By John Brownlee

Editor's Note: Welcome to an experimental Kotaku feature: Armchair Designer. In it, we're going to try to put together a tactical, squad-level turn-based strategy game that takes place in a George Romero-esque zombie apocalypse. The game will largely be randomly generated, which — combined with chooseable player goals — will allow the player to extrapolate their own story from the game. There is no boss zombie; deaths are permanent. It's all about using your wits against an inexorable, flesh-hungry horde.

Let's get one thing out of the way: this isn't meant to be an implausible, multi-page wank job about that one game idea someone's got that would just be the best game ever if someone would hand them infinite technology and a hundred million dollars. In other words, this isn't an armchair design article for Paris Hilton: The Sex Tape: The Game. The design has to be as plausible as possible, because we'd actually like some designer looking for a new project to read this and walk away with it.

Go on, designers, have it! The idea's yours! Just make the game and make it good.

Every geeky twelve year old, rolling sexless in his bed at night, is convinced that the mere happenstance of the cannibalistic dead rising from their graves would be enough to galvanize him into the greatest badass the world has ever known.

After all, the Romero zombie mythology (and, by that, I mean the undead horde, slowly lumbering, insatiable and inexorable) appeals to every facet of the nerdy pre-teen's formative psychosis: a sort of egotistical solipsism fueled from loneliness and isolation that plays itself out through a make-believe orgy of gore drenched violence.

The world that's against them — parents, teachers, the farting, braying school yard alpha male — become mindless, flesh-slavering ghouls. Better yet, all these assholes can now be displaced with a minimum of moral self-examination: a machete to the brain, or a screwdriver jammed into an ear canal, or a single blast of sawed-off shotgun exploding a putrescent, hamburger-stuffed melon. Zombies are actually p-zombies: big lumbering blister bags of gore that can be satisfyingly popped without any qualms. At least within the confines of the fantasy, Jack Thompson be damned.

And just who survives when the dead walk the earth? The ability to toss a football, endowment of bulging biceps and clear complexions — none of that matters in a zombie apocalypse. What matters is being crafty and well-prepared. The medium of zombie movies encourages the feverishly dreaming geek to cast himself as the hero, because in zombie films everyone is a fucking idiot, well-deserving a fate of having their intestines scooped like ropes of dripping playdough into the mouth of a corpse. Who hasn't watched a zombie movie and thought "Pfft! I can do better than that idiot."

If you were anything like me as a kid (and, if you're reading this, you probably were, and — maybe less probably — still are), the appeal of such a fantasy is pretty obvious. It puts into tangible conflict the geek's feeling of isolation from the outside world and lets him work it out in a berzerker's orgy of cathartic physical violence. It rewards evenings spent feverishly scrawling underground zombie bunkers on graph paper, of going over in your mind necessary supplies to fight off the horde, or prioritizing the hot babes in your high school who you'll save first... for the good of the species, natch. Hey, planning this stuff out now is going to make all the difference. The zombie movie fantasy allows even the pudgiest, wimpiest, most honestly self-appraising kid to plausibly cast himself as a hero.

We're gamers. We're all sexless twelve year old nerds at heart. So why hasn't anyone ever made a game that really understands what's just so goddamn compelling about surviving in a world where there's no more room in hell?

Well, no matter. We'll design it for 'em.

Zombie Killing's Fun In Any Genre

So what's our zombie game going to be like?

We're going to start with the controversial decision to make it a strategy game. This alone is going to lose some people who just want an action game that lets them rip through brainless hordes with a chainsaw for a few hours. For many players, strategy games do not have the visceral thrill of FPSes. Worse yet, we're going to make it largely a turn-based strategy game.

Rest assured, our game's going to let you disembowel, decapitate, gelatinize, pulverize, crush, mutilate. Ideally, we're gonna have so many arterial sprays and spilling entrails Havok's going to have to come up with a new physics engine just to capture their steaming ooginess. But surviving a zombie apocalypse is all about resourcefulness, craftiness. Strategy, in short. If you just want to blow the heads off some zombies, Resident Evil 4's as good as it gets. We're not going to try to compete.

So the model for our zombie game is largely going to be the turn-based combat engine of X-Com, Fallout or Silent Storm, interspersed with the real-time base building engine of a game like Stronghold. We'll get to how both of these figure into the game in a second.

But first, what's the plot? What super zombie is the grand foozle? How do you win? Who's the protagonist?

Make Your Own Zombie Killer

Well, you're the protagonist - after all, this is your zombie apocalypse fantasy simulator. You're the hero and, being the hero, you'll need to determine your own goals: the qualifiers of your own victory in a world where society has broken down and the only real aspiration in life is 'don't be eaten.' Our game's going to be designed in such a way that beating it doesn't involve blowing up a super-zombie with a bazooka and living happily ever after. Victories will be small in the scheme of things: human, but no less noble. And you'll choose what those victories will be.

To make this clearer, when you start the game, you will create a protagonist. During the character creation process, you will specify the character's skills (and, therefore, the character's background) and the character's goals (which will, in turn, establish the game's plot and victory scenarios). For example, you could create a character named Ben, who is good at carpentry, hand-to-hand fighting and truck driving. His goal is to get to his girlfriend in Pittsburgh. Or you could create a character named John, who can pilot a helicopter and is decent with small firearms. His goal is to escape to an unpopulated Carribean island. You could design a left-wing guerilla, who wants to kill the zombified President of the United States. And so on.

We will give the player as many different victory scenarios as possible, because we want them to make this their own zombie fantasy. But we've got a logistical problem here: how can we make a game that doesn't require infinite development resources but still allows players to specify their own goals?

Extrapolating Your Own Zombie Apocalypse

We're going to deal with this problem by dynamically randomizing as much of the damn game as possible.

Think X-Com: the player largely extrapolated their own plot based upon random enemy encounters and becoming attached to their NPC squad. There were very few "plot" missions. But talk to a player about that game today and almost all of them ca excitedly tell you the plot of their very first game.

We can use the same technique: with some cleverness on our part, a good random map generator and a handful of designed plot point missions, we can make our game specifically vague enough to allow players to create their own story in their minds as they go along... all with a minimum of hard-coded scripting. Or that's the theory, anyway.

Pantheon of Zombies?

And what kind of enemies are we going to have? Zombies, obviously. But games usually have a roster of enemies, escalating in difficulty. We want to escape that and be true to the source material, largely culled from the works of George Romero and various Italian spaghetti horror directors of the 1970s: our only monsters are going to be dead humans, in various states of decay.

Still, there's some clever things we can do to differentiate our zombie m nage. We can increase a zombie's threat by their physical fitness when that zombie was still alive. It stands to reason that someone who was relatively fit when alive will be quicker, stronger and more agile even in the throes of rigor mortis. We can also decrease a zombie's threat level by how decomposed or injured they are in death. With enough zombie models and clothing styles, this'll give us a range of zombie types and difficulties to keep things interesting.

But there aren't going to be any single zombies that are much of a threat. Zombies aren't much threat on-on-one. The threat comes from when they gang up and surround you. Overall, the strategic layer isn't going to be about emptying the map, it's going to be about controlling crowds long enough to survive a mission's objectives. If we need more traditional enemies or missions, we'll make fellow humans our enemies, in the form of fellow scavengers. This will also allow us to have some interesting (and more traditional) gun battles.

Talking about it's all well and good, but it's easier to show than to tell. So let's start playing our zombie game already.

Playing Our Zombie Game: Bob Fights In A Farmhouse

We're going to create a character named Bob. Bob's going to be a farmer — good at carpentry, a mean hand with a rifle, but not particularly intelligent or charismatic. Still, he's going to be smart enough to realize that, in a zombie apocalypse, being as far away from unpredictably dying people as possible is pretty much ideal. His goal is going to be to reach an island.

The problem? Getting to the sea. What Bob will need to do as a minimum in our game to 'win' is, a) survive, b) get to a large lake or ocean, c) get intelligence on a suitable deserted island (as you see at the end of the Dawn of the Dead remake, deserted islands are pretty rare), d) stock up with enough supplies to survive indefinitely on the island, e) find a working boat, f) find someone who can drive the boat. Each of these meta-goals roughly corresponds to plot missions, either static (designed) or instigated as objects into randomly designed maps.

Bob starts off in a randomly generated farmhouse, where he's busy trying to find supplies to fight his way clear. Alternatively, he could barricade the farmhouse and wait for help. Whatever option I choose will dynamically spawn the scenario of the next mission. If I barricade the farmhouse while fighting off the zombies, I might be picked up by a redneck posse in the next mission. If I decide to fight off zombies while barricading the house, the mission will play-out in real-time, where every second not hammering shut a door or barring shutters will bring more zombies into the farmhouse for me to have to deal with.

The intent of occasional real-time missions is to keep the game panicky and frenetic, especially in maps where there aren't too many zombies. Real-time missions will always be missions in which a viable strategy (maybe the only strategy) is to barricade yourself in to a fixed position and regroup.

But Bob's not going to barricade himself; he's going to scavenge for supplies and escape. He weaves through the house, searching closets for supplies, dodging lurching zombies when he can, initiating battle only when he has to. These zombies are relatively trivial to kill one-on-one; the danger is being cornered or letting them build up into a group, something that becomes more and more likely every second I stand still. I find a shotgun and a half-box of ammo in a closet and manage to break free of the farmhouse, escaping in a truck. The mission ends.

The ZombieScape. Bob Hits His First City. A Hot Nurse.

We'll again take a cue from X-Com or Fallout and bring the player to a Geoscape-like screen: an overhead map of heartland America indicating places that Bob has been and nearby locations of interest, as well as known landmarks. This screen is where we will pick our next meta-goal. The options we're given are a) Reconnoiter for supplies/survivors, b) Start driving towards the nearest large body of water, c) Hole Up And Rest, etc. Bob chooses to reconnoiter for supplies and survivors.

At this point, a random mission is given to Bob, in a randomly generated town or city block, with randomly generated items and zombie-types (in terms of physical fitness and number) that are appropriate to my character level. Combat in this mission will be handled-with turn-based tactics. In our game, zombies can always be killed with a bullet to the brain... but headshots are hard to get. Turn-based combat will give us time to mull over our decisions. Do we go for an easier torso shot of that zombie advancing towards us and finish him off at close range with a hand-to-hand weapon, but giving him a chance to bite us? Or do we risk the headshot, potentially losing a bullet and giving the zombie more time to get close to us? Ammunition will always be a limited commodity in our game: these decisions are meant to be agonized over. The sounds of gunshots will also attract more zombies.

In this mission, Bob doesn't see any survivors immediately. Bob does, however, find a megaphone in a police car, which gives him the option of using it to call out to survivors. However, the drawback to this is that while it may bring out a recruitable NPC, it will also draw zombies or gangs to Bob's location, which he will have to fight off.

Bob uses the megaphone, and after fighting a few zombies drawn to his location, an NPC shows up. She is also randomly generated, with a random set of skills. In our game's case, she is a nurse with very limited fighting skills, but she does have medical training. This can be used to extend the life of bitten characters and patch up shot characters.

The nurse also has her own random goals which can be optionally added to Bob's pool. She tells Bob that she has a marine boyfriend upstate, and now this (if we want it to be) can become a player goal. In other words, the randomized personal goals of NPCs can be added to your character's pool of meta-goals at the ZombieScape to expand our mission options and extend gameplay.

Scavengers, I Hate 'Em

When the mission is over, Bob can choose his next mission at the ZombieScape. The nurse's personal goal of finding her boyfriend has been added to Bob's list of meta-goals. We could disregard her boyfriend if we'd like, but we decide to head upstate to find this marine boyfriend. We could use a capable fighter as an NPC.

On the way to find him, though, a gang of scavengers randomly waylays Bob, and he has to fight them off because he fails a charisma check. If Bob had enough charisma, he might have been able to trade with these scavengers; as it is, his only option is to kill them all.

Scavenger missions will mix-up the gameplay by giving the player some more traditional turn-based tactics combat fighting against a small number of capable, armed enemies. Additionally, we will have to control any zombies that may be present on the same map (although scavengers will obviously shoot at zombies as well).

Hello, Soldier! What's With All The Disembowelment?

While fighting the scavengers off, Bob get more experience and XP, as well as supplies. He continues upstate. The boyfriend, as it turns out, is dead, but his AK47 is still at his side. In other words, the mission isn't a waste of time for the player. This is an important design decision: although the game's vague plot points may be random, each mission has to move the player closer to his ultimate goal and be a reward in and of itself.

We go back to ZombieScape and decide to make our way to a large body of water. We think we have enough supplies and there's been enough dicking around. As Bob makes his way to the nearest marina, his group fights its way through the occasional random encounter of zombies/scavengers.

Bob also passes an airport, and we are given the option of reconnoitering for vehicles/pilots. We decide to do this, because a working helicopter will reduce our trip time. Luckily, it's not for naught: we find both a working helicopter and a pilot!

But there's a problem. He's infected. According to the logic of zombie movies, that means he'll turn to a zombie within three days. The nurse can treat him, potentially expanding his period of usefulness, but this makes the pilot and the helicopter a limited commodity. There is also a random chance that the pilot will turn on us in the middle of combat and potentially infect us.

This gives us an opportunity to talk about mortality in our zombie game. I've made some mention of gaining XP, but what I haven't mentioned is that, in our game, humans are naturally frail. A single bite from a zombie is enough to kill an NPC. So what I really mean by gaining experience is that a character will grow more skilled, but no less fragile; toughening up a character will be handled by dressing them in leather and other types of armor.

And what happens if our main character dies? Well, you could, of course, reload. But another option is to continue the game by taking control over one of your NPCs. At this point, you would have the option of either choosing that NPCs personal goal as your end game goal, or continuing with your original character's.

Assault on K-ZOM. Training Your Zombie Killers.

Back on the ZombieScape, we're given the option of finding gas or just flying the helicopter as far as it will go. We choose to find gas. While in the gas station, Bob turns on the radio and we're surprised that, instead of dead radio static, we hear someone from a local radio station pleading for help.

We decide to go there. We fight our way in. The gambit pays off: we find three NPCs (!) — again, this is all randomly determined. We are lucky in that we get two beefy security guards with arms training and a more useless radio technician (who can, however, fix broken radios in the game, getting important clues to side missions and additional meta-goals). Assume, again, that like the Nurse, they all have randomly generated personal goals that can be added to your goal list as optional missions.

When recruiting the more useless radio technician, we discover that she has enough intelligence that she can learn to fly a helicopter, if someone who knows the skill will teach her. NPCs in our game will be able to teach skills to other characters, as long as they have the intelligence to do so, and the player is willing to sit stationary long enough to transfer knowledge from one character to another.

We decide to have the helicopter pilot train the radio technician in the rudimentary arts of flying. We are told it will take two days for the pilot to train the radio technician as a remedial pilot. But the pilot is two days infected. Three days is the minimum it takes for a character to succumb to the zombie infection, so there is an x% chance that, during that extra day, the pilot will turn, try to kill Bob and his gang of survivors and leave the training undone. Making the choice more complicated, the longer we stay in any one occasion, the greater the chances that zombies will break into our barricaded location and we will have to fight them off in a dangerous real-time mission. It will also use an amount of Bob's dwindling supply of fuel to train the technician as a pilot.

We decide to take the chance. We hole up in the radio station and have our pilot teach the technician piloting on the roof. Miraculously, our infected pilots doesn't turn and transfers his piloting skill to the technician... good, we'll have the helicopter for more than two more days.

But during that time, zombies break into the radio station and we have to fend them off in real-time. We successfully manage to do this, but not before one of the security guards is torn apart by the zombie horde and we use up almost all of our ammunition fighting them off. We then choose to put the helicopter pilot out of his misery, since he has no more usefulness. Bob and his crew continues on in the helicopter's scant amount of fuel.

Fast Forwarding To The End Game

Just to fast forward all this a little, in the course of our game, we will fight off many more random zombie/survivor encounters. We will both gain and lose multiple NPCs.

A marine character Bob recruits takes us to a military base, where we will substantially up our arsenal, including getting a bazooka and a couple rounds for it.

Bob finds a friendly tribe of survivors who will trade with my group, but instead, we decided to just kill them all and take their stuff. Bob tames and recruits a Dogmeat like dog, to the cheers of any Fallout 2 fan.

Eventually, our helicopter runs out of fuel in a major city and we have to fight my way out. Afterwards, Bob manages reach a small marina on a small lake which has no abandoned isles, but there, we find a boat captain and spend a hard-pressed week there with him, learning to drive a boat. He's otherwise useless, so we leave him behind.

Finally, we make our way to the ocean, and Bob is finally presented with this "end game" mission. The marina has been surrounded by flaming hulks, debris, and razor wire. Around the perimeter, thousands of zombies are mulling around. On the other side of the razorwire, the marina has been taken over as a safe haven by a group of apocalyptic scavengers, who are headquartered on the only working boat.

No time to lose. Our group cuts its way through the horde of zombies, uses the bazooka on the razor wire, and fights its way onto the ship. There, we fight the 'boss' - the leader of the rival biker gang - and succeed in killing him, while the rest of our team holds the zombies at bay on the dock. With only a couple of survivors from the end fight, Bob manages to pull away from the dock in the ship, headed for his deserted island.

The End. Bob wins. His triumph is small; he does not save the world. He just saves himself.

Summary of a Zombie Apocalypse

Although the "story" of Bob's adventure looks oddly specific, most of it is randomly generated. A smart random generator coupled with a dynamically generated pool of meta-goals allows the player to extrapolate his own story with a minimum of scripted exposition. The vast majority of levels are put together from random level templates. If a plot point needs to happen during a level, we can usually just insert an object into the level that relates to that plot-point: for example, the radio in the gas station, or an NPC.

NPCs are also randomly generated, but when being generated they draw from a large pool of their own wants and goals, which the player can then choose to follow as he sees fit. NPCs may also be linked to other NPCs, like the nurse to the soldier. Although the soldier turned out to be dead when we got there, even that outcome was random: the game "rolls" to see if he survives until the player gets there. If the NPC doesn't survive, the game still rewards the player with equipment, follow-up leads or missions.

Although most of the game's content is, in fact, determined by a number generator, the player is offered a combination of interesting choices and player rewards that allows him to follow his own distinct agenda (or 'story'). Based upon these choices, the player becomes extremely attached to his character; by being allowed to explore the goals of NPCs, coupled with the fragility of team members, he will also become attached to his team.

Most games, even open-ended ones, spend a great deal of time and resources trying to get across story. But games don't have to be cinema. By encouraging players to interact, you encourage them to imagine. By giving players a small number of interesting choices throughout the game and allowing them to specify their own goals in a vibrant, dynamic world, players can put together their own plot like Magnetic Poetry, and designers can focus on making the player dynamic as vibrant as possible.

Ultimately, our zombie game lets you live out your zombie apocalypse the way you, the player, want. And sitting enraptured near the bumper of a rocking Chevrolet as the dead lumbered across the drive-in screen, isn't that what every kid stricken by the zombie zeitgeist since 1968's Night of the Living Dead has always wanted?

... Without the threat of being actually devoured alive, of course.

Editor's Note: What do you think, Kotakuites? Does this game have any appeal? Is it a plausible design? Do you have any further ideas to suggest? Is this the sort of feature you'd like to see more of? Let us know in the comments. Remember that if you don't have a commenting account, you can type your comment in anyway, and it'll be published when we approve it.