Regressive Progression

There’s a particular mechanic in Eternal Sonata that the game uses to increase the intensity of battles as a player progresses through the game. I’ve only played the demo, but from various reviews of the game, I’ve come to understand that the real-time battle system is very forgiving in the beginning, pausing the real-time clock to allow the players time to decide upon the appropriate course of action. As the game moves forward, the system becomes less and less forgiving, getting rid of the pause and allowing the players less time to input their actions. This method of “taking away” what players once had is an interesting means of gameplay progression. None of the reviews I read had any problem with this mechanic – in fact, they seemed to like it, even though it goes against the generally accepted method of designing game progression around the unlocking or expanding of player capabilities. Instead, as the game progresses, the player becomes more restricted. This almost seems counterintuitive to the psychological basis of fun that Daniel Cook presents in his Chemistry of Game Design, in which the learning and application of new skills causes the neural pleasure response.

While it’s a bit of a stretch to call Eternal Sonata’s mechanic a progressive “restriction” of player capabilities, the idea of this “regressive progression” is an interesting one nonetheless. I would love to see a game designed around this peculiar type of progression, but the biggest question is the one I’ve already posed above: how does one make such a backward experience compelling? I’ve already shown that the formula does not appear to be a formula for fun. Sidestepping the debate of whether games need to be fun at all (maybe I’ll chime in on that topic in a later post) and making the assumption that the player should have fun with this “regressive” game in question, the task that is left for the designer is to figure out how to provide the necessary stimuli to provide an enjoyable experience in the midst of or as a response to the loss of gameplay mechanics and player capabilities.

Here’s my theory of how such a game design could possibly reconcile itself with the predominating theories of “fun”. If, in fact, players derive enjoyment from the acquisition and application of new skills, then the only solution is to ensure that these skills being acquired are separate and distinct from the mechanical skills and capabilities manifested in the gameplay. Think of it first in terms of predominating game designs: generally, the acquisition of player skills correlates to the unlocking of a new gameplay mechanic – you start the game with a pistol, learning the basic skills of aiming, firing, and reloading; after you’ve been given enough time to familiarize yourself with the basic skillset and master it, the game suddenly drops more advanced weapons onto your lap – sniper rifles, machine guns, etc. – that require a bit more specialized mastery on top of the basic skillset. The introduction of a new “player skill” (i.e. specialized weapon tactics) is made to coincide with the introduction of a new “game skill” (the physical ability to wield a sniper rifle, for instance). You can see here two different “layers” of skills – those specifically allowed by the game (“game skills” – like new powers, abilities, weapons, etc.) and those internal to the mind of the player (“player skills”). It is the “player skills” that are ultimately responsible for the sense of fun that a player experiences. In the example above, the two types of skills are presented simultaneously. Let’s take a look at a slightly different type of game – so-called “skill-based” games. Players often come to enjoy these types of games over time due to the gradual learning of how to more effectively use the game skills they already have at their disposal. As an example of this, let’s look at a typical competitive, multiplayer FPS (as opposed to the single-player or plot-driven experience that I’ve described above). In this case, all of the weapons are already at your disposal – either in your inventory or littered around the map. As you play more and more games of this multiplayer FPS, the game itself does not change. In other words, the game skills available to the player are set in stone; new weapons are not suddenly unlocked after X games or after X victories (except in games like Battlefield 2 and Call of Duty 4). However, the player still derives enjoyment because his/her player skills are constantly evolving. The player is still learning that it’s not smart to stand still when facing an opponent, or that that one particular rooftop makes an ideal sniping spot, or that the medic-heavy invuln combo is so awesome (TF2 beta… yummy).

Since there’s an apparent distinction between game skills and player skills, and it’s clearly possible for the two types of skills to evolve independently of each other in any given game, then we can only deduce that it should be a possible task to design a game in which player skills are being learned while game skills are being lost. Since enjoyment is proportional to the upward development of player skills, the way in which game skills evolve over time (or devolve, in this instance) should have no bearing on that sense of enjoyment except in the way that game skill evolution affects player skill evolution. Here are some equations if what I just said didn’t make any sense:

U = utility/enjoyment; p = player skill metric (higher number corresponds to more obtained skills); g = game skill metric

p = f(g), where f is a function determined by the design

U = kp = kf(g), where k is a positive constant; if f(g) is strictly decreasing, then U can theoretically increase as g decreases.

The difficulty lies in designing a game in which f(g) is strictly decreasing. Okay, so maybe the equations didn’t really clarify anything for some of you. Basically, if players enjoy a game more by learning more player skills, then the trick is to design a game in which player skills are implicitly obtained or exercised by the loss of game skills. Here are a few guidelines that might help in this endeavor:

• skill basis: if all of the initial game skills are convenient shortcuts or combinations of a subset of those skills (the basis), then shedding all but the basis skills will allow the player to still perform the necessary actions throughout the game by using the more primitive basis skills in the proper sequence/combination. For instance, if a player suddenly loses the ability to turn right, the player would be forced to figure out that turning left for a long enough amount of time has the same effect as a right turn – a new player skill at the expense of a game skill. That’s a really sad example, but it’s the designer’s responsibility to ensure that as the non-basis skills are shed, it becomes a challenge in itself for the player to determine how the basis skills can be used in various combinations as a substitute for the lost skills. Let me try another example in the vein of a Half-Life-style game: maybe the character, instead of starting with just a crowbar, started with all of the weapons. He could easily use the rocket launcher to blow up various reinforced blockades blocking access to subsequent areas of the game. As the game progresses and he loses access to the rocket launcher, he instead mucks around and determines that he can knock a barrel over with his crowbar, roll it to a blockade, and blow it up with his pistol. While the game has taken away his game skill of “wielding an explosive weapon,” the basis skill of “causing an explosion” is still present in the game, albeit in a more complicated form. The player just needs to gather his wits (increasing his player skill) in order to figure that out.
• training wheels: one could view a skill basis from a simpler perspective. After a game’s skill basis has been determined, you can view the other skills layered on top as “training wheels” that are related to the basis skills. By this, I mean the “derived” skills are simply super easier versions of the basis skills. If the point of a basis skill is to “move”, then the derived skills can be anything from sprinting to teleporting – anything that makes moving extremely easy. If the point of a basis skill is to “do damage” (say, via simple punches), then the derived skills can be any kind of extravagantly animated attack that just does a lot more damage. Here’s a contemporary example of how this might be used: you’re a Viper pilot that has just encountered the Cylons – an ever-adapting, mechanical enemy. (Sound familiar?) While their ships were once susceptible to missile-lock and projectile guidance technology (i.e. auto-targeting), they’ve recently discovered a way to disable it. In effect, you’ve just lost your most reliable means of destroying their attacking ships. But there’s more than one way to skin a cat – you’ve still got the basic means (a basis skill) of dealing damage: your trusty guns. Now you have to adapt and hone your method of dogfighting to deal with the enemy with less technical capabilities at your disposal. The missiles were like training wheels – the experience is easier with them – in fact, it’s practically hand-holding – but you can make do without them. Training wheels are possibly the simplest example of designing a skill basis, and while they’re not necessarily the most creative (and fun) application of regressive progression, they could certainly be used as part of the overall initial skillset.
• conscious loss: it’d probably be a lot easier for a player to accept any loss of game skills if he/she were conscious of said loss before it occurred. The game should be designed to make it clear to the player when certain game skills will be lost. The easiest way to achieve this is in a design where the player chose the game skill to lose. So we see two possible approaches to game skill loss: predetermined loss progression and player-decided loss progression. It might also be interesting to see an adaptive or dynamic loss progression, in which the game determines which skill is being used most (or least) and force the player to shed that skill. For that kind of progression, it may not be necessary to tell the player which skill will be lost until it happens, but it would still be important to alert the player as to when that skill shedding will occur. Here’s a related idea for an application of this: an RTS title in which the races controllably regress according to their actions. Maybe if a unit is overly aggressive, it’s regression accelerates. Instead of the typical arms build-up, the player would have to balance the management of the regression with the overall goals of defeating the enemy. In this example, the players would be entirely conscious of the regression mechanics and would have to build their strategy around it.

Regressive progression really isn’t an extremely revolutionary idea. If you think about it, really, regressive progression could easily be interpreted as a means of difficulty progression, but instead of the enemies getting more powerful throughout the game, the player’s character is getting weaker. There are, in fact, a lot of games that use a regression mechanic in very brief stints. As in the example above, quite a few FPS games will put the player in a situation in which all of his weapons have been seized, forcing him to use the bare basics (melee attacks or a puny pistol) in order to progress and retrieve his weapons. Rainbow Six: Vegas and Bioshock both used this gimmick. Puzzle games with a puzzle mode will often introduce a level in the middle of the rest of the puzzles in which the player is not allowed to use some of the more advanced techniques but must instead rely on clever usage of one or a couple of basic techniques/pieces. These kinds of regressive puzzles are commonplace in games like The Incredible Machine, Lemmings, and Eets. Okami’s final boss goes into a phase in which it saps all of Amaterasu’s brush abilities. In Bioshock, there’s a point at which the antagonist begins reducing your maximum health and eve over time. But again, the mechanic is only used in a very brief part of these games. All that remains is to implement a game in which this loss of game skill becomes a more central and focal gameplay element instead of a minor distraction. It would’ve been interesting and altogether different if Amaterasu gradually lost her powers as she proceeded through the plot, culminating in a boss fight where those powers are returned to her for the final, climactic battle.

I also imagine regressive progression, aside from being a creative spin on standard gameplay progression, could have some utility in pushing a narrative theme more effectively. In Final Fantasy VII, Aeris’ death is one of the most commonly cited emotionally impacting game experiences. What if an RPG amplified that kind of experience by starting the player with a full party and slowly killing off party members, leaving only the player and one companion? What if the wanderer in Shadow of the Colossus actually lost health and stamina with the defeat of each colossus, his life being taken from him piece by piece each time that ominous cloud reaches from the fallen beast and pierces him. That seems to make more sense, doesn’t it? And what if, over time, he became too weak to shoot his bow and ride his horse, having to resort to more creative and desperate means to seek out and destroy his innocent victims. Wouldn’t that have been a more effective way of leading the story toward the wanderer’s inevitable, tragic demise? Overall, regression provides a very impactful narrative effect because it has the capacity to boil down the entirety of the game into an essential, core essence over time, leading to a purified resolution, in which the all of the components and complexity of the story is distilled into that one, overarching theme. In our RPG, the narrative theme could be the broken relationship between two of the characters who’ve hated each other throughout the game and are finally left alone to band together out of necessity for survival (a little cliche, sure, but I’m no storywriter). In our alternative Shadow of the Colossus, the theme could be the ironic savagery of the wanderer’s destruction.

Of course, regressive progression has at least a couple of potential snags that must be avoided. As I described above, one approach to the design is to offer a skill basis set to which the regression eventually reduces. One problem is that the player may decide to only use the skill basis from the beginning, even in the midst of the other available skills. (Why use the super-cool flying uppercut move when a normal jab will get the job done more quickly?) The designer must ensure that the derived skills are compelling enough when they are available. Another issue is the initial learning curve. There must be a way to ramp up the player’s familiarity with the available game skills (which usually involves the standard gameplay progression curve) without prolonging that familiarity phase for too long, which would dilute the regressive mechanic.

Overall, however, regressive progression provides some interesting possibilities – both in gameplay and in the narrative. A game doesn’t necessarily need to use it exclusively. A game could certainly have a hybrid of the two, in which remaining game skills are made more powerful or more effective as other game skills are lost. Or a game could also do a push-and-pull, in which game skills are being lost, after which new, different ones are gained to provide an alternative means of progressing through the game. It’s definitely something that should be explored in more depth and could lead to games that are very different from the “ever-more-powerful hero” formula of the games we’ve become accustomed to seeing.