FTL Specifically

When I started writing my previous post, I originally meant to 1) build a vocabulary for discussing roguelikes and their derivatives, and 2) use that vocabulary to consider the strengths and weaknesses of FTL: Faster Than Light, a game I've been enjoying tremendously over the last month, but about which I do have a few reservations. I only really had time to accomplish the first goal, however, so today let's do the second.

FTL has an airtight premise. You are the commander of a federation starship. The depraved rebels will destroy the Federation unless you get there first, give your superiors the intelligence you've gathered, and successfully defend the Federation from the rebel flagship. You won't be strong enough to defeat the rebel flagship unless you upgrade your ship extensively during your journey to the Federation, but the rebels are in hot pursuit -- so you have limited time to spend in each star system. You want to make each leg of the journey last as long as possible, so that you can gather scrap, weapons, ammo, and systems for your ship. But you can't spend so long that the rebel fleet catches up, because while it's possible to escape the fleet, it's also costly.

In practice, FTL is a space exploration game where all roads lead to combat. Its rhythm is simple but compelling: when you arrive in a new system, you see a map of all the nodes available therein. For reasons that are presumably explained by the game's fiction, you can only make faster-than-light jumps to existing nodes within a system. Depending on your ship's upgrades, you generally have a pretty limited idea of what's waiting for you at any particular node, but it usually breaks down into three categories: shops, randomized events, and ship-to-ship combat. We'll tackle those in reverse order.

Roguelikes Generally

So lately I've been into roguelikes. Half of you will know exactly what I mean by this, and half will have no idea. For the latter: Rogue was (is) a very difficult computer game wherein players explored a dungeon. It features randomly generated level layouts, monsters, and gear. It was influenced by Dungeons & Dragons the way that my apartment walls were influenced by off-white paint. (Heavily.) Your character needs to find good equipment, needs to fight monsters to grow stronger, and needs to eat food in order to go on living. It's common for players to starve. The game features permadeath; that is, when you die, your character is dead for good. It was first developed in the '80s, but people are still playing and improving it today.

It's not important for our purposes now, but it's noteworthy that Rogue had ASCII graphics -- your character is represented by an @ sign, and everything else in the world is represented by the symbols on your keyboard. It looks like this:

So, a Roguelike is exactly what you'd think: it's a game that plays like Rogue. My sense is that most people believe that randomness and permadeath define the genre. As a rule, roguelikes are also almost always extremely difficult. They don't have to be about dungeon crawling or monsters, but they usually are. Roguelikes are especially interesting right now because they're providing the foundations, in terms of both philosophy and practical design, for a surprising percentage of the most interesting, most critically acclaimed games being made today.

My interest in the genre began with this thread at the Talking Time forums, wherein an experienced player gave a guided tour of Angband, one of the more popular roguelikes. Soon I installed the version that thread's author was playing and proceeded to squander a huge amount of my life. My obsession eventually led me to write this monstrosity.

I was easy prey for Angband. I was weak. The year I discovered roguelikes was one of the worst, most stressful times in my life to that point. I was studying for my MFA at New Mexico State University in Las Cruces, New Mexico. There were and are many wonderful people there, but there were also those who were very hostile to me as a person and as a writer. (Maybe they are wonderful too. It was hard for me to get past all the hostility.) I well understood hostility to my person (it is (I am) arguably loathsome) but nobody had ever really hated my writing before; I had spent my undergraduate career as a darling of the Butler University English department. The new situation -- one where I was being told that I was not a good writer, that I would never make it at this rate, that I might not even graduate from the program (a threat made both laughable and extremely hurtful by the profound rarity of literate humans failing to graduate from any MFA program) -- was almost more than I could bear.

Not because it was so terrible, objectively speaking, but because my whole identity and most of my self-worth were and are invested in the idea of myself as a writer.

Angband gave me a way to manage my stress. I played it, at the height of my interest, for several hours a day. Roguelikes are usually turn-based, (meaning that the monsters politely wait to move until you do), and they make very few demands on the systems that run them, so it was extremely simple for me to keep a game running in the background as I worked, playing a floor or two whenever I felt stuck in my thesis. (In fact, I'm playing another roguelike this way as I write this post.)

I felt bad about myself for how compulsively I played. Angband wasn't going to make me a better writer, and I would never beat Angband, which made it seem that I might play it forever. I stayed up very late some nights. I used to sleep very little. (And sleep still feels like wasted time. I should be working.)

I never got close to beating the dungeon. I still have no idea which letter represents the final boss. (I like the fact that I don't know.) There were a few times I made it very deep, I believe always playing as a dwarven paladin or cleric. They have some pretty useful spells. Eventually, I lost a character that I had invested so much time in building, I lost the will to play. I imagine this is how most players give up on most roguelikes.

Now I am going to make a numbered list of reasons that roguelikes are super-compelling. Then, when I get done making that numbered list, I will come back and put the last number into the following sentence, which will make this article appear more deliberately planned than it really was. There are seven reasons that roguelikes are super-compelling.

1. The fact that your character vanishes forever when he dies means that literally every accomplishment ratchets up the tension. Every monster you kill, every potion you find, every weapon you acquire, every floor you complete makes it that much more heartbreaking to lose your character. And you will, unless you are very good and very lucky, definitely lose that character. You're essentially fattening your  halfling thief for the slaughter. The feeling of striving against impossible odds, of delaying the inevitable, is extremely addictive. It's unbearable in the best way.

2. There's also a really interesting arc to your relationship with your characters, wherein you play carelessly at the beginning (might as well, this dude is nothing to you, he's gonna get eaten in five minutes anyway) until your little guy gets lucky, at which point you begin to explore the possibility of caring about him. Then he dies, too. Eventually, though, you start to really understand the game's basic mechanics and some of its finer points, and then your guy finds some really incredible loot, and suddenly you're ten floors deeper than you've ever been before. You fall in love with this character. You start to actually remember his name. You let yourself think that this might be the one -- the hero who will best the dungeon. And maybe he is! But he never has been for me.

3. Different roguelikes have different levels of randomness in their levels, which are generally described as "random" but are better described as procedurally generated. That is to say that each game has certain rules by which it builds its levels. In the original Rogue, most floors have nine rooms arranged in something like a 3x3 grid. But the connections between the rooms are placed somewhat randomly, and sometimes there are dead ends where you expect to find a room, among other variations. In Angband, the layouts can be much more complex. There are also several types of level with their own rules for construction -- messy levels, wide-open levels, spiraling levels, etc.

In games of other genres, you're interacting with the game's designers through the mediation of its systems and environments. Most players aren't consciously aware of this most of the time in a well-designed game, but success often requires correctly discerning the intent and desire of the game's developers. On seeing a platform in a Mario game, you might think, "The designer wants me to go up there." (If the game's world holds together well, what you'll think instead is, "I bet I can get up there." If it's badly designed, you'll probably think, "What does the designer want from me? What the hell am I supposed to do?!") In a roguelike, because the levels (and to a lesser extent, the monsters and gear) are procedurally generated, questions of the designer's intent rarely enter the equation. The developers could have never predicted that you would be in this precise situation, so it's up to you to make the best of your situation. Sometimes you're shit out of luck, but -- in the best roguelikes -- you can often solve a seemingly impossible problem with ingenuity and quick thinking. These moments are all the more thrilling because the designer didn't engineer the solution for you. You found it on your own.

4. At the same time, the designer's presence is discernible in the game. Dungeon Crawl Stone Soup, which I've been playing recently, has a number of level layouts like the one above, which very likely utilize a mix of preexisting design and randomly generated content. The scenario above is clearly designed in advance -- the developers have stranded a menagerie of monsters, some of them extremely overpowered, behind a glass wall. This gives the player the thrill of seeing these monsters without the dread of having to fight them. The random elements are the placement of this segment of the level (I've seen this variations on this structure appear in different forms several times in my playing) and the selection of monsters, which always features a mixture of levels of power, but not always the same composition.

Dungeon Crawl Stone Soup (hereafter DCSS) is a strong example of how roguelikes allow players to experience a play space that is neither wholly authored nor wholly random -- a tonally complicated mixture of automated and imagined content. You are dreaming with the designer, who dreams with the computer, which dreams through the game. This often feels like perhaps the purest expression of computer games as games. I'm especially fond of DCSS's spirals, which might well be designed by hand or algorithm. The Binding of Isaac, an action game with a strong roguelike foundation, even seems to build series of levels according to themes: I've played games of Isaac in which, by some combination of algorithm and design, the levels seemed to have a fire theme, a worm theme, a shit theme, or a flight theme. (More about this in another post.)

5. Inherent to the structure of procedurally generated levels and permadeath is the assumption that the player will see many variations on the game's content, especially in the early levels. (Later levels in roguelikes, especially Angband, tend to feature more challenging monsters with more specific powers, which has the effect of making the content feel more authored; the designers make it overwhelmingly likely that you will face certain challenges and dangers as you descend.) This frees the designers in several key ways.

For one thing, it allows them to allow you to experiment and put yourself in situations that are less than optimal without your feeling that your time is being wasted. (At least in one sense; there's another very real sense in which all time with a roguelike feels wasted. Thus my guilt over my former Angband binging.)  I love the Uncharted series and many of its cinematic fellows, which attempt to gratify and overwhelm the senses at every turn, but these strategies have severely diminishing returns. And I find them exhausting. A game that attempts to engineer my constant perfect pleasure is a game that not only costs a fortune to develop, but actively discourages me from playing from a truly playful perspective. There is tremendous pressure, in modern big-budget titles, to always be having the most fun possible. Usually, as a result, they aren't really fun so much as grating, stressful experiences that leave me grouchy and tired. Games like DCSS and Angband don't always have to be fun -- which is, for the player, very freeing.

6. It also allows them to do something that I've come to see as increasingly rare in modern games: roguelikes rigorously follow their own rules, regardless of the consequences. You won't find yourself bumping up against invisible walls or artificial barriers to progress just because the designer doesn't want you to go somewhere yet. The designer doesn't care how you play a roguelike, most of the time: if a weird solution works for you, especially one the designer didn't see coming, then that's probably a good sign for the game. You're supposed to strive to break the system, to exploit weaknesses in monsters, to go wherever you feel brave enough to explore.

The Binding of Isaac is rare among roguelikes in that it features explicit boss fights. (The boss characters are randomly selected within certain parameters, and feature many variations, but they are indeed bosses.) But if you find a one-shot insta-kill item in Isaac, you can use it to instantly kill anything you want, including a boss. Big-budget games can't afford to let you do that; they were designed so that you would see certain cinematics around and probably throughout the fight, so that you would hear certain dialogue, so that you would beat them or lose to them in a certain way. The Binding of Isaac couldn't possibly care less how you win. Surviving the game is so difficult, and you're going to play through it so many times, that the developers are more than happy to let you approach each problem any way you want. Again, this means that roguelikes can play like actual games, rather than movies with interactive segments.

7. The genre's controlled chaos frequently leads to surprising and extremely specific play situations. When fans talk about roguelikes, they usually end up sharing stories of weird deaths and skin-of-the-teeth triumphs. They talk about the time they were trapped in a mob of orcs, realized they were in over their heads, and started frantically quaffing unidentified potions -- first paralyzing themselves, then healing some damage, then poisoning themselves, then finally drinking the potion of might that bought them just enough time to cut a path to the exit, only to be eaten to death by a dire molerat. I recently read a forum post about a game of DCSS wherein a character was confronted by an "out-of-depth" (i.e., overpowered for the level of the dungeon in which it appeared) monster. He used a wand of polymorph to turn it into another randomly selected monster. The one he ended up with could have been much better, or much worse -- it was about the same, so he used the wand again, turning it into a basic human. Then he killed the human. Of course, for every story about this strategy working out, there are dozens where it ends in tears. (DCSS is especially well-designed to encourage this sort of strategic risk-taking, about which more in a future post.)

Games like Nethack and Dwarf Fortress are famously good at generating this sort of craziness. This is partly possible because these games don't mind screwing you over for a laugh, but also because of their low graphical fidelity -- it's much, much faster to write the sentence "you become a vampire" than it is to animate the same event.

Okay, so that's seven items in the numbered list. Excuse me for a second while I scroll up and put that number in the paragraph above.

Nice.

I'm sure there's at least one major omission in this post, but it gives us some shared terms to use in our future discussions of roguelikes specifically and games generally. I want to choose some key words that synthesize these ideas. Roguelikes excel through: exponential tension, challenging gameplay, robust internal logic, and emergent narrative. They are especially compelling because each play experience is coauthored by human designers and human-authored programs.

In future posts, we'll be discussing specific roguelikes (especially DCSS and Angband), games with a pervasive roguelike influence (FTL and The Binding of Isaac), and games with a philosophical basis in the roguelike genre (such as Dark Souls and Demon Souls).

We'll also talk about games that have nothing to do with the roguelike genre! My goal in this blog is to develop a critical language that advocates for the best and most interesting games while avoiding the pitfalls of the traditional review model. I also hope to move "serious" game criticism away from the developer-oriented model preferred by sites like Polygon and the Penny Arcade Report, which, though I do enjoy them, miss out on a lot of opportunities and insights because they focus on the personalities and intentions of game developers rather than the actual effects produced in players by games. That's what I really want to talk about: what games are, what they do, and how they operate in the lives of their players.