Hills and Lines: Final Fantasy XIII

Simon Ferrari


Final Fantasy has always been a JRPG (Japanese Roleplaying Game) about a group of dynamic, known mathematical values coming together in unexpected ways to tackle a static, unknown mathematical value. The former is the team of player characters, and the latter is the enemy. The major difference between Final Fantasy XIII and every past entry in the series is that it harbors no illusion that it is about anything else.1 Final Fantasy XIII is not a story about two worlds, Pulse and Cocoon, standing in opposition. It’s a process of blindly ascending hills, hills carefully placed one after the other in a line to make sure that the climber always has what she needs to make it to the top of the next in sequence. And I can tell you, as someone who lived most of his life in the foothills of Appalachia, that Final Fantasy XIII is as good as climbing hills gets.


Randomized encounters can never be as precise as explicitly designing them. Within the history of the Final Fantasy series, two constraints are placed on how random encounters work. First, zones in the world or dungeon map are delineated, and only certain enemies can spawn within those zones based on the probabilities of occurrence and volume---in the opening area of the first Final Fantasy I might have a fifty percent chance to run into 3-4 goblins, a thirty-five percent chance of two slightly stronger wolves, and a fifteen percent chance of a powerful but solitary nightmare.2 Second, these encounters can be limited by the size of the monster relative to the size of the combat screen or zone. Two dragon-type characters might take up enough room on the combat grid (in 2d) or circle (in 3D) to prevent the occurrence of any of other enemy.

Final Fantasy XIII doesn’t randomize encounters, and the player sees every threat on the map. Each encounter thus becomes a conscious choice to confront the enemy head on, to sneak up on the enemy for a preemptive strike, or to run past the enemy. What is the value of randomizing encounters over making threats visible on the map? Variety is the spice of virtual life, or so the thinking goes. The two constraints detailed above generate a modestly robust amount of difference. The downside of randomizing is that it becomes harder to account for the player’s skill level (in fact, it seems strange that more JRPGs don’t attempt to numerically gauge this somewhat intangible property). This means that a system initially set in place to provide variety often ends up creating a grinding experience---one trades the player’s time for greater configurability.

“Grind” has become a naughty word in the wake of the Everquest widow problem.3 Some Western roleplaying games have attempted to deal with this problem through adaptive difficulty. But this fix has its own problem: the elimination of any serious challenge to the player. In a Bethesda or BioWare game, enemies simply take longer to kill as the game wears on. The player is never pressured to develop novel strategies or skills. A new time sink appears to replace the old one, and, considering the amount of people who claim to enjoy the gentle massage of the grind, it is unclear where the moral high ground for designers might be. Final Fantasy XIII does away with these problems altogether by compelling its players through a tightly designed obstacle course. Its literacy model is not built into the hundreds of tutorial and help screens; rather, it resides in the carefully staged progression of combat encounters.

Final Fantasy XIII’s “paradigm” system is a natural combination of the series’ earlier “outfit”-based systems in Final Fantasy V and X-2 and the “gambit” system of XII. These are the six combat roles, roughly in order of importance: Saboteur (debuffer), Commando (primarily melee), Ravager (primarily magic), Medic (healer), Synergist (buffer), and Sentinel (tank). The player only controls the team lead, while the two other active party members always take the optimal action given the party’s current state and known information about the enemy. Magic doesn’t cost mana as it does in most games; it is simply an attack type, limited only by the time it takes to charge the action bar. Each party member is good at three roles, so there are around four viable party makeups (a combination of Fang, Lightning, and Hope being the most versatile). There are five slots for paradigms, which the player can customize in between battles. While in battle, the player can execute a “paradigm shift” to any of those five predetermined combinations.

The object of any battle is to “stagger” an enemy. When a Ravager inflicts damage, a yellow stagger bar slowly fills. Filling the bar both increases damage to the enemy and brings it closer to a stagger state, which makes it more vulnerable to afflictions and allows a Commando to launch it into the air (rendering it unable to attack or defend). Saboteurs and Commandos are the most important party members, because the stagger bar actually decreases over time. An attack from either of the two will slow down the speed at which this bar decreases. Many enemies, especially bosses, can only be significantly harmed while staggered. No battles actually require the use of a Sentinel, and the Synergist exists only to speed battles up. Many battles can be won without pausing to heal, but the Medic is almost always required for any key encounter.

Earlier I said Final Fantasy XIII is about climbing hills blindly. We’re now ready to understand what the two elements of this statement mean. First (“climbing hills”), the carefully staged progression of encounters steadily elevates challenge while teaching the player how to kill each enemy. A level will begin, say, with two soldiers, and then it will add a third soldier to the next encounter. Then the player will encounter, say, two slimes or a larger enemy such as a behemoth. After these smaller hills have been ascended, the final battle before a checkpoint will combine those enemy types: three soldiers and two slimes, or three slimes and a behemoth, etc. By slowly adding challenges and then combining different types of challenges, the game tests the tipping point where the player has to finally change her dominant strategy and develop a new cycle of paradigm shifts.

Second (“blindly”), every new enemy the player encounters has a data sheet explaining its strengths and weaknesses. This sheet always begins blank. When an enemy uses a special attack, one of its strengths gets entered into the data sheet. As the player damages the enemy with magic and melee, its weaknesses gradually become visible. Filling out the data sheet is vital, because AI teammates act on the best available information. The player can also spend a special, limited resource called “technical points” to use Libra. The player can only ever have five TP, and Libra costs one. These points are also used for summons and to revive the entire team in the event of catastrophic loss. Libra is a shortcut to the natural, gradual discovery process; it automatically tells the player most of the enemy’s weaknesses.

In past Final Fantasy games, Libra is a spell just like any other. I can distinctly remember never using it as a child playing Final Fantasy I-VII. If an enemy was aquatic, the player would assume that lightning spells worked best. If an enemy had a reflective barrier or can absorb fire damage, the player found that out naturally within the first few rounds of the battle. By making Libra a special ability, by separating it from all other spells, Final Fantasy XIII makes an argument about the essence of its system that was probably true of the series all along: the game is a matter of finding an enemy’s weakness and exploiting it. This isn’t a groundbreaking realization, and it isn’t a unique way to build a signature combat system. Final Fantasy XIII’s beauty lies not in innovation but in its minimalism, purity, and transparency. It recognizes its genealogy and invites the player to study it.

The purity of Final Fantasy XIII cannot be overstated. Absent are many traditions of the genre, such as conversation with NPCs, a world map, and villages to visit. Those subsystems that do remain---treasure hunting, weapon upgrading, and shopping---exist as options to help along players of lesser skill. They stand in for a difficulty slider and for the need to grind. A player who lets the game teach her how it works need not upgrade a single weapon or even open a single treasure chest. Experience points are still important for upgrading basic skills and attributes, but the player doesn’t need to stop at any point to harvest them. Summon spells, a staple of the series, have lost their ability to turn the tides of a battle. Instead, each character in the game must at some point confront the summon beast (called an Eidolon) within. These battles, perhaps the most difficult in the game, serve primarily to teach the player how to think about upcoming boss fights. The Eidolon are depicted as vehicles (horses, airplanes, motorcycles) for player characters within the game, while for the player they are vehicles for more nuanced knowledge about the battle system.

Final Fantasy XIII argues that no player should be left behind, that no hill should prove impossible to ascend assuming a modicum of critical thinking. In order to make good on its dedication to teaching the player, it features incredibly little of what Jesper Juul calls “setback punishment.”4 Whenever a player fails a battle, she will emerge with full health right in front of the encounter that felled her. After each battle, the entire party regains full health. Figure 1 shows what Final Fantasy XIII (left) feels like compared to Demon’s Souls, a game that surprised mainstream players in 2009 with its harsh difficulty curve (right).5

Black lines represent progress without death. Red lines indicate time spent on a failed attempt at some segment of the game. Final Fantasy XIII proves that “hard” is not “the new good.” Gentle games have just as much to offer us as brutal games do. Difficulty, like everything else about a game, serves a distinct expressive purpose. Painstakingly clawing one’s way up a mountain isn’t “better” than joyously bounding over a hill. The experiences are simply different in their demands and reward structures.


The first twenty hours of this game ask the player to follow a straight line toward a checkpoint. At intervals of fifty to one hundred paces, a group of enemies awaits. Floating treasure chests are placed after every fourth or fifth group of enemies. This corridor, perhaps the longest unbroken span of narrow, unilinear space in videogame history, makes one realize something that was true of Final Fantasy games all along: we’ve always been running in a straight line. The decision to explore or not, represented here by the floating chests, has always been a matter of whether or not any given player is the kind of person who welcomes momentary distraction.

It has become increasingly common to see others criticize linear games for their linearity, without any effort to discern what the difference between good and bad linearity might be. An example of engaging linear space is the train-hopping sequence in Uncharted 2. The modularity of a train lends itself to constrained difference. The designer of the level has a few binary values to select for any given car: is it open or covered, is it a platforming challenge or a combat challenge (the latter being further divided between assault and stealth), is the arrangement of obstructions symmetrical or asymmetrical, and, if the car is covered, can its roof be reached and traversed?

Once each of these binary values has been determined for the individual car, one must arrange relationships between each car in the string. This creates a rhythm, which can be punctuated by unique scripted events---the helicopter, the “boss,” and the heavy gunner on the log. All of this goes into describing what amounts to nothing more than a line, and a line is in no way deprecated by the fact that games can, as computational works, support other lines (and an opportunity for the player to choose between them) if its designers want them to. One of the values of identifying core pleasures of a medium in the first place---agency, immersion, and transformation in Murray’s original account---is that the withholding of these pleasures can be used for the purposes of creating challenge, intrigue, variation, or expression.

Once one understands what a good line looks like, it becomes much easier to see why the first twenty hours of Final Fantasy XIII constitute a rather boring line---structurally speaking. There is no reason to create obstructions within, or alternate paths through, this space, because interacting with space isn’t a value or strength of the JRPG. Environmental puzzles have always felt strange within the genre, especially in games featuring random encounters. Nothing is more tedious than trying to figure out how to shove a boulder from one end of a cavern to another with enemies interrupting every five paces. Golden Sun might be seen as the peak of poor JRPG spatial design, with its absurd reliance on pillar-pushing puzzles and point-and-click adventure guesswork.7 In the context of some JRPGs, environmental manipulation makes sense. These are almost always games with such a large cast of playable characters that splitting them into groups for solving interlocking puzzles in key dungeons provides an engaging diversion from standard play.

This works in the case of Final Fantasy VI’s Phoenix Cave and final dungeon.8 The encounter rate on enemies was low enough, and the cast massive enough, that dividing the heroes into three parties to solve puzzles made sense. There was also a limited variety of puzzles that changed things up without being too confusing: the player could either push a pillar or pull a switch, which might trigger the shifting of a platform or the dispersal of lava. The same party-dividing conceit doesn’t work in the more contemporary Lost Odyssey, which features a smaller cast, only two party-dividing and puzzle-solving dungeons, and only has one puzzle type, which we might call “push the transporter over the cliff.”9 Final Fantasy XIII features a small cast of characters; it splits the party up for a while, but the player can’t switch back and forth between them; thus, environmental puzzles have no place in the game.

Final Fantasy XIII was released in Japan at around the same time that Mass Effect 2 was released worldwide. It should come as no surprise that both of these series transitioned from previously multilinear level designs to these unilinear, non-interactive corridors. For years, the makers of this kind of game were told that they needed to embrace the computer’s ability to produce nonlinear game spaces. “Open” worlds of various quality proliferated, and players received hours and hours of “content” defined by the exploration of structureless, monotonous space. Everyone quickly realized that, perhaps, not every genre needs to maximize every affordance of the digital medium. This particular brand of stat-crunching, combat-focused game works just as well in a corridor as it does in a sandbox. It is also possible that many designers weren’t ready to leave the comfort of the line; designing a nonlinear space demands knowledge of the line in much the same way that abstract painting demands knowledge of representation.

And that’s the realization that the designers at BioWare ended with when they sat down to design Mass Effect 2—stripping away its “open world” hub spaces and replacing it with a number of smaller, more distinct portals to linear missions.10 The designers of Final Fantasy XIII, on the other hand, took the realization one step further: the shape of their game spaces could be used as spatial allegories. In the melodrama tacked onto this brilliant game about blindly ascending hills, two worlds (and the two factions of demigods ruling over them) called Cocoon and Pulse exist in perfect opposition. Cocoon, ruled over by the Fal’cie of the Sanctum, is a bounded sphere where humans are simultaneously provided for and controlled in every conceivable way. Hannah Arendt would identify it as the ideal centripetal totalitarian state, one in which the government controls a populace by dominating rather than destroying its public space.11

Pulse exists outside Cocoon, or below it, or around it---exactly what their spatial relationship to each other might be is vague, but Cocoon appears to be some sort of moon orbiting the planet of Pulse. The Pulse Fal’cie determined that their world would be a laissez-faire one. It is simultaneously beautiful and deadly, a place where human civilization collapsed while demigods and beasts roam free. Pulse’s absolute freedom, though, is a farce. The Fal’cie of Pulse exert a centrifugal totalitarian control, the manipulation of their human servants through the destruction of a shared public space. In the minds of Square Enix’s English localization team, this world connoted Australia. They probably did this because of Pulse’s geography, natural beauty, and extreme wildlife. By doing so, they happened to connect the divine management of Pulse to the troubled history of British imperialism. Fang’s and Vanille’s brands become convict stains.

This story of two worlds is overwrought. One can only be thankful that it is only half-delivered by Final Fantasy XIII’s myriad cutscenes. The other half of the story must be gleaned from information files generated after each cutscene. Or, if the player is smart, it can all be ignored; every cutscene can be skipped, every data file left unread. Grasping the conflict between Cocoon and Pulse requires neither video nor text, because their respective spaces structure play in a way that lets us experience the difference between them firsthand.

Reviewers of Final Fantasy XIII remarked that the game “gets better” or “truly begins to shine” when the player hits the 20-hour mark. That’s when the player transitions from Cocoon to Pulse. We trade a series of stifling hallways for a wide, open world driven by the kind of hunting quests that dominated Final Fantasy XII. In Pulse, it is somewhat difficult to find one’s way to a definite goal. Many enemies will instantly kill the player’s team on being engaged. The literacy model carefully constructed throughout the first half of the game flies out the window. Instead the player is left to fend for herself, to pick her battles and hope for the best. She has left a world where everything a human needs is provided by divine stewards, entering another where the demigods have decided to let natural selection reign. Figure 2 is a phenomenological map of the spatial difference between the two.

One can forgive reviewers for not understanding how carefully the distinction between the two worlds has been constructed. We are, after all, trained to make judgments about a game world from its story. In the case of the Final Fantasy series, we’ve come to expect this story to be delivered through elaborate cutscenes. And the cutscenes in Final Fantasy XIII tell players little about these two worlds. We might ask a negative reviewer: “How would you, without words, convey the feeling of living one’s entire life on a string?” We can accept that it isn’t necessarily fun to be forced to endure twenty hours of running in a straight line just to have a fairly simple truth bestowed upon us, but deep play demands deep structure. That so many have complained about the game’s linearity is a sure sign of the design’s success. Life in Cocoon is something worth complaining about.

Notes on method:

This critique is my first effort to use the framework developed in my Master’s thesis, “The Judgement of Procedural Rhetoric.” It contends that Ian Bogost’s theory of procedural rhetoric describes the primary meaning-making structure of games—that a game uniquely makes an argument or expression through rules alone—while diverging from Bogost’s formulation of the notion in two significant ways.12 First, I argue that the construction of space has largely been misunderstood as a layer of contextual information rather than a set of spatial, ludic rules.13 Second, that procedural rhetoric takes place not only from a designer’s perspective—where one attempts to drive players toward a desired experience or argument—but also from the player’s perspective. A critical player, armed with the proper analytic method, can articulate which rules mean what to her while ignoring others as personally non-meaningful.14

The method set forth in my thesis assumes that the non-meaningful rules of a game derive from its genre substrate. These generic rules are a form of literacy that serve to integrate players into a game’s core mechanical loop. Designers can push new aesthetic experiences to players by incorporating new rules and spatial structures into play, subverting conventions and inspiring a player’s reflection on divergence from the genre. My introductory example of this principle comes from the Call of Duty series, where infinite enemy respawns limited by a trigger point convey the futility of “eliminating” enemy numbers: enemies spawn continuously into a space until a pre-defined boundary is passed, at which point a new set of respawn locations perpetuates the experience of constant threat.15 This rule constitutes a significant divergence from the genre of run-and-gun shooters.

We can also separate a game’s rules into material and conventional constraints. In an analog game, most constraints are conventional. They are enforced by the play community, and they afford both cheating and spoil-sporting as described by Huizinga.16 A videogame, on the other hand, is composed primarily of material constraints. One cannot change the numerical values and algorithms structuring its play experience without resorting to hacking or modding, which are meta-gaming practices quite different and more technically involved than traditional cheating. Some of the only conventional constraints that persist for a videogame are endgame goals: from speedrunning, to iron-manning, to generally deciding when a play experience is sufficient and complete.

I would argue that my analytic method uniquely suits two implicit goals of the Well-Played series: that some exceptional games demand close readings, and that these close readings are demonstrably unique for all possible players. For instance, in my reading of Final Fantasy XIII, I purposefully avoided two significant portions of the game that lend themselves to generic play. In the case of FFXIII’s genre, these conventional sections imply grinding. Both the middle portion of the game and it’s post-story play experience afford grinding for achieving the side goal of hunting enemies of difficulty greater than those found in the main mission. I would describe these, in my geographical analogy, as mires dotting the hills and lines of my phenomenological experience.

My Final Fantasy XIII is a game about numbers. It asks me to blindly ascend a sequence of hills, in a line, up until the point where it sets me free. It does so for the expressive purpose of make me experience firsthand the difference between total determination and complete freedom. Final Fantasy XIII teaches me how to gauge the strengths and weakness of each type of enemy, then it asks me to adapt to various arrangements of different kinds of enemies. It argues that the Final Fantasy series has always been about this sensing process, the conflict between known, dynamic numerical values (the heroes) and a single, static number (the enemy). Final Fantasy XIII is a game that eschews grinding, adaptive difficulty, and a difficulty slider. Instead, it argues that the traditional subsystems of treasure hunting, weapon upgrading, and shopping should exist only to help players with lesser skill. Anyone who lets the game teach her how it works needs none of these. Its possibility space is narrow, as much a series of puzzles as it is a game. But it’s a good series of puzzles.