyone off too much. Problematically I watched Eternal Sunshine of the Spotless Mind – a movie I am doomed to draw from, as I saw it at the cinema in 2004 a couple of times during a significant relationship in my life. Since then that film has haunted me with conflicting emotions of wistful regret, and fond remembrance. To avoid getting too personal I will quote a Jonathan Richman lyric that sums it up, from “That Summer Feeling”:

“Do you long for her, or for the way you were? That summer feeling’s going to haunt you one day in your life.”

The film is still a beautiful triumph of a blend of visual, audio effects with beautiful music, a touching story brilliantly portrayed by all the cast. In a game that deals with memory (in not the same way, however – my memories are not being erased, but they will show the effects of their interpretation aging with time) it is inevitable that parallels will be drawn with this masterpiece. If I come anywhere near close to being mentioned favourably in that regard then I’d be honoured, though I’m putting it on the record now that I’m not trying to emulate it.

Without wanting to get emo on you all, as I’m not, it then occurred to me that my mod shouldn’t just aim to be entertaining or a decent story. My goal now is to provide a sort of tonic to that feeling. To find a sense in all the permutations of memory, and experience, and relationship – to grasp at that morsel of peace when you, even if only temporarily, put something to bed. If, in any way, I can get you in your tree talking to the Dixie Chicks, and they make you feel better, then I will have done my job.

“Memory is deceptive because it is colored by today’s events.” ~ Albert Einstein

Mike Dunbar (Follow me on Twitter at MikeDunbar and for Chronoludic updates click here)

P.S. – I should say now that Chris Green and I are both working on projects, that unbeknownst to us both bore startling similarities – though thankfully notable differences. This approach to design was independent of Chris’s adaptation of Crime and Punishment, and if that interests you but this sounds untenable and awful, then don’t let this put you off that.

 RSS Feed
This entry was posted in Collision, Devblog and tagged Collision, Dear Esther, Korsakovia, , Radiator, thechineseroom. Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.


  1. Dagda
    Posted July 12, 2010 at 2:21 am | Permalink

    Sounds very promising! Will this be your first modding attempt?

    On the conceptual side of things, it might be worth raising a point out of Psych 101- we think of our memories as recordings, but they’re actually more like reconstructions. Our minds extrapolate from scraps and associations- any new tidbit (i.e. your girlfriend making a passing reference to a police officer who pulled you over during your second date) is used to further expand on this reconstruction, even if that tidbit was actually false. There’s some interesting studies that demonstrate this.

    On the practical side of things, the mechanisms the source engine uses to pull off the scenes in Half-Life 2 where you’re teleported around repeatedly could be very useful for presenting these flashbacks. Depends how adaptable they are, of course.

  2. Posted July 12, 2010 at 9:24 am | Permalink

    Hey Dag,

    That’s a great point, referring to them as “reconstructions”. I very much had it in my mind that they’d be shifting in terms of their atmostphere, content, and therefore interpretation on repeated vistis, so that is a good way of saying what I’d not managed to yet grasp. I need to do some more reading really.

    Yeah, the teleport thing is going to be very useful. Some of the time you’ll travel (in hopefully novel ways) between memories on foot (as though you’re imagining going from one memory to the other), but there will be times where that teleport is key.

    This is my first attempt at a full scale mod, but I’ve done a bit of mapping before.

Post a Comment

Your email is never published nor shared. Required fields are marked *


You may use these HTML tags and attributes: