Don't Shoot Food

The Secret Shrine of Minecraft Doors


I think one of the greatest things about Minecraft (2011. Various)  is that because your ambitions in-game are pretty-much entirely your own (there is a way of beating the game but you are never instructed of its existence) they actually tend to shift. You might start by thinking you just want to survive the night, but it quickly expands into wanting to improve where you live, explore mines or lands, build contraptions, upgrade gear, etc etc. In my experience it ends up with juggling several plans at once, and I like the idea that just doing a bunch of different things at my own pace can either feel like juggling several things at once, or feel entirely laid back and cerebral.

I have become fascinated with the workings of villages. Villages are reasonably rare occurrences in the endless lands of Minecraft, appearing on occasion in the middle of deserts or grasslands. They have a seemingly random assortment of buildings, and a bunch of inhabitant villagers wandering among them. There are buildings, gravel paths between them, and a few crops being grown – perhaps potatoes or carrots. These villages look like they have been dropped from space – they connect with nothing else, and they respond to the random occurrences in the minecraft landscape. If there’s a river there, the village will run straight through it with the paths underwater, if that is what is there. It makes them oddly fascinating. My village was slightly on the side of a hill, so some of the paths and buildings had a warped quality to them sitting on edges and having raw foundations visible.

Then there’s the villagers, a peculiar set of AI-controlled people who are each able to offer you a simple bartered trade according to their profession (from librarian to butcher) and who hide in their houses at night and wonder about in the daytime making little quizzical noises that sound a little like Vic & Bob’s classic comedy creation The Ponderers. I ran into this village in a grassland that also had herds of wild horses nearby waiting to be tamed (I’ve never tamed a horse in Minecraft). I immediately dreamed of a populated ranch on the edge of this village, and took up residence in one of the houses at the bottom of the hill that seemed abandoned by the villagers.

What I naively didn’t realise, is that by moving in at the village, I was by design subjecting these villagers to almost certain death. The way that Minecraft works, nothing ever really happens unless you’re there – ie as soon as you wander far enough away from a thing, it just waits in a kind of artificial stasis until you return. This is a simple practicality of running a game world – you can’t have the whole world alive at once, so Minecraft breaks it into chunks around you. By living in this village, by just being there, I was actually subjecting the villagers to the random occurrences that actually occur when near to an active player. So every night that I was pottering away in my house, zombies were spawning and bashing at the doors of the poor villagers.

I realised this very late, when I checked the village one morning and found a bunch of zombies in some of the houses, and only two inhabitants actually left alive from the bunch that I assumed would be endlessly wandering to and from their houses in perpetuity. There may have been a zombie siege, a rare phenomena that may or may not exist – there seems to be some confusion whether or not Notch actually removed this occurrence from the game at some point. I now know that some of these zombies in the houses were in fact infected villagers – there are very obscure and elaborate ways to ‘cure’ these zombies, but I had no idea and killed them. These zombies only came because I was there allowing random events to happen, and perhaps because I didn’t skip the night-time a few times by sleeping at sunset.

Now I’m no fool – with only two villagers remaining I was acutely aware that the ecosystem of this village was under threat. With no villagers all I have is a ghost-town, and the thought of living there with regret may be overwhelmingly unappealing. The first thing I did in my panic, was create walls of stone directly around these villagers to keep them safe from attack. Then I went to the internet.

This is one of the crucial successes of Minecraft – even in a solo game you do not feel alone, because the lack of clear instruction basically forces you online to look for potential solutions to problems. The funny thing about the internet, of course, is that knowing what information you can trust is an art in itself. Look up villager breeding on youtube, and you’ll see a bunch of elaborately designed building contraptions that may or may not encourage villagers to procreate with each other. Some of these I regard with great suspicion – the videos encourage you to make buildings ‘so wide’, because ‘it helps the villagers breed’. There’s a hell of a lot of pseudo-science basically. And of course with a game that is always being updated, a solution for an old patch may not apply to a new patch.

But looking for solutions online forces you into a duality of states that is rarely so clearly obvious. On the one hand an immersive world that you care about, where you create your own future often according to random encounters and chance. On the other hand a game to be endlessly grokked – ie with complex rules at which you can pick at a thread to try and understand and solve an issue about the way it works.

One example is light. Play Minecraft and you realise that light stops monsters spawning. But how much light, and in what precise spread, is left open to the player. I have never felt moved to find out exactly how much light in what interval stops monsters spawning. A general idea has seemed to be enough to carry on. But there is a precise answer – Notch at some point created the rules at which level of light a square will be x times more likely to spawn an enemy that could be a direct threat to the player. At one level the simple concept of light and dark, and its metaphorical connection with good and evil, safety and threat. On the other level a set of rules which can be understood and adhered to for perfection, but which in practice exist to approximate something more theoretical.


0.35. This is number very important to villagers. Actually, more importantly doors. Doors are the key to villages and villagers. A villager responds not to houses, but actually doors that signify houses – the game looks at doors in the villager’s vicinity, measures that they have ‘outside’ on one side and ‘inside’ on the other, and deems that a house, ie a place of safety that a villager can exploit at night to be ‘safe’. But also doors trigger population growth – for every door in the vicinity, the game allows for 0.35 villagers to exist.

So in my panic, I actually came to realise that, despite the pseudo-science on the internet about the subject, all I needed was these two villagers to be together, and a third one will pop up at some point as long as I have enough houses (ie doors) to make something in the game system trigger. And hey presto, some love hearts started appearing between the two surviving, sexless villagers, and a little villager at some point popped up beside them.

But how to keep my tiny population alive? I proceeded to spend a panicked hour or so completely enclosing the village within a cobblestone wall. And making sure that the whole village was completely and relentlessly bright. So that most monsters cannot spawn there. This village will give its inhabitants a headache, but that’s the price of freedom.

The thing with progress, you get ambitious. I started thinking that the village currently allowed for a population of about 10, but that this wasn’t really that practical in the long-term. I think when things go wrong, there’s just a chance that a population of 10 could be wiped out in one night, as had nearly happened before. Get to 20-30 people, and your population is probably robust enough to survive the slings and arrows of bad fortune. And actually the game has a perk that it starts to spawn protective iron golems to defend the villagers once you get past a certain population level. This felt like the way to go. I started to dream of a city of bustling villagers who I could then call city-folk. Citizens.

And in my eagerness I started to grok the system. I built an entirely enclosed uninhabitable building enclosed in stone, but which was entirely full of doors hidden inside from view, each near enough to daylight to be considered as houses by Notch’s game. A secret shrine to doors. The doors that nobody could ever know about. And as expected the population soon started to explode – each door was being accepted as a house, and out popped a new villager. These villagers were being lied to, of course – there were no more houses, only hidden doors. But in my eagerness for progress I forgot the deception.

But with progress came new issues. Because this temple full of secret doors was actually the key to the population’s growth, I actually became aware that it was affecting the way my villagers lived. They were subconsciously being drawn towards the temple, without really knowing why, as if drawn there by an irresistible aura of door. Whereas the further reaches of the village were now desolate. The population was being concentrated in one area, just as London is at its densest in the very centre – nobody really wants to live in Croydon, they do because that’s where the free doors are. So my burgeoning village was actually a condensed, unsatisfying group of chattering villagers who just weren’t getting full value out of the facilities. And were obsessed with doors they couldn’t see or ever get access to. The population was bigger, the village itself seemed to become smaller in scope.

The way that villagers and villages actually work is fascinating I think.

The game is constantly measuring the world around each villager (around a 30-block radius I think). So their conception is of a village full of doors that they can potentially hide in at night. But if you think about it, they may be standing at one end of the village without being able to see the other end – in which case their perception is limited. But what Minecraft cleverly does is note the way that each villagers perception overlaps – imagine their 60×60 perception overlapping each others like a bunch of circles on a massive venn diagram. This is where the villagers ideas of a ‘village’ come from, their shared experience of how many doors they can see together.

But what happens is that the villagers will doubtless get drawn towards where the doors actually are. The only clear decision they really seem to make (other than running from danger etc) is at sunset, where they look for the nearest empty house (door) to go inside for the night. I think what happens is that regardless of the layout of a village, they will end up making a random decision each night that is nevertheless likely to eventually draw them towards where the greater amount of houses (doors) are. And the more central they become, the less likely they are as a group to be able to see the houses towards the edge of the village. They come to forget the suburbs, they like the bright lights.

This came to a head when a group of villagers started simply gathering outside the door temple where I had hidden my extra doors. They were looking for an empty house, were drawn towards the hidden doors, couldn’t get to them, and had no other empty houses to head for. So they simply stood outside the shack like eager worshippers waiting for a glimpse of their door God for some sort of empty salvation. Not only a dangerous way to act at night, but rather hopeless and depressing – a sort of malaise had descended on my village. I started to wonder if progress had been good for these people.

The only right thing to do seemed to be to tear down this shrine to doors, and destroy the hidden doors that were transfixing the villagers. At the time it felt like my only shot at redemption.

Then, tragedy. As I started to tear down the secret shrine, and the second that the hidden doors were exposed to the daylight, the villagers ran into the doors a full speed. They were simply going from door to door in there, opening and closing the door, and then moving onto the next door, and the next. It was like one of those disorientating funfair mazes full of doors. It was actually a cacophony of doors opening and closing. Every one of my 30 or so villagers just became transfixed and obsessed by these doors, each in a sort of mad stupor of simply opening and closing these doors for empty gratification.

Now the one warning I had heard about villager breeding, was that there is an inherent danger with doing it – one youtuber had mentioned that unfettered breeding can lead to a sort of villager-breeding disaster in which hundreds and thousands of villagers breed so quickly that it can destroy your game as it cannot cope with the demands of so many autonomous beings in the same world. Like crossing the streams, or meeting a version of yourself from the past, it just felt like one of those unspoken apocalypses that suddenly seemed like a potential reality as these villagers opened and shut these doors so that all I could hear was the opening and closing of doors, and villagers all stuck in a maze of doors from which they may never escape…

So I took out my axe and tried desperately to start smashing down these wooden doors to cure them. And as I hit the doors, so I started hitting the villagers too, and I didn’t know whether to stop or what else I could do! Collateral damage seemed like a unavoidable by-product of the situation, but for the greater good.

Of course what I didn’t consider was the stone golem trained to protect the villagers. He raised his giant fists and killed me with what seemed like one crushing blow. The last thing I remember was my inventory, largely consisting of wooden doors, spilling out onto the gravel path beside my broken body. I didn’t feel pain, only regret.


Most great empires eventually fail.

I feel like I’m a more benign influence on my village these days. I build little houses for the village folk, and get a little sense of satisfaction if and when a villager might visit them for the first time. I feel like I’m running the village as more of a collective these days. I’m enthralled by the way the villagers act, the way that they spread changes a tiny bit with every door that I add to every new building. I respect their opinion – if the villagers tend towards one area of the village I get the message and react accordingly. One funny thing is that I no longer have my own house as part of the village, mainly because the villagers themselves might like to use it for themselves. At night I just look for any empty house and sleep there.

I never built my ranch. I just like the way the wild horses are already, running free.

GTA – Soft in the Middle?


Every Grand Theft Auto is the same – a series of early exhilarating events, and a memorable set piece at the end, and a sprawling open-ended mess in the middle. Why are so many blockbuster games saggy in the middle?

Well it’s largely a structural thing. If you look at movies, they last 90-120 minutes on the most part. TV series episodes 42-47 minutes. Yet GTA and Assassins Creed want to last 50 hours plus. These games are basically swimming in uncharted territory in terms of mainstream entertainment media, and actually in many cases they’re flailing desperately in deep water trying to stay afloat. Luckily players, and audiences in general, remember strong beginnings, and memorable moments, but can quite easily be persuaded to forget the boring bits in the middle.

There used to be long stories everywhere. They were called novels. They could last for hundreds of pages, and basically go on for ages. But as many critics have noted and readers experienced, novels have never been as reliant on structure. When you’re holding a private conversation with somebody’s mind, and using words themselves as a kind of intimate structure with its own internal rhythm, you’re able to riff on pretty much any detail for pretty much as long as you like. Novels can literally talk about the angle of a shirt collar for pages, and no reader will flinch. If Hitchcock had done that in a movie, only the French would have praised him for it.

The trouble with a structure is that it is kind of a natural thing – there is a structure that an audience will be able to stand, and the exceptions to that will be few and far between. It’s been noted that any movie beyond two hours has to work damned hard to find an audience. It’s interesting to wonder whether that prediliction originates with the audiences, or with the commercial distributors who want to get in as many showings per day as possible. It used to be, of course, that a B-picture would accompany the main film like a support act at a concert or boxing bout – those who wanted full value braved the whole evening, while the cool kids just arrived for the main event. Maybe it’s a decision that comes out of convention. It’s a bit like the length of Mario’s jump – he jumped so far and it just became the de facto standard for videogame jumping – disobey it in your game at your peril.

In a way this must be the issue facing the makers of GTAV – the convention set by their own precedent is that their game must be vaster, more varied, and have a story arc that will last somewhere around 50 hours. Of course some of that is travelling, mucking about in the city, doing the sub-tasks that GTA excels in. But they know that generally their story has to pack in lots and lots of gameplay within an overarcing story. It’s one hell of an ask. The problem is that story structure, as we know it, doesn’t often allow for 50 hours sprawling gameplay odysseys.

Hollywood’s three-act structure and the videogame.

Presumably each and every writer working in videogames knows the Hollywood paradigm, the 3-act structure that is said to be the building block of every mainstream movie. A first act setting up the conflict and throwing the hero into the unknown, the second act where the hero battles to get to the final confrontation in act three. The hero’s journey. I think most mainstream games show the signs of being prepped according to the three-act structure, and then stretched past breaking point to fill a game with too much content to correspond to the structure.

I think the three-act structure in GTAV is based around Michael. His story and ours begins with a botched heist, after which Michael has entered witness protection and effectively retired as a gangster. But a series of mid-life travails and an encounter with a young street hustler called Franklin, throws him back into the dangerous life of crime that he has been trying to avoid. And I think the first act often works in games – it transplants totally successfully. The way you set up a character in peril, thrown into an adventure, that is the purest stuff of the mainstream videogame. Games are about action, just like the hero’s journey.

If you look at Space Invaders, you can imply a three-act structure with a clear and strong first act if you care to look for it. The Earth is being attacked, a sole defensive turret/ship has lined up to defend it, and the game’s middle act is then the sense of impending dread as the enemies get closer and closer to the Earth’s defences. Clear the enemies, and in a cruel twist more spring up in their place. The third act, I suppose is the heroic loss of life. After which the player is supposed to put in a new coin to carry on the endless fight.

The problem in GTA is the middle. We’ve set up Michael, we’ve put him into this life of crime, now what? We kind of sort of kinda know what the end is – a sort of final confrontation between Michael and the forces of crime from his past that he has spent much of his life trying to avoid. We kind of know where it’s going and we play along. Or we don’t – full disclosure, I’m stuck somewhere in the middle of GTA5 searching for that ending.

But what does the game do in the meantime? Second acts are not easy, because really what you’re attempting to do is to recreate the first act a number of times, and tell a series of mini-stories that tie into the main story and push it forward. So the hero faces a new hurdle, struggles to overcome it, and as a result is thrust slightly closer to the conclusion of the journey by being a little wiser, a little stronger, or with a new ally or a weakened enemy. Like a series of end-level bosses basically – classic story structure descriptions often use the metaphor of a series of monsters in a series of caves.

The trick is to tie in each boss battle to the greater story. So that it seems like it changes things, rather than just being a series of boss battles. And actually in much of gaming history, that has unfortunately been the problem – a series of boss battles is literally a series of boss battles. We come to accept that the boss battles are part of the game, rather than a means to an end in terms of a story. Often because the story is threadbare and not connected to the gameplay, so we’re not encouraged to ever interpret the game emotionally at all. This is the reason for the massive disconnect between one player and the next – while one player might interpret defeating a boss emotionally empathising with the lead character, the next player is simply checking their score and ignoring the cut-scene. This is the monster we’ve all created with videogames, and it will always be hard to break free from.

Seeing the King and checkmate – knowing where you’re going.

Chess has no problem with the middle of the story – it is an essential part of the process, and actually generally the most skillful part of the process. You know what you’re aiming to do (checkmate) and you can even see the end-boss (the king) and you realise that every move and every situation directly takes you further or closer to the goal. The excitement often comes from situations where you turn adversity into success, or even fake adversity to bait the enemy into a mistake that leads to success – this is drama created by gameplay.

In GTAV I think this is a crucial error in the game – you kind of know what the end of the story is, but you just have no idea what form it will take. I think this is intentional, because the game uses psychopathic Trevor as a kind of angel of death who signifies Michael’s fatal end, and in a way we’re just waiting with Michael for the powder keg to explode. And it could be that GTA realises that it simply cannot sustain the tension of being able to see the King on the board throughout – over a 50-hour game I suppose it might just get frustrating.

Assassins Creed III has exactly this problem I think. You’re introduced to the hero and to the villain, and it just gets more and more ludicrous and frustrating the ways that the conclusion of their antagonism is avoided. The game is even daring enough to have the hero and villain co-operate in missions in the game as part of their interlinked role in the history of the US. I actually really like AC3 and I think it’s a daring, original game, but I think this weakness is a major reason it goes under-appreciated.

But the alternative is just as problematic – consider Mexico. The weakest part of the excellent Red Dead Redemption is universally pinpointed as the Mexico section of the game. The problem not necessarily being that Mexico is unpleasant or uninteresting, in fact many of the most memorable sections of the game are in Mexico, but the problem is that there’s just not enough connection to the main storyline. The political squabbles in Mexico feel like an aside to the main story, indeed this is mainly how they are presented, and the whole thing ends up kind of pointless for that reason. We can’t see the king and we can’t see the checkmate, and we can’t understand how anything we’re doing gets us closer to the end of our journey.

Another approach – make sure that every encounter clearly brings one closer to the conclusion of the game. This is what the original game Assassins Creed does – for all its complexities the story is very very clearly signified as a series of assassinations. And the main complaint about Assassins Creed – repetition. Too many players saw this clear structure and found it dull – the road to the conclusion was too obvious, and the path too straight, so many players were dissatisfied. They could see the king and see the route to checkmate, so the fun of working it out was removed. In many ways I think the first Assassins Creed is the richest, most original and thoughtful mainstream game ever made (okay that’s a bold assertion – I do love that game). But I note how the actual content of the story is irrelevant if what you do as a hero is not satisfying. I’ve read many reviews of the game and none mention the daring themes or artistic content of AC, but all mention the simplistic mission structure, its fatal flaw.

GTA and Breaking Bad – the game as a series.

I think somewhere around the time of GTAIV Rockstar started to look at television series for its structural inspiration. Drama series, by their nature, contain an enclosed story in each episode which satisfies an audience for that hour. There may be greater story arcs told over seasons, but the main focus is on a workable story that will be contained in that episode. I suspect this is how GTA missions are now prepared, each as their own mini-episode made to satisfy an audience for half an hour plus. And there’s something efficient about that, as an aging demographic of gamers tend to experience bigger games in smaller chunks these days. And by considering each story mission an enclosed story GTA has managed to add variety and scale to these chunks of the story – many of the missions in earlier GTAs were actually just dead boring, and there was a huge variety in terms of excitement, length and quality. In a modern GTA you enter a story mission and you know it will be half an hour plus, and will involve an action set-piece.

But this is a solution to one problem, but not to the wider malaise over story structure. We may complete a satisfying mission that contains great thrills, and makes sense within itself (ie it has a three-act structure and doesn’t overstay its welcome etc). But does it drive the greater story forward? I remember one of my favourite missions in GTAV is the heist of a cargo plane – Michael shoots its engine with a high-powered rifle in a van at the observatory, and Trevor then trails the failing plane cross-country on a motorbike to retrieve the contents – the exhilaration of jumping a train on a motorbike while a jet falls out of the sky in the distance is incredible. But what happens in the story because of it? Well, next to nothing. It’s just a cool series of set-pieces. It would be cut out of a movie, or tweaked to make sense in terms of the hero’s journey. It could exist in the episode of a drama series, but only in terms of the enclosed story of that episode. And the trouble with GTA is that we don’t care about the episodes I think. We could skip 10 or 20 and still be satisfied, maybe moreso.

There is a peculiar aspect of GTAV that I think is a missed opportunity. In a drama series you’ll be introduced to a character, and the episode might centre around their issues, and then there’s a sort of conclusion at the end of it. The murderer in Columbo gets their comeuppance. The poor farmers get protected from the evil cartel by the A Team. But in GTA you meet a schmorgasbord of characters, but I’ve noticed how rarely any of their stories come to any definite conclusion. Even when they do it can be undramatic and almost incidental – Trevor murders a couple of characters to cut off the strand of a story, but it’s presented as an incidental joke rather than the conclusion of anything. If GTA wants to go forward I think it needs to consider having its missions come to a proper conclusion – have them complete a character’s arc. Too often the missions end with one of the three characters standing in a place having done a thing – nearly always anticlimactic.

Drama series are of course undergoing a renaissance directly because they are becoming less like drama series and more like movies, with more epic scale and higher production values. So it’s ironic to me that GTA has recently become less like a movie and more like a drama series.

Being stuck with characters that won’t change

This is an ongoing problem for videogames. How long can you stand being with one person? And GTAV’s decision to actually have three playable characters (four if you include the dog) is in many ways a response to that – get bored of one and you can switch to another. That;s the theory anyway. I don’t think it works because the characters don’t change, and I wonder why that is – is it simply to retain consistency over such a huge storyline?

While Niko Bellic in GTAIV was a terrific memorable character, at the same time there was a slight disconnect between what he thought as a character and what you did as a player. This problem has become one of the standard examples of ‘narrative dissonance’ which is a key buzz-phrase in modern story design. I think aspects of the characters in GTAV are an attempt to ‘solve’ that. These are archetypes, and they are pretty consistent with their aims and worldview, so that you know exactly where you stand with them, and hence no disconnect. When I play as Trevor I run every red light and act as psychotic as I want, as I know that is consistent with his depravity. As Michael or Franklin I am much more unlikely to carjack someone or commit sporadic violence, because I just think they’d be less likely to do that.

But by solving one problem I think they’ve created another – if you have broadly drawn characters who simply don’t change and don’t really ever consider changing, then having them ‘develop’ through a 40-hour middle act is all but an impossibility.

I think this is where other game characters have had the edge. With Gordon Freeman in first-person (in Half-Life and sequels) the character is a blank slate who is essentially you, the player, so what you think is what they think – there is no disconnect, and no problem. With female or male Shepard in the Mass Effect series, you can mould your avatar to a point of view and a playstyle that you decide and the game allows for that. Again the emphasis is on the playable character as more of a blank slate. The other successful approach I think is to be enigmatic – to have a strong character but to also have an ambivalence in which playstyle in many ways defines character, John Marston in Red Dead being the perfect example. You can’t ‘play’ Trevor as anything but a psychopath, and there’s just something fundamentally uninteresting and limiting about that, however fun or memorable he may be in the short term.

Not only are you stuck with characters that don’t change in GTAV, you are stuck with characters who don’t change who play out the same fractured relationships hour after hour after hour. While the prospect of missions carried out by swappable characters is a very good idea, in practice it means that so many missions have a similarity to them in terms of dynamic. How many times in the game do two of the characters get in a car and drive across the city chatting about something, or even if one is in a helicopter communicating with the other in a car the same relationships play out… again and again. In GTAIV Niko might encounter and interact with a different character in each mission, revealing aspects of his character and relationships as he goes.

A key part of a middle act is variety, of taking a story and the characters in new and different directions. The characters in GTAV really don’t allow this to happen.

Middle acts going forward – GTAVI and beyond

It’s all very well criticising GTAV. It’s an easy target, but it’s also a phenomenal success with gamers. Or should I say it’s a phenomenal success with its target audience. Like Call of Duty it has a devoted fanbase, but also a majority of gamers and other game-makers can see its faults. I feel like I fall somewhere inbetween, as I have a real love/hate relationship with mainstream games. But what can GTA and mainstream games do to avoid this middle-act sag? It’s not easy – I’ve spent this piece criticising games but many many movies fall foul of the exact same problems.

A problem that I have no easy answer for is the scale of these blockbuster games. Almost by design these games must be chopped up into manageable sections to make the practical creation of them possible, which makes subtle ramping of tension or character development a major headache. I can’t imagine a GTA or an Assassins Creed without missions that loosely fit together to create an over-arcing story. That’s the way it has to be, and for that reason the broad sweep will almost always be problematic. Because presumably missions get cut or drastically change, and if each were as dramatically crucial to the story as I’d like, games would risk failing entirely. Assassins Creed 2 is pretty clearly a game where the late missions have got into some sort of mess late in production, and the game comes close to simply falling apart with massive holes in the narrative – luckily the rest of the game is so satisfying it gets away with it.

I do think GTA could look at Assassins Creed to find a very effective idea, and that is what I would call ‘story acts’. Assassins Creed games have an open world, but they also have epic stories that are clearly split into acts, literally jumping years into the future (or past) to be clearly distinct. Actually Red Dead has a few similar jumps. I think it works for the player – we know that what we were playing has finished, and we expect a new emphasis or perhaps a calming of tension leading up to a new conclusion to a new section of the story. GTAV pretty clearly lacks this in my opinion – I struggle to actually remember any clear breaks in the story there. A plane trip to somewhere else in the country is the only clearly noticable break in the story, and has a clear plot point to go with it – I think future GTAs need more of these.

I also don’t understand why GTAV doesn’t use the clear (if crude) decision making that occured in GTAIV. One of the most memorable moments in any game that I’ve ever played occured when deciding whether to kill or spare a character in GTAIV – I had to literally stop the game and think about it overnight. Now I can see that these choices are the equivalent of a choose your own adventure choice, as basic as that, but I totally think GTAV lacks this. The reason I mention it is that decisions like these can get a player to invest in a choice in a middle act that might come to feel like something. In GTAIV you actually get to choose one of two key allies to kill – nobody can say that isn’t a dramatic moment with consequences, that feels like it thrusts you down a new path in a middle act. These are the tools that games have to deliver stories – it feels to me like GTA and Assassins Creed both choose to ignore them, presumably because of the complexities and sheer man hours involved in accommodating different outcomes. But actually this doesn’t have to be the case – a choice may seem dramatic and lead nowhere in reality, but the consequence of just taking it can seem to thrust a story and a character forward. GTAV too often feels like the story isn’t going anywhere but round in circles.

One other practical suggestion is to make the stories shorter, not in hours but in terms of dramatic beats. This probably sounds like disaster for players, but I think it would lead to a better experience. Imagine a GTA with perhaps more gaps between missions, but a load more side events. I really think this is the way to go. Perhaps the side missions could have more of an arc to them – they could be stories in their own right, perhaps multi-missions with proper drama in them, designed for dramatic weight rather than light relief. I look at Skyrim and Fallout 3 – both amazing games with short-ish main stories but incredibly strong side-missions. They are simply better games I think – their structure works so much better.

And maybe that’s another answer – do away with the structures, and set the player free to create their own. A player left to their own devices in Minecraft seems to have that inspirational knack of creating their own stories more memorable than anything a team of designers could come up with. But then maybe that’s what GTA Online is for, I haven’t touched it yet. I suspect that points to the future of GTA more than anything Trevor says or does in a cut-scene.

Alan Turing and the Imitation Game


In 1948 the legendary mathematician and computer scientist Alan Turing was busy writing a workable chess program for computers before a machine existed that could possibly run it. Turing had been instrumental in helping the British government secretly break key the ‘Enigma’ German codes for communications during the second world war – he was a brilliant mathematician and now widely regarded as a key figure in the 20th century and in the history of computing.

In 1951 Turing joined the crowds at the festival of Britain and saw the NIMROD machine – a digital computer about the size of a garden shed designed specifically to simply play one game, Nim. It’s a traditional game where there are four piles of matches and you take turns to grab a handful of matches and the aim is to not be left picking up the last match. NIMROD used lights instead of matches, and members of the public were invited to challenge it to a game. In the accompanying booklet it was explained, and in some ways, sought to be excused…

It may appear that, in trying to make machines play games, we are wasting our time. This is not true as the theory of games is extremely complex and a machine that can play a complex game can also be programmed to carry out very complex practical problems.

Alan Turing proposed his own game in 1950. It was called The Imitation Game and it was designed to test the hypothesis of whether or not machines could be made to think. It’s popularly known as the ‘Turing Test’. It involved three rooms – one with a human, another with a computer, and in the third another human as interrogator. The interrogator’s task was simply to work out through questioning which of the other two was the human and which the computer. If the computer could convince the interrogator, machines could truly think and the game would be beaten.

Turing died in 1954, and his game was still only at the theoretical stage. But he predicted that by the year 2000 machines would be considered to engage in what would be considered as thought, and would be getting close to beating his game. But more than 50 years later we’re still waiting for the Turing test to be passed – it’s now regarded more as a philosophical concept than a serious challenge. But that doesn’t stop anyone playing the game.

The early history of videogames is almost entirely undertaken by scientists mucking about with machines imagined for more serious uses. AS Douglas, a professor of computer science, has his noughts and crosses game OXO up and running at Cambridge University in 1952. In 1958 physicist Willy Higinbotham had a tennis game >Tennis For Two running on an oscilloscope in his lab in Brookhaven to impress visitors. By the mid-60s, everywhere that there was a computer, there was usually a small bunch of scientists cum hackers writing and playing games to play on it in their free time.

Some of these games passed as research. By 1966, a computer program called Eliza, developed by Joseph Weizenbaum at MIT,  provided its own brand of imitation game. Posing as an artificial psychologist, Eliza prompts you “How do you do. Please state your problem.” It replies with catch-all phrases like “Can you elaborate on that?” “Tell me more”. And by scanning for key words in your input like “yes” or “no”, “mother” or “father”, the program is able to offer convincing responses (“You seem to be quite negative”, “Tell me more about you family”). Although never presented as any more than just a clever trick, Eliza was often able to fool its users. And it was and is fun.

Perhaps the more practical sibling of the imitation game was the massive interest in creating a chess-playing computer that could hope to match the best human mind. It’s not hard to see why this sort of project could inspire minds – this wasn’t just fun, it could be science! 1962’s Kotoc-McCarthy, an MIT-produced chess program could play chess convincingly to an amateur level; the only drawback being that it took up to twenty minutes to make each move, and required an IBM costing $2.9 million to run it.

As for Turing’s original game from 1950, The Imitation Game is still played annually as players compete for the Loebner prize, with competitors aiming for the $100,000 prize offered for the first computer to successfully pass Turing’s test and ‘become human’. It’s hall of fame includes generations of conversational programmes, these days called ‘chatterbots’ – Albert One, ALICE, Ella, Jabberwok, Ultra Hal. The game has many critics – some scientists dismiss it as trivial and unhelpful to the development of artificial intelligence. Those crusty old scientists! It’s a game that illuminates and encourages discussion, and popularises a key scientific discussion, probably better than any scientist could.

You might also enjoy a similar but less enjoyable version of the imitation game when you try to ring your bank. Press the hash key to hear this message again, and have a nice day.

Devil World


I came across Devil World (Nes, 1984) while looking at the early releases for the Famicom or NES, particularly Nintendo’s roster of games in 1984. The Famicom had been launched in Japan, and now Nintendo had its eyes firmly on an American release for its console. But this was also crucially a world of gaming that existed before the classic releases Super Mario Bros (1985) and The Legend Of Zelda (1986).

Their 1984 roster basically seems to be a port of Galaxian that was never released in the US, a very servicable clone of Pole Position called F1 Race, a HudsonSoft single-screen platformer inspired by Donkey Kong called Nuts ‘n Milk, and Miyamoto’s pre-Mario PacMan clone Devil World. These are basically all arcade-style games, influenced hugely by coin-op design. Having a game that lasts longer than 180 seconds on any of them is a major challenge.

This was in huge contrast with what was going on in Western development, particularly on home computers. RPGs, tactical strategy games, arcade adventures, and simply more elaborate games were being created in this era in the West, and it sounds weird now to admit that Nintendo seem slow to have caught on. The likes of Pitfall on the VCS, Jet Set Willy on the Sinclair Spectrum, and wildly ambitious open-world games like Elite and Ultima. While Nintendo was cloning Pole Position as a simple arcade racer, computers were already running fully featured flight simulators.

The way I like to imagine it is that Miyamoto, perhaps on his first visit to the West, goes to some trade show and his jaw drops at the ideas available in Western titles. He travels back to Japan inspired, and reinvents these ideas in his own style, with the help of the artisans at Nintendo who have schooled themselves on tight arcade rules and execution. Hence Super Mario Bros and The Legend Of Zelda. It needed that confluence of factors coming together to create that revolution in mainstream gaming.

Devil World, however, belongs very much to that pre-Mario thought pattern and that’s why I find it interesting in comparison. It’s strictly an arcade game. You’re in a maze, there are dots to eat and enemies to avoid, so it passes that first test of arcade games in that it’s clear exactly what you must do. Obviously its roots are in PacMan, the solid gold hit from 1981. But actually even arcade games had moved on significantly from that era, and whether Devil World could have possibly competed in arcades with the likes of Marble Madness, Paperboy or Pac-Land is very dubious indeed. Already it was to some extent a game out of time. But the Atari VCS had operated in a very similar manner, porting arcade games from years prior and giving the home user a chance to enjoy some of that thrill well after the event. This was Nintendo’s version of PacMan, and it had no more ambition than that.

But the choice of imagery in the game is something quite else, and the game was rejected for release by Nintendo of America for its religious iconography. It’s a game about the devil with flashing bibles and crosses used as weapons against Satan. Of course a modern company would simply change the sprites and launch the game with a non-Satanic name. Maybe it shows how low down the priority list Devil World was by the time the NES launched in the US in 1985 – it was a game out of time almost as soon as it was made.

But the game’s a good deal more surreal than it sounds – it may be a fight against the devil with crosses and bibles as the power-ups, but there are also fried egg monsters, and you play as a green or red dragon! The meaning of all of this is not clear, though I wonder if there’s an Eastern connection between dragons and spirituality that might make the game play less absurdly in Japan. Either way, it’s a bizarre choice of motifs, that suggest to me that Miyamoto was basically naive of the power of images, themes and motifs at this stage.

1984_devilworld_1Regardless, there are a bunch of fun ideas working through the game that lift it above the multiple PacMan clones of that era. The devil looks on from the top of the screen, dictating through dance to his minions which way the map must scroll – he looks part cute dragon, part disco-dancer. Areas of the map look like crosses, the game plays with this motif. You pick up crosses like a mini-exorcist to fight the enemies and pick up the dots. One could imagine this gamespace as a level of hell with never-ending catacombs leading to a central altar. But the ritual dances around the screen are undertaken by ice creams.

What the game does have is phases of play that I think work nicely. You go from dodging enemies to chasing them with crosses, all the meantime making sure you’re aware of the movement of the screen in order to avoid getting trapped. I do think the first level can be effectively dealt with through simply dodging enemies and staying pretty central – at some point you’ll probably be in contact with the dots you need. And with no timer the danger doesn’t escalate – a crucial issue which surely would have to be tackled for an arcade version?

What remains is a game that was limited in terms of its arcade roots, that probably couldn’t compete in arcades, and felt out of place on home consoles alongside the more elaborate long-form games that followed it. And with religious-themed sprites naively employed, that immediately limited its commercial potential. And that’s what makes it a fascinating relic of its time.