On Permadeath

    This past weekend, while I was busy watching what was an underwhelming start to the year's NCAA basketball tournament, a few friends talked me into attempting what's called a Nuzlocke run of Pokemon Emerald. For the uninitiated, a Nuzlocke is a run of a Pokemon game in which you treat any knockout as a death -- that is to say, if a pokemon drops to zero health, it's treated as permanently dead, and you have to release it or permanently store it, not using it for the rest of the game. There are a few other rules (and a nearly limitless number of variant rulesets), but the permadeath constraint is key to the game's appeal.
    Nuzlocke runs are by nature time- and grinding-intensive, so I've avoided them in the past, but one particular comment intrigued me: specifically, that Nuzlocke runs give Pokemon a sense of stakes, something the series generally lacks outside of the unusual case that is Pokemon Mystery Dungeon. Given my complaints last week about Pokemon's stagnation and its dated gameplay, I thought committing to a Nuzlocke might give me a renewed appreciation for the series.
    Long story short, it did not.
    I'm glad to have completed most of a Nuzlocke, if only for the experience -- I stopped just before the game's ending sequence, as there's a significant jump in enemy level at the end of the game and I didn't feel like grinding for the final boss -- but it's not something I see myself doing again. It did, however, get me thinking about the concept of permadeath as a narrative and gameplay tool. It's a fairly rare thing and it always seems to be polarizing when it does appear, so I thought I'd delve into the ways in which it does and doesn't work.
    It's impossible to talk about permadeath, of course, without addressing Fire Emblem. FE is an old tactical RPG series that used to be known for its large casts and its permadeath mechanics. If a party member dies in battle in Fire Emblem, the character is permanently dead, and the story dialogue changes to reflect this. Fire Emblem Awakening and Fire Emblem Fates allowed players to turn this mechanic off, a decision which was met with mixed feelings. Longtime players felt that the very option of disabling this series stable undermined the purity of the games, while newer players felt the more forgiving "casual" mode made the games considerably more accessible to a broader audience -- especially important given the games' renewed focus on character appeal over raw strategic gameplay.
    The thing about permadeath in Fire Emblem is in most practical cases it doesn't work as advertised. Typically, players respond to character deaths not by proceeding through the game without the dead characters, but rather by reloading their saves and replaying the fights until they can be cleared without any deaths. In one sense, this serves to heighten the games' difficulty, as the effective win condition changes from "defeat the enemy" to "defeat the enemy without losing any allies." This makes the permadeath concept somewhat misleading. It doesn't raise the stakes of individual decisions so much as it slows progress through the game, and it also means bad luck can be much more of a factor in Fire Emblem than in looser tactical RPGs (such as Final Fantasy Tactics).
    What makes Nuzlocke runs of Pokemon games interesting is the commitment it forces on the part of the players. As Nuzlocke rulesets are self-imposed rather than mechanical, it feels against the spirit of the run to reload after an unfortunate fight. Fire Emblem's system-imposed permadeath has the counterintuitive effect of leading players to work around the imposed stakes rather than embracing them. An "honest" run of Fire Emblem would require never reloading from a prior save -- but this isn't realistic, and players could easily get stuck or discouraged after losing a few key units. The older games are balanced around permadeath in the sense that it's possible to win fights with no unit deaths, but they're not balanced around permadeath in the sense that later fights are balanced around you having a full team of powerful units. The games' narratives adjust for loss of character life, but gameplay-wise they do not.
    It's a clunky system, ultimately, and the decision in Awakening and Fates to allow players to disable permadeath feels to me like an acknowledgement that the mechanic doesn't work well -- or, at least, that it doesn't work the way it was initially intended to. It adds difficulty (and potentially frustration), but it does not create a sense of stakes.
    It's also worth noting that Pokemon games are not balanced around Nuzlocke runs, which means the Nuzlockes essentially boil down to "grind forever" and then "steamroll boss fights and hope you don't get unlucky." It's a reverse of Fire Emblem, essentially -- it definitely creates a sense of stakes, but it lacks any coherent difficulty curve, unless you consider the patience required a form of strategic challenge.
    This is not to say that permadeath is an inherently flawed mechanic. It can work well, but it requires two factors to be present: first, it must be presented such that the player voluntarily opts-in to the system and cannot easily "reset" to avoid character deaths, and second, the game must be balanced around the presence of the mechanic.
    Before I get to the main example I want to use for this, I'd like to address a strange DS game called Valkyrie Profile: Covenant of the Plume. For reference, I haven't played the game, but the way it handles permadeath is interesting enough that it occasionally gets brought up in discussions of the topic regardless. The game is a tactical RPG in the vein of Final Fantasy Tactics, and in-battle character deaths do not result in permanent death. However, the player can choose to temporarily strengthen party members during a battle in exchange for having them die permanently afterward. Permadeath in this case is not a penalty for strategic failure but rather a trade-off associated with a powerful strategic tool. It is entirely in the hands of the player and not subject to luck (a common frustration with Fire Emblem and with Nuzlocke runs), and it opens up interesting mechanical trades. You can, for example, choose to strengthen an underperforming party member for the duration of a particularly difficult boss fight, with the knowledge that after the character dies you can replace them with someone more powerful. It's a clever take on permadeath, though it's also not really what people usually mean when they talk about the mechanic.
    The best usage of a permadeath-like system I've seen in an RPG comes in Atlus's Devil Survivor games. These games are structured somewhat like Fire Emblem, but they approach permadeath on two separate-but-related fronts.
    The first and most important of these is narrative. Permanent character deaths happen not because of gameplay failures but rather as a result of narrative decisions the player makes over long periods of time. This creates the sense of stakes permadeath seeks to create, but it does so narratively instead of mechanically, and the drawn-out nature of these sequences of decisions makes them difficult to reset and fix. In Devil Survivor 2, for example, there is a character who is presented with a harrowing personal trial near the game's end. If the player has taken the time to speak with her and help her throughout the course of the game, she finds the mental and emotional strength to succeed and survive. If the player has ignored her, she lacks the mental fortitude she needs, and she fails and dies. The plot and dialogue change to accommodate character deaths, as in Fire Emblem, but the deaths are the result of accumulated player choices rather than freak dice rolls, which adds a significant weight to any failures.
    Additionally, the games contain a sort of mechanically-monitored equivalent to Pokemon's Nuzlocke runs. The Devil Survivor games each have a wide array of endings, and players are encouraged to replay the game five-or-so times to complete them all. Depending on how well the player does, the player is allowed to carry various tools over to the next run, making future playthroughs easier. If the player manages to go through the entire game without any in-battle character deaths, the player is rewarded with a considerable bonus in future runs.
    This means there is an incentive to play as if the game had permadeath, even though there are no direct consequences to losing characters. It allows and encourages players to add the additional level of strategic difficult Fire Emblem's permadeath system creates, but it doesn't punish players who choose not to do so, and its status as a strictly mechanical option (as opposed to Fire Emblem's ludonarrative hybrid) keeps the inevitable resets to avoid deaths from undermining the weight of the games' stories. Cool, no?
     Most importantly, Devil Survivor also rewards playing through without grinding, and as such the games are balanced around no-deaths, no-grinding runs. They're crazy hard games, as RPGs go, but they rely on true strategic difficulty rather than sheer numbers (as in Fire Emblem) or luck and patience (as in Shin Megami Tensei: Nocturne). Devil Survivor is the pinnacle of RPG difficulty that is legitimately challenging but also entirely fair. You have access to the exact same tools as your enemies do, and you play by the exact same rules, so the games become 20% preparation and 80% strategy. It's super fun, and the bulk of the game is spent in the actual meat of the fights rather than in the time-sink that is grinding.
    I'm generally an advocate for making gameplay forgiving and keeping stakes to storytelling. By "forgiving" I don't mean easy, either -- I mean forgiving in the sense that games shouldn't heavily penalize failure. Look to Super Meat Boy as an example of this -- it's a difficult platformer, but there's no penalty for failure. Players are encouraged to try again as many times as needed, which keeps the difficulty from feeling frustrating or unfair. Severe penalties for death are more likely to breed frustration than to create a legitimate sense of challenge, and except in the most well-balanced of games permadeath generally falls into that mix.
    It's entirely possible to make character death feel meaningful and fights feel dangerous without actually punishing the player for failure, too. I have a love-hate relationship with Dark Souls, but it's excellent at this. For all its difficulty it's a highly forgiving game, but even though death tends to only set players back a minute or two close calls and dangerous fights inevitably lead to the burst of adrenaline you might feel in a game where a death can mean hours of lost progress. The game accomplishes this not through practical stakes so much as through aesthetic ones -- its foes are visually imposing, and its stellar sound design creates psychological tension without artificial mechanical difficulty. If death in Dark Souls meant more significant losses that its small experience penalty -- if, for example, you lost all of your equipment when you died -- this would make failure more frustrating, but it wouldn't meaningfully raise the sense of stakes. Strange as it seems, gameplay stakes are best created through narrative means rather than through mechanical methods.
    Ultimately, permadeath is somewhat of a dangerous game, especially when it isn't predicated on player buy-in. It can be done well, but it usually isn't. It makes games more difficult, but difficulty is better achieved through careful balance and strong enemy design than through mechanical quirks like permadeath. Like any other mechanic, it has its place, but it shouldn't be used thoughtlessly.